20170922Attachment 3-2A - 2017 Electric IRP Final.pdf

2017 Electric Integrated
Long Lake Dam
SC_PR_3-2 Attachment A Page 1 of 205
Safe Harbor Statement

This document contains forward-looking statements. Such statements are subject to a
variety of risks, uncertainties and other factors, most of which are beyond the Company’s
control, and many of which could have a significant impact on the Company’s operations,
results of operations and financial condition, and could cause actual results to differ
materially from those anticipated.

For a further discussion of these factors and other important factors, please refer to the
Company’s reports filed with the Securities and Exchange Commission. The forward-
looking statements contained in this document speak only as of the date hereof. The
Company undertakes no obligation to update any forward-looking statement or
statements to reflect events or circumstances that occur after the date on which such
statement is made or to reflect the occurrence of unanticipated events. New risks,
uncertainties and other factors emerge from time to time, and it is not possible for
management to predict all of such factors, nor can it assess the impact of each such factor
on the Company’s business or the extent to which any such factor, or combination of
factors, may cause actual results to differ materially from those contained in any forward-
looking statement.

SC_PR_3-2 Attachment A Page 2 of 205
Production Credits

Primary Avista 2017 Electric IRP Team

Individual Title
Clint Kalich Manager of Resource Planning & Analysis
James Gall IRP Manager
John Lyons Senior Resource Policy Analyst
Grant Forsyth Senior Forecaster & Economist
Richard Maguire System Planning Engineer

2017 Electric IRP Contributors

Name Title
Thomas Dempsey Manager, Generation Joint Projects
Tom Pardee Natural Gas Planning Manager
Amber Gifford DSM Planning and Analytics Manager
Ryan Finesilver DSM Analyst
Jeff Schlect Senior Manager of FERC Policy and Transmission Services
Dave Schwall Senior Engineer
Darrell Soyars Manager of Corporate Environmental Compliance
Xin Shane Senior Power Supply Analyst
Debbie Simock Senior External Communications Manager
Jason Graham Mechanical Engineer

Contact contributors via email by placing their names in this email address format:
[e-mail]

SC_PR_3-2 Attachment A Page 3 of 205
Avista Corp 2017 Electric IRP
2017 Electric IRP Introduction

Avista has a 128-year tradition of innovation and a commitment to providing safe,
reliable, low-cost, clean energy to our customers. We meet this commitment
through a diverse mix of generation resources.

The 2017 Integrated Resource Plan (IRP) continues this legacy by looking 20 years into
the future to determine the energy needs of our customers. The IRP, updated every two
years, analyzes and outlines a strategy to meet the projected demand and renewable
portfolio standards through energy efficiency and a diverse mix of renewable and
traditional energy resources.

Summary
The 2017 IRP shows Avista has adequate resources between owned and contractually
controlled generation, combined with conservation and market purchases, to meet
customer needs through 2026. In the longer term, plant upgrades, energy efficiency
measures, solar, demand response, energy storage and additional natural gas-fired
generation are integral parts of Avista’s 2017 Preferred Resource Strategy.

Changes
Major changes from the 2015 IRP include:
 The 2017 Expected Case energy forecast grows 0.47 percent per year, replacing
the 0.6 percent annual growth rate in the last IRP.
 Peak load growth is lower than energy growth, at 0.42 percent in the winter and
0.46 percent in the summer.
 Lower expected load growth combined with recent Mid-Columbia hydroelectric
contracts, energy efficiency, and demand response delay the need for additional
resources from the end of 2020 until 2026.
 The return of demand response (temporarily reducing the demand for energy)
and the addition of energy storage and solar.
 Lower expected emissions from Avista owned and controlled resources with
fewer natural-gas fired peaking plants and no new combined-cycle plants.

Highlights
Some highlights of the 2017 IRP include:
 Avista’s current generation resources remain cost effective and reliable sources
of power to meet future customer needs over the next 20 years.
 Energy storage costs are significantly lower than the last IRP which for the first
time makes the technology operationally attractive in meeting energy needs in
the 20-year timeframe of the 2017 IRP.
 Avista is working to construct a 15 MW (DC) solar facility for the company’s new
Solar Select Program for commercial and industrial customers.
 This study estimates conservation will serve 53.3 percent of future load growth.

SC_PR_3-2 Attachment A Page 4 of 205
Avista Corp 2017 Electric IRP
IRP Process
Each IRP is a thoroughly researched and data-driven document that identifies and
describes a Preferred Resource Strategy to meet customer needs while balancing costs
and risk measures with environmental and other policy mandates. Avista’s professional
energy analysts use sophisticated modeling tools and input from over 100 invited
participants to develop each plan. The participants in the public process include
customers, academics, environmental organizations, government agencies,
consultants, utilities, elected officials, state utility commission stakeholders and other
interested parties.

Conclusion
This document is mostly technical in nature. The IRP has an Executive Summary and
chapter highlights at the beginning of each section to help guide the reader. Avista
expects to begin developing the 2019 IRP in mid-2018. Stakeholder involvement is
encouraged and interested parties may contact John Lyons at [phone] or
[e-mail] for more information on participating in the IRP process.
SC_PR_3-2 Attachment A Page 5 of 205
Table of Contents

Avista Corp 2017 Electric IRP

i
Table of Contents

1. Executive Summary ...................................................................................................... 1-1
Resource Needs ....................................................................................................................... 1-1
Modeling and Results ............................................................................................................... 1-2
Electricity and Natural Gas Market Forecasts .......................................................................... 1-2
Energy Efficiency Acquisition ................................................................................................... 1-3
Preferred Resource Strategy ................................................................................................... 1-4
Energy Independence Act Compliance .................................................................................... 1-5
Greenhouse Gas Emissions .................................................................................................... 1-6
Action Items .............................................................................................................................. 1-7
2. Introduction and Stakeholder Involvement ................................................................ 2-1
IRP Process ............................................................................................................................. 2-1
2017 IRP Outline ...................................................................................................................... 2-4
Regulatory Requirements ........................................................................................................ 2-6
3. Economic & Load Forecast .......................................................................................... 3-1
Introduction & Highlights .......................................................................................................... 3-1
Economic Characteristics of Avista’s Service Territory ............................................................ 3-1
IRP Long-Run Load Forecast ................................................................................................ 3-13
Monthly Peak Load Forecast Methodology ............................................................................ 3-19
Simulated Extreme Weather Conditions with Historical Weather Data ................................. 3-20
Extreme Temperature Analysis .............................................................................................. 3-24
4. Existing Supply Resources .......................................................................................... 4-1
Introduction & Highlights .......................................................................................................... 4-1
Spokane River Hydroelectric Developments ........................................................................... 4-2
Clark Fork River Hydroelectric Development ........................................................................... 4-4
Total Hydroelectric Generation ................................................................................................ 4-4
Thermal Resources .................................................................................................................. 4-5
Power Purchase and Sale Contracts ....................................................................................... 4-6
Customer-Owned Generation .................................................................................................. 4-9
Solar ....................................................................................................................................... 4-11
5. Energy Efficiency & Demand Response ..................................................................... 5-1
Introduction ............................................................................................................................... 5-1
The Conservation Potential Assessment ................................................................................. 5-2
Overview of Energy Efficiency Potential .................................................................................. 5-5
Conservation Targets ............................................................................................................... 5-7
NPCC’s Seventh Power Plan Benchmarking ........................................................................... 5-8
Energy Efficiency-Related Financial Impacts ......................................................................... 5-11
Integrating Results into Business Planning and Operations .................................................. 5-11
Conservation’s T&D Deferral Analysis ................................................................................... 5-13
Generation Efficiency Audits of Avista Facilities .................................................................... 5-14
Demand Response ................................................................................................................. 5-16
6. Long-Term Position ....................................................................................................... 6-1
Introduction & Highlights .......................................................................................................... 6-1
Reserve Margins ...................................................................................................................... 6-1
Energy Imbalance Market ........................................................................................................ 6-6
Balancing Loads and Resources ............................................................................................. 6-6
Washington State Renewable Portfolio Standard .................................................................... 6-9
7. Policy Considerations ................................................................................................... 7-1
Environmental Issues ............................................................................................................... 7-1
Avista’s Climate Change Policy Efforts .................................................................................... 7-3
8. Transmission & Distribution Planning ........................................................................ 8-1
Introduction ............................................................................................................................... 8-1
Avista Transmission System .................................................................................................... 8-1
Transmission Planning Requirements and Processes ............................................................ 8-3
SC_PR_3-2 Attachment A Page 6 of 205
Table of Contents

Avista Corp 2017 Electric IRP

ii
Annual Transmission Planning Report ..................................................................................... 8-5
IRP Generation Interconnection Options ................................................................................. 8-6
Distribution Planning ................................................................................................................ 8-7
9. Generation Resource Options...................................................................................... 9-1
Introduction ............................................................................................................................... 9-1
Assumptions ............................................................................................................................. 9-1
Natural Gas-Fired Combined Cycle Combustion Turbine ........................................................ 9-3
Hydroelectric Project Upgrades and Options ......................................................................... 9-14
Thermal Resource Upgrade Options ..................................................................................... 9-16
Ancillary Services Valuation ................................................................................................... 9-17
10. Market Analysis ........................................................................................................... 10-1
Introduction ............................................................................................................................. 10-1
Marketplace ............................................................................................................................ 10-1
Fuel Prices and Conditions .................................................................................................... 10-6
Greenhouse Gas Emissions and the Clean Power Plan ..................................................... 10-10
Risk Analysis ........................................................................................................................ 10-12
Market Price Forecast .......................................................................................................... 10-19
Scenario Analysis ................................................................................................................. 10-27
11. Preferred Resource Strategy ...................................................................................... 11-1
Introduction ............................................................................................................................. 11-1
Supply-Side Resource Acquisitions ....................................................................................... 11-1
Resource Deficiencies............................................................................................................ 11-2
Preferred Resource Strategy ................................................................................................. 11-7
Efficient Frontier Analysis ..................................................................................................... 11-13
Determining the Avoided Costs of Energy Efficiency ........................................................... 11-17
Determining the Avoided Cost of New Generation Options ................................................. 11-18
12. Portfolio Scenarios ...................................................................................................... 12-1
Introduction ............................................................................................................................. 12-1
Colstrip Scenarios .................................................................................................................. 12-2
Other Resource Scenarios ................................................................................................... 12-10
Washington State Emission Goal Analysis .......................................................................... 12-14
13. Action Items ................................................................................................................. 13-1
Summary of the 2015 IRP Action Plan................................................................................... 13-1
2017 IRP Two Year Action Plan ............................................................................................. 13-3

SC_PR_3-2 Attachment A Page 7 of 205
Table of Contents

Avista Corp 2017 Electric IRP

iii
Table of Figures

Figure 1.1: Load-Resource Balance—Winter Peak Load & Resource Availability ...................... 1-1
Figure 1.2: Average Mid-Columbia Electricity Price Forecast ...................................................... 1-2
Figure 1.3: Stanfield Natural Gas Price Forecast ......................................................................... 1-3
Figure 1.4: Annual and Cumulative Energy Efficiency Acquisitions ............................................. 1-3
Figure 1.5: Efficient Frontier ......................................................................................................... 1-5
Figure 1.6: Avista’s Qualifying Renewables for Washington State’s EIA ..................................... 1-6
Figure 3.1: MSA Population Growth and U.S. Recessions, 1971-2016 ....................................... 3-2
Figure 3.2: Avista and U.S. MSA Population Growth, 2007-2016 ................................................ 3-3
Figure 3.3: MSA Non-Farm Employment Breakdown by Major Sector, 2016 .............................. 3-4
Figure 3.4: Avista and U.S. MSA Non-Farm Employment Growth, 2007-2016 ........................... 3-4
Figure 3.5: MSA Personal Income Breakdown by Major Source, 2015 ....................................... 3-5
Figure 3.6: Avista and U.S. MSA Real Personal Income Growth, 1970-2013 ............................. 3-6
Figure 3.7: Forecasting IP Growth................................................................................................ 3-9
Figure 3.8: Industrial Load and Industrial (IP) Index .................................................................. 3-10
Figure 3.9: Population Growth vs. Customer Growth, 2000-2016 ............................................. 3-11
Figure 3.10: Forecasting Population Growth .............................................................................. 3-12
Figure 3.11: Long-Run Annual Residential Customer Growth ................................................... 3-16
Figure 4.1: 2018 Avista Capability & Energy Fuel Mix ................................................................. 4-1
Figure 4.2: Avista’s Net Metering Customers ............................................................................. 4-10
Figure 5.1: Historical Conservation Acquisition (system) ............................................................. 5-2
Figure 5.2: Analysis Approach Overview ..................................................................................... 5-3
Figure 5.3: Achievable Conservation Potential Assessment (20-Year Cumulative) .................... 5-6
Figure 5.4: Washington Annual Achievable Potential Energy Efficiency (Megawatt Hours)........ 5-7
Figure 5.5: 2017 Avista CPA / Seventh Power Plan Benchmark Comparison ............................ 5-9
Figure 6.1: Winter One-Hour Capacity Load and Resources ....................................................... 6-7
Figure 6.2: Summer One-Hour Capacity Load and Resources ................................................... 6-7
Figure 6.3: Annual Average Energy Load and Resources ........................................................... 6-9
Figure 8.1: Avista Transmission System ...................................................................................... 8-1
Figure 8.2: Avista 230 kV Transmission System .......................................................................... 8-2
Figure 8.3: Avista Transmission System Planning Regions ......................................................... 8-3
Figure 8.4: NERC Interconnection Map ....................................................................................... 8-4
Figure 9.1: Northwest Wind Project Levelized Costs per MWh ................................................... 9-7
Figure 9.2: Solar Nominal Levelized Cost ($/MWh) ..................................................................... 9-9
Figure 9.3: Historical and Planned Hydro Upgrades .................................................................. 9-14
Figure 9.4: Storage’s Value Stream ........................................................................................... 9-18
Figure 10.1: NERC Interconnection Map ................................................................................... 10-2
Figure 10.2: 20-Year Annual Average Western Interconnect Energy ........................................ 10-3
Figure 10.3: Resource Retirements (Nameplate Capacity) ....................................................... 10-4
Figure 10.4: Cumulative WECC Generation Resource Additions (Nameplate Capacity) .......... 10-5
Figure 10.5: Henry Hub Natural Gas Price Forecast .................................................................. 10-7
Figure 10.6: Northwest Expected Energy ................................................................................... 10-9
Figure 10.7: Regional Wind Expected Capacity Factors .......................................................... 10-10
Figure 10.8: Historical Stanfield Natural Gas Prices (2004-2015) ........................................... 10-12
Figure 10.9: Stanfield Annual Average Natural Gas Price Distribution .................................... 10-13
Figure 10.10: Stanfield Natural Gas Distributions .................................................................... 10-14
Figure 10.11: Stanfield Natural Gas Annual Price Statistical Comparison............................... 10-14
Figure 10.12: Wind Model Output for the Northwest Region ................................................... 10-18
Figure 10.13: 2016 Actual Wind Output BPA Balancing Authority ........................................... 10-19
Figure 10.14: Mid-Columbia Electric Price Forecast Range .................................................... 10-21
Figure 10.15: Western States Greenhouse Gas Emissions ..................................................... 10-22
Figure 10.16: Emission Intensity Metric.................................................................................... 10-23
Figure 10.17: Instate Emission Intensity Change from 2018 to 2037 ...................................... 10-24
Figure 10.18: Base Case Western Interconnect Resource Mix ............................................... 10-24
SC_PR_3-2 Attachment A Page 8 of 205
Table of Contents

Avista Corp 2017 Electric IRP

iv
Figure 10.19: Western Interconnect Resource Mix Changes .................................................. 10-25
Figure 10.20: Northwest Greenhouse Gas Emission Shadow Prices ...................................... 10-26
Figure 10.21: Washington Clean Air Rule Pricing .................................................................... 10-27
Figure 10.22: Annual Mid-Columbia Flat Price Forecast Colstrip Retires Scenario ................ 10-28
Figure 10.23: No Colstrip Scenario Annual Western U.S. Greenhouse Gas Emissions ......... 10-29
Figure 10.24: Colstrip Emissions & Pricing .............................................................................. 10-29
Figure 10.25: Greenhouse Gas Reduction ............................................................................... 10-31
Figure 10.26: Mid-Columbia Electric Price Comparison........................................................... 10-31
Figure 10.27: 2037 Generation Mix Comparison ..................................................................... 10-32
Figure 11.1: Resource Acquisition History ................................................................................. 11-2
Figure 11.2: Physical Resource Positions (Includes Energy Efficiency) .................................... 11-3
Figure 11.3: REC Requirements versus Qualifying RECs for EIA ............................................. 11-4
Figure 11.4: Conceptual Efficient Frontier Curve ....................................................................... 11-6
Figure 11.5: New Resources to Meet Winter Peak Loads ......................................................... 11-8
Figure 11.6: Load Forecast with and without Energy Efficiency .............................................. 11-10
Figure 11.7: Avista Owned and Controlled Resource’s Greenhouse Gas Emissions ............. 11-11
Figure 11.8: Projected Power Supply Expense Range ............................................................ 11-13
Figure 11.9: Expected Case Efficient Frontier .......................................................................... 11-14
Figure 11.10: Risk Adjusted PVRR of Efficient Frontier Portfolios ........................................... 11-15
Figure 11.11: Risk Adjusted PVRR of Efficient Frontier Portfolios ........................................... 11-16
Figure 12.1: Colstrip Retires Scenario Cost versus Risk ........................................................... 12-4
Figure 12.2: High-Cost Colstrip Retention Scenario Efficient Frontier ....................................... 12-7
Figure 12.3: High-Cost Colstrip Scenarios Annual Cost ............................................................ 12-8
Figure 12.4: Greenhouse Gas Emissions: Retire Colstrip in 2023 versus PRS ........................ 12-8
Figure 12.5: 50 Percent Colstrip Dispatch Reduction Scenario Cost & Risk Comparison ......... 12-9
Figure 12.6: Colstrip Dispatch Reduction Scenario Greenhouse Gas Comparison ................ 12-10
Figure 12.7: Other Resource Strategy Portfolio Cost and Risk (Millions) ................................ 12-11
Figure 12.8: Avista Direct Greenhouse Gas Emissions ........................................................... 12-15

SC_PR_3-2 Attachment A Page 9 of 205
Table of Contents

Avista Corp 2017 Electric IRP

v
Table of Tables

Table 1.1: The 2017 Preferred Resource Strategy ...................................................................... 1-4
Table 2.1: TAC Meeting Dates and Agenda Items ....................................................................... 2-2
Table 2.2: External Technical Advisory Committee Participating Organizations ......................... 2-3
Table 2.3: Idaho IRP Requirements ............................................................................................. 2-6
Table 2.4: Washington IRP Rules and Requirements .................................................................. 2-6
Table 3.1: UPC Models Using Non-Weather Driver Variables ..................................................... 3-8
Table 3.2: Customer Growth Correlations, January 2005 – December 2013 ............................ 3-11
Table 4.1: Avista-Owned Hydroelectric Resources ...................................................................... 4-4
Table 4.2: Avista-Owned Thermal Resources .............................................................................. 4-5
Table 4.3: Mid-Columbia Capacity and Energy Contracts ........................................................... 4-8
Table 4.4: PURPA Agreements .................................................................................................... 4-8
Table 4.5: Other Contractual Rights and Obligations ................................................................... 4-9
Table 5.1: Cumulative Potential Savings (Across All Sectors for Selected Years) ...................... 5-5
Table 5.2: Annual Achievable Potential Energy Efficiency (Megawatt Hours) ............................. 5-8
Table 5.3: Annual Achievable Potential Energy Efficiency (Megawatt Hours) ............................. 5-8
Table 5.4: Transmission and Distribution Benefit ....................................................................... 5-14
Table 5.5: Preliminary Generation Facility Efficiency Upgrade Potential ................................... 5-15
Table 5.6: Planned Generation Facility Efficiency Upgrades 2017 – 2018 ................................ 5-16
Table 5.7: Commercial and Industrial Demand Response Achievable Potential (MW) ............. 5-18
Table 6.1: Washington State EIA Compliance Position Prior to REC Banking (aMW) .............. 6-10
Table 7.1: Avista Owned and Controlled PM Emissions .............................................................. 7-7
Table 8.1: 2017 IRP Generation Study Transmission Costs ........................................................ 8-7
Table 8.2: Third-Party Large Generation Interconnection Requests ............................................ 8-7
Table 8.3: Capital Deferment Analysis ......................................................................................... 8-9
Table 8.4: Planned Feeder Rebuilds .......................................................................................... 8-10
Table 9.1: Natural Gas-Fired Plant Levelized Costs per MWh .................................................... 9-3
Table 9.2: Natural Gas-Fired Plant Cost and Operational Characteristics................................... 9-5
Table 9.3: Solar Capacity Credit by Month ................................................................................... 9-8
Table 9.4: Storage Power Supply Value .................................................................................... 9-18
Table 9.5: Natural Gas-Fired Facilities Ancillary Service Value ................................................. 9-19
Table 10.1: AURORAXMP Zones ................................................................................................. 10-2
Table 10.2: Added Northwest Renewable Generation Resources ............................................. 10-6
Table 10.3: Natural Gas Price Basin Differentials from Henry Hub ........................................... 10-8
Table 10.4: Monthly Price Differentials for Stanfield from Henry Hub ........................................ 10-8
Table 10.5: January through June Load Area Correlations ..................................................... 10-15
Table 10.6: July through December Load Area Correlations ................................................... 10-16
Table 10.7: Area Load Coefficient of Determination (Standard Deviation/Mean) .................... 10-16
Table 10.8: Area Load Coefficient of Determination (Standard Deviation/Mean) .................... 10-16
Table 10.9: Annual Average Mid-Columbia Electric Prices ($/MWh) ....................................... 10-21
Table 11.1: Qualifying Washington EIA Resources ................................................................... 11-4
Table 11.2: 2017 Preferred Resource Strategy .......................................................................... 11-7
Table 11.3: 2015 Preferred Resource Strategy .......................................................................... 11-9
Table 11.4: PRS Rate Base Additions from Capital Expenditures ........................................... 11-12
Table 11.5: Alternative Resource Strategies (2035) along the Efficient Frontier (MW) ........... 11-17
Table 11.6: 2017 IRP Avoided Costs ....................................................................................... 11-19
Table 12.1: Load Forecast Scenarios (2018-2037) .................................................................... 12-1
Table 12.2: Resource Selection for Load Forecast Scenarios ................................................... 12-2
Table 12.3: Colstrip Retires- Resource Strategy Options (ISO Conditions MW) ....................... 12-3
Table 12.4: Colstrip Retires in 2023 Scenario Resource Strategy ............................................. 12-7
Table 12.5: No New Thermal Resource Scenario .................................................................... 12-12
Table 12.6: No New Thermal Resource and Colstrip Replacement Scenario ......................... 12-13
Table 12.7: New CCCT Replaces Lancaster Scenario ............................................................ 12-14

SC_PR_3-2 Attachment A Page 10 of 205

SC_PR_3-2 Attachment A Page 11 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-1

1. Executive Summary

Avista’s 2017 Electric Integrated Resource Plan (IRP) shapes its resource strategy over
the next two years and procurements over the next 20 years. It provides a snapshot of
existing resources and loads and evaluates acquisition strategies over expected and
possible future conditions. The 2017 Preferred Resource Strategy (PRS) includes a mix
of solar, demand response, energy efficiency, storage, upgrades to existing assets, and
new natural gas-fired generation.

The PRS relies on modeling methods to balance cost, reliability, rate volatility, and
renewable requirements. Avista’s management and the Technical Advisory Committee
(TAC) guide IRP development through their input on modeling and planning
assumptions. TAC members include customers, Commission staff, the Northwest
Power and Conservation Council, consumer advocates, academics, environmental
groups, utility peers, government agencies, and other interested parties.

Resource Needs
Under extreme weather conditions, Avista expects its highest peak loads in the winter.
Its peak planning methodology includes operating reserves, regulation, load following,
wind integration, a 14 percent planning margin over winter-peak load levels, and a
seven percent planning margin over summer-peak load levels. The company has
adequate resources combined with conservation to meet peak load requirements
through October 2026. Figure 1.1 shows Avista’s resource position through 2037.
Chapter 6 – Long-Term Position details Avista’s resource needs.

Figure 1.1: Load-Resource Balance—Winter Peak Load & Resource Availability

0
500
1,000
1,500
2,000
2,500
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
g
a
w
a
t
t
s
Existing Resources & Rights
Load w/o Conservation + Cont.
Load w/ Conservation + Cont.
SC_PR_3-2 Attachment A Page 12 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-2

Modeling and Results
Avista uses multiple steps to develop its PRS; beginning with identifying and quantifying
potential new generation resources to serve projected electricity demand across the
Western Interconnect. This study determines the impact of external markets on the
Northwest electricity marketplace. It then maps existing Avista resources to the
transmission grid in a model simulating hourly operations for the Western Interconnect
in the 2018 to 2037 IRP timeframe. The model adds new resources and transmission to
the Western Interconnect as regional loads grow and resources retire. Monte Carlo-
style analyses vary hydroelectric and wind generation, loads, forced outages and
natural gas price data over 500 iterations of potential future market conditions to
develop the Mid-Columbia electricity marketplace through 2037.

Electricity and Natural Gas Market Forecasts
Figure 1.2 shows the 2017 IRP Mid-Columbia electricity price forecast for the Expected
Case, including the range of prices resulting from 500 Monte Carlo iterations. The
levelized price is $35.85 per MWh in nominal dollars over the 2018-2037 timeframe.

Figure 1.2: Average Mid-Columbia Electricity Price Forecast

Electricity and natural gas prices are highly correlated because natural gas fuels
marginal generation in the Northwest during most of the year. Figure 1.3 presents
nominal Expected Case natural gas prices at the Stanfield trading hub, located in
northeastern Oregon, as well as the forecast range from the 500 Monte Carlo iterations
performed for the Expected Case. The average is $4.20 per dekatherm (Dth) over the
next 20 years. See Chapter 10 – Market Analysis for natural gas and electricity price
forecasts.

$/MWh
$20/MWh
$40/MWh
$60/MWh
$80/MWh
$100/MWh
$120/MWh
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
20
1
8
-
3
7
SC_PR_3-2 Attachment A Page 13 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-3

Figure 1.3: Stanfield Natural Gas Price Forecast
Energy Efficiency Acquisition
Avista commissioned a 20-year Conservation Potential Assessment (CPA) to determine
potential residential, commercial and industrial energy efficiency applications. Data from
this study formed the basis of the IRP’s conservation analysis. This study estimates
conservation will serve 53.3 percent of future load growth. Since 1978, Avista’s load is
12.3 percent lower due to conservation. Figure 1.4 illustrates the historical efficiency
acquisitions as blue bars and the dashed line shows the amount of energy efficiency still
reducing loads due to the 18-year assumed measure life. See Chapter 5 – Energy
Efficiency and Demand Response for details.

Figure 1.4: Annual and Cumulative Energy Efficiency Acquisitions

$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
$12/Dth
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
20
1
8
-
3
7
aMW
20 aMW
40 aMW
60 aMW
80 aMW
100 aMW
120 aMW
140 aMW
160 aMW
180 aMW
200 aMW
220 aMW
aMW
2 aMW
4 aMW
6 aMW
8 aMW
10 aMW
12 aMW
14 aMW
16 aMW
18 aMW
20 aMW
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
20
0
4
20
0
6
20
0
8
20
1
0
20
1
2
20
1
4
20
1
6
Cu
m
u
l
a
t
i
v
e

S
a
v
i
n
g
s
An
n
u
a
l

S
a
v
i
n
g
s
SC_PR_3-2 Attachment A Page 14 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-4

Preferred Resource Strategy
The PRS results from careful consideration and input by Avista’s management, the
TAC, and from the information gathered and analyzed in the IRP process. It meets
future load growth with upgrades at existing generation facilities, energy efficiency,
natural gas-fired technologies, storage, energy efficiency, and demand response, as
shown in Table 1.1.
Table 1.1: The 2017 Preferred Resource Strategy

Resource By the End of
Year
ISO Conditions
(MW)
Winter Peak
(MW)
Energy
(aMW)
Solar 2018 15 0 3
Natural Gas Peaker 2026 192 204 178
Thermal Upgrades 2026-2029 34 34 31
Storage 2029 5 5 0
Natural Gas Peaker 2030 96 102 89
Natural Gas Peaker 2034 47 47 43
Total 389 392 344
Efficiency
Improvements
Acquisition
Range
Winter Peak
Reduction
(MW)
Energy
(aMW)
Energy Efficiency 2018-2037 203 108
Demand Response 2025-2037 44 0
Distribution Efficiencies <1 <1
Total 247 108

The 2017 PRS describes a reasonable low-cost plan along the Efficient Frontier of
potential resource portfolios accounting for fuel supply and price risks. Major changes
from the 2015 IRP include a lower contribution from natural gas-fired peakers and
inclusion of demand response, solar and storage resources.

Each new generation resource and energy efficiency option is valued against the
Expected Case’s Mid-Columbia electricity market forecast to identify its future energy
value, as well as its inherent risk measured by year-to-year portfolio power cost
volatility. These values, and their associated capital and fixed operation and
maintenance (O&M) costs, form the input into Avista’s Preferred Resource Strategy
Linear Programming Model (PRiSM). PRiSM assists Avista by developing optimal mixes
of new resources along an efficient frontier. Chapter 11 – Preferred Resource Strategy
provides a detailed discussion of the efficient frontier concept.

The PRS provides a least reasonable-cost portfolio, minimizing future costs and risks
within actual and expected environmental constraints. The Efficient Frontier helps
determine the tradeoffs between risk and cost. The approach is similar to finding an
optimal mix of risk and return in an investment portfolio, as potential returns increase, so
do risks. Conversely, reducing risk generally reduces overall returns. Figure 1.5
presents the change in cost and risk from the PRS on the Efficient Frontier. Lower
power cost variability comes from investments in more expensive, but less risky,
SC_PR_3-2 Attachment A Page 15 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-5

resources such as wind and hydroelectric upgrades. The PRS is the portfolio selected
on the Efficient Frontier where reduced risk justifies the increased cost.

Figure 1.5: Efficient Frontier

Chapter 12 – Portfolio Scenarios, includes several scenarios identifying tipping points
where the PRS could change under different conditions from the Expected Case. It also
evaluates the impacts of, among others, varying load growth, resource capital costs,
and greenhouse gas policies.

Energy Independence Act Compliance
Washington’s Energy Independence Act (EIA), or Initiative 937, requires utilities with
over 25,000 customers to meet nine percent of retail load from qualified renewable
resources by 2016 and 15 percent by 2020. The initiative also requires utilities to
acquire all cost-effective conservation and energy efficiency measures. Avista will meet
or exceed its EIA requirements through the IRP timeframe with a combination of
qualifying hydroelectric upgrades, the Palouse Wind project, and Kettle Falls Generating
Station output. Figure 1.6 shows Avista’s EIA-qualified generation; Chapter 6 – Long-
Term Position covers this topic in-depth.

$20 Mil
$30 Mil
$40 Mil
$50 Mil
$60 Mil
$70 Mil
$80 Mil
$90 Mil
$350 Mil $400 Mil $450 Mil $500 Mil $550 Mil
20
3
0

S
t
d
e
v

o
f

P
o
w
e
r

S
u
p
p
l
y

C
o
s
t
s
Levelized Cost 2018-2042
Least Cost Preferred Resource Strategy
Least Risk
SC_PR_3-2 Attachment A Page 16 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-6

Figure 1.6: Avista’s Qualifying Renewables for Washington State’s EIA

Greenhouse Gas Emissions
The regulation of greenhouse gases, or carbon emissions, has changed since the 2015
IRP with the change in presidential administrations, resulting in evolving federal and
additional state-driven regulation. Some states have active cap and trade programs,
emissions performance standards, renewable portfolio standards, or a combination of
current and proposed regulations affecting emissions from electric generation
resources.

Figure 1.7 shows that Avista emissions will decrease over the IRP timeframe. The 2017
IRP’s emissions forecast is 29 percent lower for 2035 than the 2015 IRP’s forecast.
Figure 1.8 shows the western-region emissions likely will fall from historic levels.
Regional emissions fall below 1990 levels by the end of the study period due to coal
retirements and potential state and federal policies. More details on state and federal
greenhouse gas policies are in Chapter 7 – Policy Consideration. Results of
greenhouse-gas policy scenarios are in Chapter 10 – Market Analysis and Chapter 12 –
Portfolio Scenarios.

-
20
40
60
80
100
120
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Banking
Kettle Falls
Palouse Wind
Hydro Upgrades
Requirement
SC_PR_3-2 Attachment A Page 17 of 205
Chapter 1- Executive Summary
Avista Corp 2017 Electric IRP 1-7

Figure 1.7: Avista Owned and Controlled Resource’s Greenhouse Gas Emissions

Figure 1.8: U.S. Western Interconnect Greenhouse Gas Emissions

Action Items
The 2017 Action Items chapter updates progress made on Action Items in the 2015 IRP
and outlines activities Avista intends to perform between the publication of this report
and publication of the 2019 IRP. It includes input from Commission Staff, Avista’s
management team, and the TAC. Action Item categories include generation resource-
related analysis, energy efficiency, and transmission planning. Refer to Chapter 13 –
Action Items for details about each of these categories.

(0.10)
0.02
0.14
0.26
0.38
0.50
Mil
1 Mil
2 Mil
3 Mil
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
t
r
i
c

T
o
n
s

p
e
r

M
W
h
Me
t
r
i
c

T
o
n
s
Expected Case: PRS
Metric Tons per MWh
-
50
100
150
200
250
300
350
19
9
0
19
9
2
19
9
4
19
9
6
19
9
8
20
0
0
20
0
2
20
0
4
20
0
6
20
0
8
20
1
0
20
1
2
20
1
4
20
1
6
20
1
8
20
2
0
20
2
2
20
2
4
20
2
6
20
2
8
20
3
0
20
3
2
20
3
4
20
3
6
Mi
l
l
i
o
n

M
e
t
i
c

T
o
n
s

o
f

C
O
2
Historical
Expected Case
10th Percentile
90th Percentile
SC_PR_3-2 Attachment A Page 18 of 205

SC_PR_3-2 Attachment A Page 19 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
2. Introduction and Stakeholder Involvement

Avista submits an IRP to the Idaho and Washington public utility commissions
biennially.1 Including its first plan in 1989, the 2017 IRP is Avista’s fifteenth plan. It
identifies and describes a PRS for meeting load growth while balancing cost and risk
measures with environmental mandates.

Avista is statutorily obligated to provide safe and reliable electricity service to its
customers at rates, terms, and conditions that are fair, just, reasonable, and sufficient.
Avista assesses different resource acquisition strategies and business plans to acquire
a mix of resources meeting resource adequacy requirements and optimizing the value
of its current portfolio. The IRP is a resource evaluation tool, not a plan for acquiring a
particular set of assets. Actual resource acquisition generally occurs through
competitive bidding processes.

IRP Process
The 2017 IRP is developed and written with the aid of a public process. Avista actively
seeks input from a variety of constituents through the TAC. The TAC is a mix of over
100 invited external participants, including staff from the Idaho and Washington
commissions, customers, academics, environmental organizations, government
agencies, consultants, utilities, and other interested parties, who joined the planning
process.

Avista sponsored six TAC meetings for the 2017 IRP. The first meeting was on June 2,
2016 and the last occurred on June 20, 2017. Each TAC meeting covers different
aspects of IRP planning activities. At the meetings, members provide contributions to,
and assessments of, modeling assumptions, modeling processes, and results of Avista
studies. Table 2.1 contains a list of TAC meeting dates and the agenda items covered in
each meeting.

Agendas and presentations from the TAC meetings are in Appendix A and on Avista’s
website at https://www.myavista.com/about-us/our-company/integrated-resource-
planning. The website link contains all past IRPs and TAC meeting presentations back
to 1989.

1 Washington IRP requirements are contained in WAC 480-100-238 Integrated Resource Planning. Idaho
IRP requirements are in Case No. U-1500-165, Order No. 22299 and Case No. GNR-E-93-3, Order No.
25260.
SC_PR_3-2 Attachment A Page 20 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
Table 2.1: TAC Meeting Dates and Agenda Items

Meeting Date Agenda Items
TAC 1 – June 2, 2016  TAC Meeting Expectations
 2015 IRP Commission Acknowledgements
 2015 Action Plan Update
 Energy Independence Act Compliance
 Energy Efficiency Modeling Discussion
 Resource Adequacy – Preliminary Results
 Draft 2017 Electric IRP Work Plan
TAC 2 – September 28, 2016  Introduction & TAC 1 Recap
 TAC 1 Action Item Update
 Electrification Update
 Load and Economic Forecasts
 Supply Side Options
 Clean Energy Fund 2 Grant Project
TAC 3 – November 8, 2016  Introduction & TAC 2 Recap
 Colstrip Discussion
 Clean Power Plan and Clean Air Rule
 IRP Modeling Overview
 Cost of Carbon
 Avista’s Power Planner Simulator
TAC 4 – February 15, 2017  Introduction & TAC 3 Recap
 Resource Needs Assessment
 Natural Gas Price Forecast
 Electric Price Forecast
 Transmission Planning
 Market and Portfolio Scenarios
TAC 5 – March 28, 2017  Introduction & TAC 4 Recap
 Updated Electric Price Forecast
 Energy Storage and Ancillary Services
 Conservation Potential Assessment
 Distribution Planning
 Draft Preferred Resource Strategy
TAC 6 – June 20, 2017  Introduction & TAC 5 Recap
 Conservation Assessment
 Final 2017 Preferred Resource Strategy
 Scenario Analysis
 C&I Solar Select Program
 2019 IRP Action Items
 2017 IRP Document Overview

SC_PR_3-2 Attachment A Page 21 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
Avista greatly appreciates the valuable contributions of its TAC members and wishes to
acknowledge and thank the organizations that allow their attendance. Table 2.2 is a list
of the organizations participating in the 2017 IRP TAC process.

Table 2.2: External Technical Advisory Committee Participating Organizations

Organization
AEG
City of Spokane
Clearwater Paper
Eastern Washington University
GE Energy
Idaho Conservation League
Idaho Department of Environmental Quality
Idaho Power
Idaho Public Utilities Commission
Inland Empire Paper
NW Energy Coalition
Northwest Power and Conservation Council
PacifiCorp
Pend Oreille PUD
Puget Sound Energy
Renewable Northwest
Residential and Small Commercial Customers
Sierra Club
Snake River Alliance
Spokane Neighborhood Action Partners
The Energy Authority
Washington State Office of the Attorney General
Washington Department of Enterprise Services
Washington Utilities and Transportation Commission
Whitman County Commission

Issue Specific Public Involvement Activities
In addition to TAC meetings, Avista sponsors and participates in several other
collaborative processes involving a range of public interests. A sampling is below.

Energy Efficiency Advisory Group
The energy efficiency Advisory Group provides stakeholders and public groups biannual
opportunities to discuss Avista’s energy efficiency efforts.

FERC Hydro Relicensing – Clark Fork and Spokane River Projects
Over 50 stakeholder groups participated in the Clark Fork hydro-relicensing process
beginning in 1993. This led to the first all-party settlement filed with a FERC relicensing
application, and the eventual issuance of a 45-year FERC operating license in February
2003. This collaborative process continues in the implementation of the license and
Clark Fork Settlement Agreement, with stakeholders participating in various protection,
mitigation, and enhancement efforts. Avista received a 50-year license for the Spokane
SC_PR_3-2 Attachment A Page 22 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
River Project following a multi-year collaborative process involving several hundred
stakeholders. Implementation began in 2009 with a variety of collaborating parties.

Low Income Rate Assistance Program
This program is coordinated with four community action agencies in Avista’s
Washington service territory. The program began in 2001, and quarterly reviews ensure
changing administrative issues and needs are met.

Regional Planning
The Pacific Northwest generation and transmission system operates in a coordinated
fashion. Avista participates in the efforts of many regional planning processes.
Information from this participation supplements Avista’s IRP process. A partial list of the
regional organizations Avista participates in includes:

 Western Electricity Coordinating Council
 Peak Reliability
 Northwest Power and Conservation Council
 Northwest Power Pool
 Pacific Northwest Utilities Conference Committee
 ColumbiaGrid
 Northern Tier Transmission Group
 North American Electric Reliability Corporation

Future Public Involvement
Avista actively solicits input from interested parties to enhance its IRP process. We
continue to expand TAC membership and diversity, and maintain the TAC meetings as
an open public process.

2017 IRP Outline
The 2017 IRP consists of 13 chapters plus an executive summary and this introduction.
A series of technical appendices supplement this report.

Chapter 1: Executive Summary
This chapter summarizes the overall results and highlights of the 2017 IRP.

Chapter 2: Introduction and Stakeholder Involvement
This chapter introduces the IRP and details public participation and involvement in the
IRP planning process.

Chapter 3: Economic and Load Forecast
This chapter covers regional economic conditions, Avista’s energy and peak load
forecasts, and load forecast scenarios.

Chapter 4: Existing Supply Resources
This chapter provides an overview of Avista-owned generating resources and its
contractual resources and obligations.

SC_PR_3-2 Attachment A Page 23 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
Chapter 5: Energy Efficiency and Demand Response
This chapter discusses Avista energy efficiency programs. It provides an overview of
the conservation potential assessment and summarizes energy efficiency and demand
response modeling results.

Chapter 6: Long-Term Position
This chapter reviews Avista reliability planning and reserve margins, resource
requirements, and provides an assessment of its reserves and flexibility.

Chapter 7: Policy Considerations
This chapter focuses on some of the major policy issues for resource planning,
including state and federal greenhouse gas policies and environmental regulations.

Chapter 8: Transmission & Distribution Planning
This chapter discusses Avista distribution and transmission systems, as well as regional
transmission planning issues. It includes detail on transmission cost studies used in IRP
modeling and provides a summary of our 10-year Transmission Plan. The chapter
concludes with a discussion of distribution efficiency and grid modernization projects;
including storage benefits to the distribution system.

Chapter 9: Generation Resource Options
This chapter covers the costs and operating characteristics of the generation resource
options modeled for the IRP.

Chapter 10: Market Analysis
This chapter details Avista IRP modeling and its analyses of the wholesale market.

Chapter 11: Preferred Resource Strategy
This chapter details the resource selection process used to develop the 2017 PRS,
including the efficient frontier and resulting avoided costs.

Chapter 12: Portfolio Scenarios
This chapter discusses the portfolio scenarios and tipping point analyses.

Chapter 13: Action Items
This chapter discusses progress made on Action Items contained in the 2015 IRP. It
details the action items Avista will focus on between publication of this plan and the
2019 IRP.

SC_PR_3-2 Attachment A Page 24 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
Regulatory Requirements
The IRP process for Idaho has several requirements documented in IPUC Orders Nos.
22299 and 25260. Table 2.3 summarizes them.

Table 2.3: Idaho IRP Requirements

Requirement Plan Citation
Identify and list relevant operating characteristics
of existing resources by categories including:
hydroelectric, coal-fired, oil or gas-fired, PURPA
(by type), exchanges, contracts, transmission
resources, and others.
Chapter 4- Existing Supply Resources
Identify and discuss the 20-year load forecast
plus scenarios for the different customer classes.
Identify the assumptions and models used to
develop the load forecast.
Chapter 3- Economic & Load Forecast
Chapter 12- Portfolio Scenarios
Identify the utility’s plan to meet load over the 20-
year planning horizon. Include costs and risks of
the plan under a range of plausible scenarios.
Chapter 11- Preferred Resource
Strategy
Chapter 12- Portfolio Scenarios
Identify energy efficiency resources and costs. Chapter 5- Energy Efficiency & Demand
Response
Provide opportunities for public participation and
involvement.
Chapter 2- Introduction and Stakeholder
Involvement
Explain the present load/resource position,
expected responses to possible future events,
and the role of conservation in those responses.
Chapter 6- Long-Term Position
Chapter 12- Portfolio Scenarios
Chapter 5- Energy Efficiency & Demand
Response
Discuss any flexibilities and analyses considered,
such as: (1) examination of load forecast
uncertainties; (2) effects of known or potential
changes to existing resources; (3) consideration
of demand- and supply-side resource options,
and (4) contingencies for upgrading, optioning
and acquiring resources.

Chapter 3- Economic & Load Forecast
Chapter 4- Existing Supply Resources
Chapter 9- Generation Resource
Options
Chapter 11- Preferred Resource
Strategies

The IRP process for Washington has several requirements documented in Washington
Administrative Code (WAC). Table 2.4 summarizes where in the document Avista
addressed each requirement.

Table 2.4: Washington IRP Rules and Requirements

Rule and Requirement Plan Citation
–

–

SC_PR_3-2 Attachment A Page 25 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
WAC 480-100-238(2)(a) – Plan describes mix of
energy supply resources. Chapter 4- Existing Supply Resources
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(a) – Plan describes
conservation supply.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(a) – Plan addresses
supply in terms of current and future needs of
utility ratepayers.
Chapter 3- Economic & Load Forecast
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – Plan uses lowest
reasonable cost (LRC) analysis to select mix of
resources.
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – LRC analysis
considers resource costs.
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – LRC analysis
considers market-volatility risks.
Chapter 10- Market Analysis
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – LRC analysis
considers demand side uncertainties.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 12- Portfolio Scenarios
WAC 480-100-238(2)(b) – LRC analysis
considers resource dispatchability. Chapter 9- Generation Resource Options
Chapter 10- Market Analysis
WAC 480-100-238(2)(b) – LRC analysis
considers resource effect on system operation. Chapter 10- Market Analysis
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – LRC analysis
considers risks imposed on ratepayers.
Chapter 7- Policy Considerations
Chapter 9- Generation Resource Options
Chapter 10- Market Analysis
Chapter 11- Preferred Resource Strategy
Chapter 12- Portfolio Scenarios
WAC 480-100-238(2)(b) – LRC analysis
considers public policies regarding resource
preference adopted by Washington state or
federal government.
Chapter 3- Economic & Load Forecast
Chapter 4- Existing Supply Resources
Chapter 7- Policy Considerations
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(2)(b) – LRC analysis
considers cost of risks associated with
environmental effects including emissions of
carbon dioxide.
Chapter 7- Policy Considerations
Chapter 11- Preferred Resource Strategy
Chapter 12- Portfolio Scenarios
WAC 480-100-238(2)(c) – Plan defines
conservation as any reduction in electric power
consumption that results from increases in the
efficiency of energy use, production, or
distribution.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 11- Preferred Resource Strategy
WAC 480-100-238(3)(a) – Plan includes a range
of forecasts of future demand.
Chapter 3- Economic & Load Forecast
Chapter 12- Portfolio Scenarios
WAC 480-100-238(3)(a) – Plan develops
forecasts using methods that examine the effect
of economic forces on the consumption of
electricity.
Chapter 3- Economic & Load Forecast
Chapter 12- Portfolio Scenarios
WAC 480-100-238(3)(a) – Plan develops
forecasts using methods that address changes
in the number, type and efficiency of end-uses.
Chapter 3- Economic & Load Forecast
Chapter 5- Energy Efficiency & Demand
Response
Chapter 8- Transmission & Distribution

SC_PR_3-2 Attachment A Page 26 of 205
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2017 Electric IRP
WAC 480-100-238(3)(b) – Plan includes an
assessment of commercially available
conservation, including load management.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 8- Transmission & Distribution
WAC 480-100-238(3)(b) – Plan includes an
assessment of currently employed and new
policies and programs needed to obtain the
conservation improvements.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 8- Transmission & Distribution

WAC 480-100-238(3)(c) – Plan includes an
assessment of a wide range of conventional and
commercially available nonconventional
generating technologies.
Chapter 9- Generation Resource Options
Chapter 11- Preferred Resource Strategy
Chapter 12- Portfolio Scenarios
WAC 480-100-238(3)(d) – Plan includes an
assessment of transmission system capability
and reliability (as allowed by current law).
Chapter 8- Transmission & Distribution

WAC 480-100-238(3)(e) – Plan includes a
comparative evaluation of energy supply
resources (including transmission and
distribution) and improvements in conservation
using LRC.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 8- Transmission & Distribution
Chapter 11- Preferred Resource Strategy
WAC-480-100-238(3)(f) – Demand forecasts
and resource evaluations are integrated into the
long range plan for resource acquisition.
Chapter 5- Energy Efficiency & Demand
Response
Chapter 8- Transmission & Distribution
Chapter 9- Generation Resource Options
Chapter 12- Portfolio Scenarios
WAC 480-100-238(3)(g) – Includes a two-year
action plan implementing the long range plan.
Chapter 13- Action Items
WAC 480-100-238(3)(h) – Plan includes a
progress report on the implementation of the
previously filed plan.
Chapter 13- Action Items
WAC 480-100-238(5) – Plan includes
description of consultation with commission staff
and public participation
Chapter 2- Introduction and Stakeholder
Involvement
WAC 480-100-238(5) – Plan includes
description of work plan.
Appendix B
WAC 480-107-015(3) – Proposed request for
proposals for new capacity needed within three
years of the IRP.
Chapter 10- Preferred Resource Strategy
RCW 19.280.030-1(e) – An assessment of
methods, commercially available technologies,
or facilities for integrating renewable resources,
and addressing overgeneration events, if
applicable to the utility's resource portfolio;
Chapter 9- Generation Resource Options
Chapter 10- Market Analysis

RCW 19.280.030-1(f) – Integration of demand
forecasts and resource evaluations into a long-
range assessment describing the mix of supply
side generating resources and conservation and
efficiency resources that will meet current and
projected needs, including mitigating
overgeneration events, at the lowest reasonable
cost and risk to the utility and its ratepayers.
Chapter 9- Generation Resource Options
Chapter 10- Market Analysis

SC_PR_3-2 Attachment A Page 27 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-1
3. Economic & Load Forecast

Introduction & Highlights
An explanation and quantification of Avista’s loads and resources are integral to the IRP.
This chapter summarizes Expected Case customer and load projections, load growth
scenarios, and recent enhancements to our forecasting models and processes.

Economic Characteristics of Avista’s Service Territory
Avista’s core service area for electricity includes a population of more than a half million
people residing in Eastern Washington and Northern Idaho. Three metropolitan statistical
areas (MSAs) dominate its service area: the Spokane-Spokane Valley, WA MSA
(Spokane-Stevens counties); the Coeur d’Alene, ID MSA (Kootenai County); and the
Lewiston-Clarkson ID-WA, MSA (Nez Perce-Asotin counties). These three MSAs account
for just over 70 percent of both customers (i.e., meters) and load. The remaining 30
percent are in low-density rural areas in both states. Washington accounts for about two-
thirds of customers and Idaho the remaining one-third.

Population
Population growth is increasingly a function of net migration within Avista’s service area.
Net migration is strongly associated with both service area and national employment
growth through the business cycle. The regional business cycle follows the U.S. business
cycle, meaning regional economic expansions or contractions follow national trends.1
Econometric analysis shows that when regional employment growth is stronger than U.S.
growth over the business cycle, it is associated with increased in-migration. The reverse
holds true. Figure 3.1 shows annual population growth since 1971 and highlights the
recessions. During all deep economic downturns since the mid-1970s, reduced
population growth rates in Avista’s service territory led to lower load growth.2 The Great
Recession reduced population growth from nearly two percent in 2007 to less than one
percent from 2010 to 2013. Accelerating service area employment growth in 2013 helped
push population growth to around one percent starting in 2014.

1 An Exploration of Similarities between National and Regional Economic Activity in the Inland Northwest,
Monograph No. 11, May 2006. http://www.ewu.edu/cbpa/centers-and-institutes/ippea/monograph-
series.xml.
2 Data Source: Bureau of Economic Development, U.S. Census, and National Bureau of Economic
Research.
Chapter Highlights
 Population and employment growth are recovering from the Great Recession.
 The 2017 Expected Case energy forecast grows 0.47 percent per year,
replacing the 0.6 percent annual growth rate in the 2015 IRP.
 Peak load growth is lower than energy growth, at 0.42 percent in the winter and
0.46 percent in the summer.
 Retail sales and residential use per customer forecasts continue to decline from
2015 IRP projections.
SC_PR_3-2 Attachment A Page 28 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-2
Figure 3.1: MSA Population Growth and U.S. Recessions, 1971-2016

Figure 3.2 shows population growth since the start of the Great Recession in 2007.3
Service area population growth over the 2010-2012 period was weaker than the U.S.; it
was closely associated with the strength of regional employment growth relative to the
U.S. over the same period. The same can be said for the increase in service area
population growth in 2014 relative to the U.S. The association of employment growth to
population growth has a one year lag. The relative strength of service area population
growth in year “y” is positively associated with service area population growth in year
“y+1”. Econometric estimates based on historical data show that, holding U.S.
employment-growth constant, every one percent increase in service area employment
growth is associated with a 0.4 percent increase in population growth in the next year.
Employment
It is useful to examine the distribution of employment and employment performance since
2007 given the correlation between population and employment growth. The Inland
Northwest has transitioned from a natural resources-based manufacturing economy to a
services-based economy. Figure 3.3 shows the breakdown of non-farm employment for
all three service area MSAs.4 Approximately 70 percent of employment in the three MSAs
is in private services, followed by government (17 percent) and private goods-producing
sectors (14 percent). Farming accounts for one percent of total employment.

Spokane and Coeur d’Alene MSAs are major providers of health and higher education
services to the Inland Northwest. A recent addition to these sectors is approval from
Washington’s legislature for Washington State University to open a medical school in
Spokane, Washington.

3 Data Source: Bureau of Economic Analysis, U.S. Census, and Washington State OFM.
4 Data Source: Bureau of Labor and Statistics.
-1.0%
-0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
19
7
1
19
7
3
19
7
5
19
7
7
19
7
9
19
8
1
19
8
3
19
8
5
19
8
7
19
8
9
19
9
1
19
9
3
19
9
5
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
An
n
u
a
l

G
r
o
w
t
h
SC_PR_3-2 Attachment A Page 29 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-3
Figure 3.2: Avista and U.S. MSA Population Growth, 2007-2016

Non-farm employment growth averaged 2.7 percent per year between 1990 and 2007.
However, Figure 3.4 shows that service area employment lagged the U.S. recovery from
the Great Recession for the 2010-2012 period.5 Regional employment recovery did not
materialize until 2013, when services employment started to grow. Prior to this, reductions
in federal, state, and local government employment offset gains in goods producing
sectors. Service area employment growth began to match or exceed U.S. growth rates
by the fourth quarter 2014.

Figure 3.5 shows the distribution of personal income, a broad measure of both earned
income and transfer payments, for Avista’s Washington and Idaho MSAs.6 Regular
income includes net earnings from employment, and investment income in the form of
dividends, interest and rent. Personal current transfer payments include money income
and in-kind transfers received through unemployment benefits, low-income food
assistance, Social Security, Medicare, and Medicaid.

5 Data Source: Bureau of Labor and Statistics.
6 Data Source: Bureau of Economic Analysis.
1.9%
1.4%
1.2%
0.8%
0.5%0.5%
0.8%
1.1%
0.9%
1.0%1.0%1.0%
0.9%
0.8%0.8%0.8%
0.7%
0.8%0.8%0.8%
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
1.4%
1.6%
1.8%
2.0%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
U.S.
SC_PR_3-2 Attachment A Page 30 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-4
Figure 3.3: MSA Non-Farm Employment Breakdown by Major Sector, 2016

Figure 3.4: Avista and U.S. MSA Non-Farm Employment Growth, 2007-2016

Private Goods
Producing, 14%
Private Service
Producing, 69%
Federal Government,
2%
State Government,
4%
Local Government,
11%
2.2%
0.5%
-4.7%
-1.6%
0.2%
0.6%
2.0%1.8%2.0%
3.0%
1.1%
-0.5%
-4.3%
-0.7%
1.2%
1.7%1.6%1.9%2.1%
1.8%
-5.5%
-4.5%
-3.5%
-2.5%
-1.5%
-0.5%
0.5%
1.5%
2.5%
3.5%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
U.S.
SC_PR_3-2 Attachment A Page 31 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-5
Figure 3.5: MSA Personal Income Breakdown by Major Source, 2015

Transfer payments in Avista’s service area in 1970 accounted for 12 percent of the local
economy. The income share of transfer payments has nearly doubled over the last 40
years to 23 percent. The relatively high regional dependence on government employment
and transfer payments means continued federal fiscal consolidation and transfer program
reform may reduce future growth. Although 57 percent of personal income is from net
earnings, transfer payments account for more than one in every five dollars of personal
income. Recent years have seen transfer payments become the fastest growing
component of regional personal income. This growth reflects an aging regional
population, a surge of military veterans, and the Great Recession; the later significantly
increased payments from unemployment insurance and other low-income assistance
programs.

Figure 3.6 shows the real (inflation adjusted) average annual growth per capita income
by MSA for Avista’s service area and the U.S. overall. Note that in the 1980 – 1990 period
the service area experienced significantly lower income growth compared to the U.S. as
a result of the back-to-back recessions of the early 1980s.7 The impacts of these
recessions were more negative in the service area compared to the U.S. as a whole, so
the ratio of service area per capita income to U.S. per capita income fell from 93 percent
in the previous decade to around 85 percent. The income ratio has not since recovered.

7 Data Source: Bureau of Economic Analysis.
Net Earnings,
57%
Dividends,
Interest, and
Rent, 20%
Transfer
Receipts, 23%
SC_PR_3-2 Attachment A Page 32 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-6
Figure 3.6: Avista and U.S. MSA Real Personal Income Growth, 1970-2013

Five-Year Load Forecast Methodology
In non-IRP years, the retail and native load forecasts have a five-year time horizon. Avista
conducts the forecasts each spring with the option of second forecast in the winter if
changing economic conditions warrant a new forecast. The results are fed into Avista’s
revenue model, which converts the load forecast into a revenue forecast. In turn, the
revenue forecast feeds Avista’s earnings model. In IRP years, the long-term forecast
boot-straps off the five-year forecast by applying growth assumptions beyond year five.

Overview of the Five-Year Retail Load Forecast
The five-year retail load forecast is a two-step process. For most schedules in each class,
there is a monthly use per customer (UPC) forecast and a monthly customer forecast.8
The load forecast is generated by multiplying the customer and UPC forecasts. The UPC
and customer forecasts are generated using time-series econometrics, as shown in
Equation 3.1.

Equation 3.1: Generating Schedule Total Load
𝐹(𝑘𝑊ℎ𝑡,𝑦𝑐+𝑗,𝑠) = 𝐹(𝑘𝑊ℎ/𝐶𝑡,𝑦𝑐+𝑗,𝑠) × 𝐹(𝐶𝑡,𝑦𝑐+𝑗,𝑠)

Where:
 F(kWht,yc+j,s) = the forecast for month t, year j = 1,…,5 beyond the
current year, yc ,for schedule s.
 F(kWh/Ct,yc+j,s) = the UPC forecast.
 F(Ct,yc+j,s) = the customer forecast.

8 For schedules representing a single customer, where there is no customer count and for street lighting,
total load is forecast directly without first forecasting UPC.
2.3%
1.4%
2.3%
0.7%
1.6%
2.1%
2.3%2.4%
0.7%
2.1%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
1970 to 1980 1980 to 1990 1990 to 2000 2000 to 2010 2010 to 2015
Re
a
l

A
v
e
r
a
g
e

A
n
n
u
a
l

G
r
o
w
t
h

R
a
t
e
Avista WA-ID MSAs
U.S.
SC_PR_3-2 Attachment A Page 33 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-7
UPC Forecast Methodology
The econometric modeling for UPC is a variation of the “fully integrated” approach
expressed by Faruqui (2000) in the following equation:9

Equation 3.2: Use Per Customer Regression Equation

𝑘𝑊ℎ/𝐶𝑡,𝑦,𝑠= 𝛼𝑊𝑡,𝑦+ 𝛽𝑍𝑡,𝑦+ 𝜖𝑡,𝑦

The model uses actual historical weather, UPC, and non-weather drivers to estimate the
regression in Equation 3.2. To develop the forecast, normal weather replaces actual
weather (W) along with the forecasted values for the Z variables (Faruqui, pp. 6-7). Here,
W is a vector of heating degree day (HDD) and cooling degree day (CDD) variables; Z is
a vector of non-weather variables; and εt,y is an uncorrelated N(0,σ) error term. For non-
weather sensitive schedules, W = 0.

The W variables will be HDDs and CDDs. Depending on the schedule, the Z variables
may include real average energy price (RAP); the U.S. Federal Reserve industrial
production index (IP); non-weather seasonal dummy variables (SD); trend functions (T);
and dummy variables for outliers (OL) and periods of structural change (SC). RAP is
measured as the average annual price (schedule total revenue divided by schedule total
usage) divided by the consumer price index (CPI), less energy. For most schedules, the
only non-weather variables are SD, SC, and OL. See Table 3.1 for the occurrence RAP
and IP.

If the error term appears to be non-white noise, then the forecasting performance of
Equation 3.3 can be improved by converting it into an ARIMA “transfer function” model
such that Єt,y = ARIMAЄt,y(p,d,q)(pk,dk,qk)k. The term p is the autoregressive (AR) order,
d is the differencing order, and q is the moving average (MA) order. The term pk is the
order of seasonal AR terms, dk is the order of seasonal differencing, and qk is the seasonal
order of MA terms. The seasonal values relate to “k,” or the frequency of the data. With
the current monthly data set, k = 12.

For certain schedules, such as those related to lighting, simpler regression and smoothing
methods are used because they offer the best fit for irregular usage without seasonal or
weather related behavior, is in a long-run steady decline, or is seasonal and unrelated to
weather.

Normal weather for the forecast is defined as a 20-year moving average of degree-days
taken from the National Oceanic and Atmospheric Administration’s Spokane International
Airport data. Normal weather updates only when a full year of new data is available. For
example, normal weather for 2015 is the 20-year average of degree-days for the 1995 to
2014 period; and 2016 is the 1996 to 2015 period.

9 Faruqui, Ahmad (2000). Making Forecasts and Weather Normalization Work Together, Electric Power
Research Institute, Publication No. 1000546, Tech Review, March 2000.
SC_PR_3-2 Attachment A Page 34 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-8
The choice of a 20-year moving average for defining normal weather reflects several
factors. First, recent climate research from the National Aeronautics and Space
Administration’s (NASA) Goddard Institute for Space Studies (GISS) shows a shift in
temperature starting about 20 years ago. The GISS research finds the summer
temperatures in the Northern Hemisphere increased one degree Fahrenheit above the
1951-1980 reference period; the increase started roughly 20 years ago in the 1981-1991
period.10 An in-house analysis of temperature in Avista’s Spokane-Kootenai service area,
using the same 1951-1980 reference period, also shows an upward shift in temperature
starting about 20-years ago. A detailed discussion of this analysis is in the peak-load
forecast section of this chapter.

The second factor in using a 20-year moving average is the volatility of the moving
average as function of the years used to calculate the average. Moving averages of ten
and 15 years showed considerably more year-to-year volatility than the 20-year average.
This volatility can obscure longer-term trends and lead to overly sharp changes in
forecasted loads when the updated definition of normal weather is applied each year.
These sharp changes would also cause excessive volatility in the revenue and earnings
forecasts.

As noted earlier, if RAP and IP appear in Equation 3.2, then they must also be in the
forecast for five years to generate the UPC forecast. The assumption in the five-year
forecast for this IRP is the RAP will increase two percent annually. This rate reflects the
average annual real growth rate for the 2005-2013 period.

Table 3.1: UPC Models Using Non-Weather Driver Variables

Schedule Variables Comment
Washington:
Residential Schedule 1 RAP
Commercial Schedule 31 RAP Commercial pumping schedule
Industrial Schedule 31 RAP
Industrial Schedules 11, 21, and 25 IP
Idaho:
Residential Schedule 1 RAP
Commercial Schedule 31 RAP Commercial pumping schedule
Industrial Schedules 11 and 21 IP

IP forecasts generate from a regression using the GDP forecast. Equation 3.3 and Figure
3.7 describes this process.

10 See Hansen, J.; M. Sato; and R. Ruedy (2013). Global Temperature Update Through 2012,
http://www.nasa.gov/topics/earth/features/2012-temps.html.
SC_PR_3-2 Attachment A Page 35 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-9
Equation 3.3: IP Regression Equation

𝐺𝐼𝑃𝑦,𝑈𝑆= 𝜈0 + 𝜈1𝐺𝐺𝐷𝑃𝑦,𝑈𝑆+ 𝜖𝑦
Where:
 GIPy,US = the annual growth in IP in year y.
 GGDPy,US = the annual growth in real GDP in year y.
 εy = a random error term.

Equation 3.3 uses historical data and incorporates forecasts for GDP to forecast GIP over
five years. GIP is an input for the generation of a forecast for the level of the IP index. The
forecasts for GGDP reflect the average of forecasts from multiple sources. Sources
include the Bloomberg survey of forecasts, the Philadelphia Federal Reserve survey of
forecasters, the Wall Street Journal survey of forecasters, and other sources. Averaging
forecasts reduces the systematic errors of a single-source forecast. This approach
assumes that macroeconomic factors flow through UPC in the industrial schedules. This
reflects the relative stability of industrial customer growth over the business cycle.

Figure 3.8 shows the historical relationship between the IP and industrial load for
electricity.11, The load values have been seasonally adjusted using the Census X12
procedure. The historical relationship is positive for both loads. The relationship is very
strong for electricity with the peaks and troughs in load occurring in the same periods as
the business cycle peaks and troughs.

Figure 3.7: Forecasting IP Growth

11 Data Source: U.S. Federal Reserve and Avista records.
12 Figure 3.8 excludes one large industrial customer with significant load volatility.
Growth Forecasts:
IMF, FOMC,
Bloomberg, etc.
Average
forecasts out 5-
yrs.
Index (IP) Growth Model:
Model links year y GDP
growth year y IP
growth.
Federal Reserve
industrial production
index is measure of IP
growth.

Low IP Forecast:
Forecast annual IP growth
using the GDP forecast
average (the baseline
scenario), a “high” scenario,
and a “low” scenario.
The high and low GDP
forecasts are the annual high
and low values from the
various sources used to
generate the average GDP
growth rate in each year.
Apply scenario that makes
most sense given the most
current economic analysis.
Convert annual growth
scenario to a monthly basis
to project out the monthly
GD
P IP
SC_PR_3-2 Attachment A Page 36 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-10
Figure 3.8: Industrial Load and Industrial (IP) Index

Customer Forecast Methodology
The econometric modeling for the customer models range from simple smoothing models
to more complex autoregressive integrated moving average (ARIMA) models. In some
cases, a pure ARIMA model without any structural independent variables is used. For
example, the independent variables are only the past values of the schedule customer
counts, the dependent variable. Because the customer counts in most schedules are
either flat or growing in a stable fashion, complex econometric models are generally
unnecessary for generating reliable forecasts. Only in the case of certain residential and
commercial schedules is more complex modeling required.

For the main residential and commercial schedules, the modeling approach needs to
account for customer growth between these schedules having a high positive correlation
over 12-month periods. This high customer correlation translates into a high correlation
over the same 12-month periods. Table 3.2 shows the correlation of customer growth
between residential, commercial, and industrial users of Avista electricity and natural gas.
To assure this relationship in the customer and load forecasts, the models for the
Washington and Idaho Commercial Schedules 11 use Washington and Idaho Residential
Schedule 1 customers as a forecast driver. Historical and forecasted Residential
Schedule 1 customers become drivers to generate customer forecasts for Commercial
Schedule 11 customers.

Figure 3.9 shows the relationship between annual population growth and year-over-year
customer growth.13 Customer growth has closely followed population growth in the
combined Spokane-Kootenai MSAs over the last 15 years. Population growth averaged
1.2% over the 2000-2016 period, and customer growth averaged 1.1 percent annually.

13 Data Source: Bureau of Economic Analysis, U.S. Census, Washington State OFM, and Avista records.
70
80
90
100
110
70 GWh
80 GWh
90 GWh
100 GWh
110 GWh
120 GWh
130 GWh
Ja
n
-
9
7
Se
p
-
9
7
Ma
y
-
9
8
Ja
n
-
9
9
Se
p
-
9
9
Ma
y
-
0
0
Ja
n
-
0
1
Se
p
-
0
1
Ma
y
-
0
2
Ja
n
-
0
3
Se
p
-
0
3
Ma
y
-
0
4
Ja
n
-
0
5
Se
p
-
0
5
Ma
y
-
0
6
Ja
n
-
0
7
Se
p
-
0
7
Ma
y
-
0
8
Ja
n
-
0
9
Se
p
-
0
9
Ma
y
-
1
0
Ja
n
-
1
1
Se
p
-
1
1
Ma
y
-
1
2
Ja
n
-
1
3
Se
p
-
1
3
Ma
y
-
1
4
Ja
n
-
1
5
Se
p
-
1
5
Ma
y
-
1
6
In
d
u
s
t
r
i
a
l

P
r
o
d
u
c
t
i
o
n

(
B
l
u
e

L
i
n
e
)
Lo
a
d
Industrial, SA Industrial, Trend-Cycle Industrial Production
SC_PR_3-2 Attachment A Page 37 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-11
Table 3.2: Customer Growth Correlations, January 2005 – December 2013

Customer Class
(Year-over-Year)
Residential Commercial Industrial Streetlights
Residential 1
Commercial 0.892 1
Industrial -0.285 -0.167 1
Streetlights -0.273 -0.245 0.209 1

Figure 3.9 demonstrates population growth can be used as a proxy for customer growth.
As a result, forecasted population is an adjustment to Expected Case forecasts of
Residential Schedule 1 customers in Washington and Idaho. An Expected Case forecast
is made using an ARIMA times-series model, for Schedule 1 in Washington and Idaho. If
the growth rates generated from this approach differ from forecasted population growth,
the Expected Case forecasts are adjusted to match forecasted population growth. Figure
3.10 summarizes the forecasting process for population growth for use in Residential
Schedule 1 customers.

Figure 3.9: Population Growth vs. Customer Growth, 2000-2016

0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
20
0
0
20
0
1
20
0
2
20
0
3
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
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
System Customers
SC_PR_3-2 Attachment A Page 38 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-12
Figure 3.10: Forecasting Population Growth

Forecasting population growth is a process that links U.S. GDP growth to service area
employment growth and then links regional and national employment growth to service
area population growth.
The forecasting models for regional employment growth are:

Equation 3.4: Spokane Employment Forecast

𝐺𝐸𝑀𝑃𝑦,𝑆𝑃𝐾= 𝜗0 + 𝜗1𝐺𝐺𝐷𝑃𝑦,𝑈𝑆+ 𝜗2𝐺𝐺𝐷𝑃𝑦−1,𝑈𝑆
+ 𝜗3𝐺𝐺𝐷𝑃𝑦−2,𝑈𝑆+ 𝜔𝑆𝐶𝐷𝐾𝐶,1998−2000=1+ 𝜔𝑆𝐶𝐷𝐻𝐵,2005−2007=1 + 𝜖𝑡,𝑦

Equation 3.5: Kootenai Employment Forecast

𝐺𝐸𝑀𝑃𝑦,𝐾𝑂𝑂𝑇= 𝛿0 + 𝛿1𝐺𝐺𝐷𝑃𝑦,𝑈𝑆+ 𝛿2𝐺𝐺𝐷𝑃𝑦−1,𝑈𝑆
+ 𝛿3𝐺𝐺𝐷𝑃𝑦−2,𝑈𝑆+ 𝜔𝑂𝐿𝐷1994=1+ 𝜔𝑂𝐿𝐷2009=1 + 𝜔𝑆𝐶𝐷𝐻𝐵,2005−2007=1 + 𝜖𝑡,𝑦
Where:
 SPK = the Spokane, WA MSA.
 KOOT = the Kootenai, ID MSA.
 GEMPy = employment growth in year y.
 GGDPy,US, GGDPy-1,US, and GGDPy-2,US = U.S. real GDP growth in years y, y-1,
and y-2.
 DKC = structural change (SC) dummy variables for the closing of Kaiser
Aluminum in Spokane.
 DHB = for the housing bubble, specific to each region.
 D1994=1 and D2009=1 = outlier (OL) dummy variables for 1994 and 2009 in
Kootenai.
 εy = a random error term.

Average GDP
Growth Forecasts:
IMF, FOMC,
Bloomberg, etc.
Average
forecasts out 5-
years.
Model links regional, U.S., and CA year
y-1 employment growth to year y county
population growth. Forecast out 6-years for Spokane, WA;
Kootenai, ID; and Jackson, OR.
Averaged with IHS forecasts in ID and
OFM forecasts in WA.
Growth rates used to generate
population forecasts for customer
Growth Model:
Model links year y, y-
1, and y-2 GDP
growth to year y
regional employment
growth.
Forecast out 6-
years.
Averaged with IHS
GDP
EMP
SC_PR_3-2 Attachment A Page 39 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-13
The same average GDP growth forecasts used for the IP growth forecasts are inputs to
the five-year employment growth forecast. Employment forecasts are averaged with IHS
Connect’s (formerly Global Insight) forecasts for the same counties. Averaging may
reduce the systematic errors of a single-source forecast. The averaged employment
forecasts become inputs to generate population growth forecasts. The forecasting models
for regional population growth are:

Equation 3.6: Spokane Population Forecast

𝐺𝑃𝑂𝑃𝑦,𝑆𝑃𝐾= 𝜅0 + 𝜅1𝐺𝐸𝑀𝑃𝑦−1,𝑆𝑃𝐾+ 𝜅2𝐺𝐸𝑀𝑃𝑦−2,𝑈𝑆+ 𝜔𝑂𝐿𝐷2001=1+𝜖𝑡,𝑦

Equation 3.7: Kootenai Population Forecast

𝐺𝑃𝑂𝑃𝑦,𝐾𝑂𝑂𝑇= 𝛼0 + 𝛼1𝐺𝐸𝑀𝑃𝑦−1,𝐾𝑂𝑂𝑇
+ 𝛼2𝐺𝐸𝑀𝑃𝑦−2,𝑈𝑆+ 𝜔𝑂𝐿𝐷1994=1 + 𝜔𝑂𝐿𝐷2002=1+ 𝜔𝑆𝐶𝐷𝐻𝐵,2007↑=1 + 𝜖𝑡,𝑦

Where:
 SPK = the Spokane, Washington MSA.
 KOOT = the Kootenai, Idaho MSA.
 GPOPy = employment growth in year y.
 GEMPy-1 and GEMPy-2 = employment growth in y-1 and y-2.
 D1994=1, D2001=1, and D2002=1 = outlier (OL) dummy variables for recession
impacts
 DHB,2007=1 = structural change (SC) dummy variable that adjusts for the after
effects of the housing bubble collapse in the Kootenai, Idaho MSA.

Equations 3.6 and 3.7 are estimated using historical data. Next, the GEMP forecasts (the
average of Avista and IHS forecasts) become inputs to Equations 3.6 and 3.7 to generate
population growth forecasts. The Kootenai forecast is averaged with IHS’s forecasts for
the same MSA. The Spokane forecast is averaged with Washington’s Office of Financial
Management forecast for the same MSA. These averages produce the final population
forecast for each MSA. These forecasts are then converted to monthly growth rates to
forecast population levels over the next five years.

IRP Long-Run Load Forecast

The Basic Model
The long-run load forecast extends the five-year projection out to 2035. It includes the
impacts from growing electric vehicle (EV) fleets and residential rooftop photovoltaic solar
(PV). The long-run modeling approach starts with Equation 3.8.

SC_PR_3-2 Attachment A Page 40 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-14
Equation 3.8: Residential Long-Run Forecast Relationship

ℓ𝑦= 𝑐𝑦+ 𝑢𝑦
Where:
 ℓy = residential load growth in year y.
 cy = residential customer growth in year y.
 uy = UPC growth in year y.

Equation 3.8 sets annual residential load growth equal to annual customer growth plus
the annual UPC growth.14 Cy is not dependent on weather, so where uy values are
weather normalized, ℓy results are weather-normalized. Varying cy and uy generates
different long-run forecast simulations. This IRP varies cy for economic reasons and uy
for increased usage of PV, EVs, and LED lighting.

Expected Case Assumptions
The Expected Case forecast makes assumptions about the long-run relationship between
residential, commercial, and industrial classes, as documented below.

1. Long-run residential and commercial customer growth rates are the same for 2022 to
2040, consistent with historical growth patterns over the past decade. Figure 3.11
shows the Expected Case time path of residential customer growth. The average
annual growth rate after 2021 is approximately 0.8 percent, assuming a gradual
decline starting in 2022. The values shown in Figure 3.11 were generated with the
Employment and Population forecast Equations 3.4, 3.5, 3.6, and 3.7 in conjunction
with IHS’s employment and population forecasts and Washington’s OFM population
forecasts. The annual industrial customer growth rate assumption is zero, matching
historical patterns for the past decade.

2. Commercial load growth follows changes in residential load growth, but with a spread
of 0.5 percent. This high correlation assumption is consistent with the high historical
correlation between residential and commercial load growth. The 0.5 percent spread
is within the range of historical norms and the forecasted growth spread from the five-
year model.

3. Consistent with historical behavior, industrial and streetlight load growth projections
are not correlated with residential or commercial load. Annual industrial load growth is
set at 0.5 percent and streetlight load growth at 0.1 percent for 2022-2037. Both
growth rates are in the range of historical norms and forecasted growth trends from
the five-year model.

4. The real residential price per kWh increases at 2 percent per year until 2027. Up to
2027, this is the same as the nominal price increasing 4 percent per year assuming a

14 Since UPC = load/customers, calculus shows the annual percentage change UPC ≈ percentage change
in load - percentage change in customers. Rearranging terms, the annual percentage change in load ≈
percentage change in customers + percentage change in UPC.

SC_PR_3-2 Attachment A Page 41 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-15
non-energy inflation rate of 2 percent. The real price increase assumption is zero
starting in 2027. This assumption means the nominal price is increasing at the same
rate as consumer inflation, excluding energy. This assumption relies on historical
trends in residential prices and current capital spending plans.

5. The own-price elasticity of UPC is set at -0.11. Own price elasticity was estimated from
the five-year UPC forecast equations for Residential Schedule 1 in Washington and
Idaho. Specifically, the own-price elasticity calculation uses the customer-weighted
average between Washington and Idaho.

6. From 2022 to 2024, depressed UPC growth results from new lighting and other
efficiency standards. The impact is more gradual than the Energy Information
Administration’s (EIA) modeling assumptions in its 2016 Annual Energy Outlook. The
EIA assumes a large decline in UPC growth in 2020 with a subsequent sharp rebound
in 2021 that Avista believes is too volatile.

7. Electric vehicles (EVs) grow at a rate consistent with present adoption rates. Using
Electric Power Research Institute data, Avista estimates that as of 2015 there were
around 400 EVs registered in its service area. The forecasted rate of adoption over
the 2020-2040 period is a function of and EV forecast provided by Avista’s EV
management team. This forecast reflects a low, middle, and, high forecast for EVs in
our electric service area. The low forecast predicts 20,000 EVs by 2040; the middle
predicts 70,000; and the high predicts 118,000. The final 2040 forecast used for the
IRP weights the low forecast at 70 percent, the middle a 20 percent weight; and the
high with a 10 percent weight. Therefore, the IRP forecast for 2040 is 0.70 x 20,000 +
0.20 x 70,000 + 0.10 x 118,000 = 39,800 EVs. Between 2016 and 2040, the implied
growth rate is 19 percent, which puts total EVs in 2037 as 22,395. The forecast
assumes each EV uses 2,500 kWh per year.

8. Rooftop PV penetration, measured as the share of PV residential customers to total
residential customers, continues to grow at present levels in the forecast. The average
PV system is forecast at the current median of 5.0 kW (DC) and a 13 percent capacity
factor, or about 5,578 kWh per year per customer. It assumed that this median system
size will increase annually to 6.0 kW (DC) by 2040, or about 6,694 kWh per year per
customer. This is equal to an annual growth rate in PV kWh of about 0.8 percent per
year. In addition, the IRP assumes the penetration rate (share of residential
customers) will follow the historical regression relationship between the historical
penetration rate in year t and the historical number of residential customers in year t
for the 2008-2015 period. Using this relationship, residential PV penetration will
increase from 0.09 percent in 2016 to about 0.42 percent in 2037. Residential solar
adoption in Avista’s service area continues at a very modest pace even though solar
prices have fallen significantly and state subsidies for solar are still in place. One
important factor restricting solar adoption in our service territory is the stable real price
of residential electricity. Adjusting the average residential price for CPI inflation, less
energy, shows that real prices have been largely flat since 2009. The IRP assumes
the real price of residential power will continue to rise at a very modest pace which, in
SC_PR_3-2 Attachment A Page 42 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-16
turn, will keep solar adoption in line with the historical data used to forecast future
solar adoption. Clarity on federal energy policy would help make possible adjustments
to the forecast now based on historical behavior alone.

Figure 3.11: Long-Run Annual Residential Customer Growth

Native Load Scenarios with Low/High Economic Growth
The high and low load scenarios use population growth Equations 3.6 and 3.7, holding
U.S. employment growth constant at 1.1 percent, but varying MSA employment growth
at higher and lower levels to gauge the impacts on population growth and utility loads.
See Table 3.3. The high/low range for service area employment growth reflects historical
employment growth variability. Simulated population growth is a proxy for residential and
customer growth in the long-run forecast model, and produces the high and low native
load forecasts shown in Figure 3.12.

Table 3.3: High/Low Economic Growth Scenarios (2017-2037)

Economic
Growth
Annual U.S.
Employment Growth
(percent)
Annual Service Area
Employment Growth
(percent)
Annual Population
Growth
(percent)
Expected Case 1.1 1.3 0.9
High Growth 1.1 2.0 1.6
Low Growth 1.1 0.1 0.8

0.5%
0.6%
0.7%
0.8%
0.9%
1.0%
1.1%
1.2%
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
An
n
u
a
l

G
r
o
w
t
h
SC_PR_3-2 Attachment A Page 43 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-17
Figure 3.12: Average Megawatts, High/Low Economic Growth Scenarios

Table 3.4 is the average annual load growth rate over the 2017-2037 period. The low
growth scenario predicts a slight load decline over 2022-2024 due to the impact of the
phased-in efficiency standards discussed in Item 6 of the Expected Case’s assumptions
listed above.

Table 3.4: Load Growth for High/Low Economic Growth Scenarios (2018-2037)

Economic Growth Average Annual Native Load
Growth
(percent)
Expected Case 0.47
High Growth 0.82
Low Growth 0.19

Long-Run Forecast Residential Retail Sales
Focusing on residential kWh sales, Figure 3.13 is the Expected Case residential UPC
growth plotted against the EIA’s annual growth forecast of U.S. residential use per
household growth. The EIA’s forecast is from the 2016 Annual Energy Outlook. Both
Avista’s and EIA’s forecasts show positive UPC growth returning around 2035. The EIA
forecast reflects a population shift to warmer-climate states where air conditioning is
typically required most of the year. In contrast, Avista’s forecast reflects the impact of
EVs.

1,000
1,050
1,100
1,150
1,200
1,250
1,300
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Expected Case
High Growth Rate
Low Growth Rate
SC_PR_3-2 Attachment A Page 44 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-18
Figure 3.13: UPC Growth Forecast Comparison to EIA

Figure 3.14 shows the EIA and Expected Case residential load growth forecasts of
residential load growth. Avista’s forecast is higher in the 2015-2020 period, reflecting an
assumption that service area population growth will be stronger than the U.S. average,
consistent with government and consultant’s forecasts for the far west and Rocky
Mountain regions where Avista’s service territory is located.

Figure 3.14: Load Growth Comparison to EIA

-2.5%
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
0.5%
1.0%
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
An
n
u
a
l

G
r
o
w
t
h
EIA Refrence Case Use Per Household Growth
UPC Growth, Base-Line with Renewables and EV/PHEVs
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
An
n
u
a
l

G
r
o
w
t
h
EIA Purchased Residential Electricity Growth (Quad. BTU)
Expected Case's Load Growth
SC_PR_3-2 Attachment A Page 45 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-19
Monthly Peak Load Forecast Methodology

The Peak Load Regression Model
The peak load forecast helps Avista determine the amount of resources necessary to
meet peak demand. In particular, Avista must build generation capacity to meet winter
and summer peak periods. Looking forward, the highest peak loads are most likely to
occur in the winter months, although in some years a mild winter followed by a hot
summer could find the annual maximum peak load occurring in a summer hour. On a
planning basis where extreme weather is expected to occur in the winter, peak loads
occur in the winter throughout the IRP timeframe. Equation 3.9 shows the current peak
load regression model.

Equation 3.9: Peak Load Regression Model

ℎ𝑀𝑊𝑑,𝑡,𝑦𝑛𝑒𝑡𝑝𝑒𝑎𝑘= 𝜆0 + 𝜆1𝐻𝐷𝐷𝑑,𝑡,𝑦+ 𝜆2(𝐻𝐷𝐷𝑑,𝑡,𝑦)2
+ 𝜆3𝐻𝐷𝐷𝑑−1,𝑡,𝑦+ 𝜆4𝐶𝐷𝐷𝑑,𝑡,𝑦 + 𝜆5𝐶𝐷𝐷𝑑,𝑡,𝑦𝐻𝐼𝐺𝐻+ 𝜆6𝐶𝐷𝐷𝑑−1,𝑡,𝑦+ 𝜙1𝐺𝐷𝑃𝑞(𝑡).𝑦−1
+ 𝜔𝑊𝐷𝐷𝑑,𝑡,𝑦+ 𝜔𝑆𝐷𝐷𝑡,𝑦+ 𝜔𝑂𝐿𝐷𝑀𝑎𝑟 2005=1 + 𝜔𝑂𝐿𝐷𝐹𝑒𝑏 2012=1 + 𝜔𝑂𝐿𝐷𝐽𝑎𝑛 2015=1
+ 𝜔𝑆𝐶𝐷𝑊𝑖𝑛𝑡𝑒𝑟 2016 + 𝜔𝑆𝐶𝐷𝑆𝑢𝑚𝑚𝑒𝑟 2016 + 𝜖𝑑,𝑡,𝑦

Where:
 hMWd,t,y
netpeak = metered peak hourly usage on day of week d, in month t, in year
y, and excludes two large industrial producers. The data series starts in June
2004.
 HDDd,t,y and CDDd,t,y = heating and cooling degree days the day before the
peak.
 (HDDd,t,y)2 = squared value of HDDd,t,y.HDDd−1,t,y and CDDd−1,t,y = heating and
cooling degree days the day before the peak.
 CDDd,t,yHIGH = maximum peak day temperature minus 65 degrees.15
 GDPq(t).y−1 = level of real GDP in quarter q covering month t in year y-1.
 ωWDDd,t,y = dummy vector indicating the peak’s day of week.
 ωSDDt,y = seasonal dummy vector indicating the month; and the other dummy
variables control for outliers in March 2005, February 2012, and January
2015.
 ωSDDWinter 2016 and ωSDDSummer 2016 = dummy variables to control for the extreme
La Nina effects on peak load.
 εd,t,y = uncorrelated N(0, σ) error term.

Generating Weather Normal Growth Rates Based on a GDP Driver
Equation 3.9 coefficients identify the month and day most likely to result in a peak load in
the winter or summer. By assuming normal peak weather and switching on the dummy
variables for day (dMAX) and month (tMAX) that maximize weather normal peak conditions
in winter and summer, a series of peak forecasts from the current year, yc, are generated

15 This term provides a better model fit than the square of CDD.
SC_PR_3-2 Attachment A Page 46 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-20
out N years by using forecasted levels of GDP as shown in Equation 3.3.16 All other
factors besides GDP remain constant to determine the impact of GDP on peak load. For
winter, this is defined as the forecasted series W:

𝑊 = {𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊),𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+2
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊),…,𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+𝑁
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊)}

For summer, this is defined as the forecasted series S:

𝑆 = {𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆),𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+2𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆),…,𝐹(ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+𝑁𝑊𝑁,𝑛𝑒𝑡 𝑝𝑒𝑎𝑘,𝑆)}

Both S and W are convertible to a series of annual growth rates, GhMW. Peak load growth
forecast equations are shown below as winter (WG) and summer (SG.):

𝑊𝐺= {𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊),𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+2𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊),…,𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+𝑁𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊)}

𝑆𝐺= {𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆),𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+2
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆),…,𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+𝑁
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆) }

In Equation 3.10, holding all else constant, growth rates are applied to simulated peak
loads generated for the current year, yc, for each month, January through December.
These peak loads are generated by running actual extreme weather days observed since
1890. The following section describes this process.

Simulated Extreme Weather Conditions with Historical Weather Data
Equation 3.10 generates a series of simulated extreme peak load values for heating
degree days.

Equation 3.10: Peak Load Simulation Equation for Winter Months

ℎ𝑀𝑊̂𝑡,𝑦𝑊= 𝑎 + 𝜆1̂𝐻𝐷𝐷𝑡,𝑦,𝑀𝐼𝑁 + 𝜆2̂(𝐻𝐷𝐷𝑡,𝑦,𝑀𝐼𝑁 )2 𝑓𝑜𝑟 𝑡 = 𝐽𝑎𝑛,…,𝐷𝑒𝑐 𝑖𝑓 𝑚𝑎𝑥𝑖𝑢𝑚 𝑎𝑣𝑔.𝑡𝑒𝑚𝑝
<65 𝑎𝑛𝑑 𝑦 =1890,…,𝑦𝑐

Where:
 hMŴt,yW = simulated winter peak megawatt load using historical weather data.
 HDDt,y,MIN = heating degree days calculated from the minimum (MIN) average
temperature (average of daily high and low) on day d, in month t, in year y if
in month t the maximum average temperature (average of daily high and low)
is less than 65 degrees.
 a = aggregate impact of all the other variables held constant at their average
values.

16 Forecasted GDP is generated by applying the averaged GDP growth forecasts used for the employment and
industrial production forecasts discussed previously.
SC_PR_3-2 Attachment A Page 47 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-21
Similarly, the model for cooling degree days is:

Equation 3.11: Peak Load Simulation Equation for Summer Months

ℎ𝑀𝑊̂𝑡,𝑦𝑆= 𝑎 + 𝜆4̂𝐶𝐷𝐷𝑡,𝑦,𝑀𝐴𝑋 𝑓𝑜𝑟 𝑡 = 𝐽𝑎𝑛,…,𝐷𝑒𝑐 𝑖𝑓 𝑚𝑎𝑥𝑖𝑢𝑚 𝑎𝑣𝑔. 𝑡𝑒𝑚𝑝 >65 𝑎𝑛𝑑 𝑦
= 1890,…,𝑦𝑐

Where:
 hMŴt,yS = simulated winter peak megawatt load using historical weather data.
 CDDt,y,MAX = cooling degree days calculated from the maximum (MAX) average
temperature. The average of daily high (H) and low (L) on day d, in month t, in
year y if in month t if the maximum average temperature (average of daily high
and low) is greater than 65 degrees.
 a = aggregate impact of all the other variables held constant at their average
values.

With over 100 years of average maximum and minimum temperature data, Equations
3.10 and 3.11 applied to each month t will produce over 100 simulated values of peak
load that can be averaged to generate a forecasted average peak load for month t in the
current year, yc. The average for each month are shown by Equations 3.12 and 3.13.

Equation 3.12: Current Year Peak Load for Winter Months

𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐
𝑊) =1
(𝑦𝑐−1890)+ 1 ∑ ℎ𝑀𝑊̂𝑡,𝑦𝑊𝑦𝑐
𝑦=1890
𝑓𝑜𝑟 𝑒𝑎𝑐ℎ ℎ𝑒𝑎𝑡𝑖𝑛𝑔 𝑚𝑜𝑛𝑡ℎ 𝑡
𝑤ℎ𝑒𝑟𝑒 𝑚𝑎𝑥𝑖𝑢𝑚 𝑎𝑣𝑔.𝑡𝑒𝑚𝑝 <65

Equation 3.13: Current Year Peak Load for Summer Months

𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐
𝑆) =1
(𝑦𝑐−1890)+ 1 ∑ ℎ𝑀𝑊̂𝑡,𝑦𝑆𝑦𝑐
𝑦=1890
𝑓𝑜𝑟 𝑒𝑎𝑐ℎ 𝑐𝑜𝑜𝑙𝑖𝑛𝑔 𝑚𝑜𝑛𝑡ℎ 𝑡
𝑤ℎ𝑒𝑟𝑒 𝑚𝑎𝑥𝑖𝑢𝑚 𝑎𝑣𝑔.𝑡𝑒𝑚𝑝 >65

Forecasts beyond yc are generated using the appropriate growth rate from series WG and
SG. For example, the forecasts for yc+1 for winter and summer are:

𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐+1
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊) = 𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐
𝑊) ∗ [1 + 𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑊)]

𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐+1
𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆) = 𝐹(ℎ𝑀𝑊𝑡,𝑦𝑐
𝑆) ∗ [1 + 𝐹(𝐺ℎ𝑀𝑊𝑑𝑀𝐴𝑋,𝑡𝑀𝐴𝑋,𝑦𝑐+1𝑊𝑁,𝑛𝑒𝑡𝑝𝑒𝑎𝑘,𝑆)]

The peak load forecast is finalized when the loads of two large industrial customers
excluded from the Equation 3.12 and 3.13 estimations are added back in.

SC_PR_3-2 Attachment A Page 48 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-22
Table 3.5 shows estimated peak load growth rates with and without the two large
industrial customers. Figure 3.15 shows the forecasted time path of peak load out to 2040,
and Figure 3.16 shows the high/low bounds based on a one-in-20 event (95 percent
confidence interval) using the standard deviation of the simulated peak loads from
Equations 3.12 and 3.13.

Table 3.5: Forecasted Winter and Summer Peak Growth, 2017-2037

Category Winter
(Percent)
Summer
(Percent)
Excluding Large Industrial Customers 0.42 0.46
Including Large Industrial Customers 0.38 0.42

Table 3.6 shows the summer peak is forecast to grow faster than the winter peak. Under
current growth forecasts, the orange summer line in Figure 3.15 will converge with the
blue winter line in approximately year 2100. Figure 3.16 shows that the winter high/low
bound considerably larger than summer, and reflects a greater range of temperature
anomalies in the winter months. Table 3.6 shows the energy and peak forecasts.

Figure 3.15: Peak Load Forecast 2017-2037

1,000
1,100
1,200
1,300
1,400
1,500
1,600
1,700
1,800
1,900
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
20
2
5
20
2
7
20
2
9
20
3
1
20
3
3
20
3
5
20
3
7
Me
g
a
w
a
t
t
s
Winter Peak
Summer Peak
SC_PR_3-2 Attachment A Page 49 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-23
Figure 3.16: Peak Load Forecast with 1 in 20 High/Low Bounds, 2017-2037

Table 3.6: Energy and Peak Forecasts

Year
Energy
(aMW)
Winter Peak
(MW)
Summer Peak
(MW)
2018 1,087 1,690 1,616
2019 1,094 1,697 1,623
2020 1,101 1,703 1,630
2021 1,109 1,710 1,637
2022 1,109 1,716 1,643
2023 1,109 1,723 1,650
2024 1,108 1,729 1,657
2025 1,114 1,736 1,664
2026 1,120 1,743 1,671
2027 1,126 1,749 1,678
2028 1,132 1,756 1,685
2029 1,138 1,763 1,692
2030 1,144 1,770 1,699
2031 1,150 1,776 1,707
2032 1,156 1,783 1,714
2033 1,162 1,790 1,721
2034 1,169 1,797 1,728
2035 1,175 1,804 1,735
2036 1,182 1,811 1,743
2037 1,189 1,818 1,750

1,000
1,200
1,400
1,600
1,800
2,000
2,200
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
20
2
5
20
2
7
20
2
9
20
3
1
20
3
3
20
3
5
20
3
7
Me
g
a
w
a
t
t
s
Winter Peak Summer Peak
Winter- High Winter- Low
Summer- High Summer- Low
SC_PR_3-2 Attachment A Page 50 of 205
Chapter 3: Economic & Load Forecast

Avista Corp 2017 Electric IRP 3-24
Extreme Temperature Analysis
The impact of temperature changes and the relevance of historical temperature data
drives much of the recent load forecasting debates regarding peak load forecasts. To
validate the use of historical temperatures in the peak load forecast, an analysis using the
same GISS methodology and reference periods referenced in the UPC forecast
methodology section. In particular, using 1951-1980 as the reference period, Avista
examined daily temperature anomalies using daily temperature data from the Spokane
International Airport going back to 1947.

The analysis focuses on the core summer months (June, July, and August) and winter
months (December, January, and February). The GISS study only considers summer
months and found, in addition to an increase in the average temperature in the summer,
the variance around the average increased. Specifically, the frequency of extreme
temperature anomalies three or more standard deviations above the summer average
increased compared to the 1950-1981 reference period. In contrast, Avista’s analysis
shows average temperature increases compared to the reference period, but there was
no significant shift in the frequency of extreme temperature events. This finding supports
continued use of historical temperature extremes for peak load forecasting.

SC_PR_3-2 Attachment A Page 51 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-1
4. Existing Supply Resources

Introduction & Highlights
Avista relies on a diverse portfolio of assets to meet customer loads, including owning
and operating eight hydroelectric developments on the Spokane and Clark Fork rivers.
Its thermal assets include partial ownership of two coal-fired units, five natural gas-fired
projects, and a biomass plant. Avista purchases energy from several independent
power producers (IPPs), including Palouse Wind, Rathdrum Power, and the City of
Spokane.

Figure 4.1 shows Avista capacity and energy mixes. Winter capability is the share of
total capability of each resource type the utility can rely upon to meet peak load absent
outages. The annual energy chart represents the energy as a percent of total supply;
this calculation includes fuel limitations (for water, wind, and wood), maintenance and
forced outages. Avista’s largest supply in the peak winter months is hydroelectric at 51
percent, followed by natural gas. On an energy capability basis, natural gas-fired
generation can produce more energy, at 43 percent, than hydroelectric at 38 percent,
because it is not constrained by fuel limitations. In any given year, the resource mix will
change depending on streamflow conditions and market prices.

Figure 4.1: 2018 Avista Capability & Energy Fuel Mix

Owned Hydro
40%
Contracted
Hydro
11%
Natural Gas
37%
Coal
9%
Biomass & Wind
3%
Winter Capability
Owned Hydro
28%
Contracted
Hydro10%
Natural Gas
43%
Coal
13%
Biomass & Wind
6%
Annual Energy
Section Highlights
Hydroelectric represents about half of Avista’s winter generating capability.
fired plants represent the largest portion of Avista’s thermal
Six percent of Avista’s generating potential is biomass and wind.


SC_PR_3-2 Attachment A Page 52 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-2
Avista reports its fuel mix annually in the Washington State Fuel Mix Disclosure. The
State calculates the resource mix used to serve load, rather than generation potential,
by adding regional estimates for unassigned market purchases and Avista-owned
generation minus environmental attributes from renewable energy credit (REC) sales.

Spokane River Hydroelectric Developments
Avista owns and operates six hydroelectric developments on the Spokane River. Five
operate under 50-year FERC operating licenses issued in June 2009. The sixth, Little
Falls, operates under separate authorization from the U.S. Congress. This section
describes the Spokane River developments and provides the maximum on-peak and
nameplate capacity ratings for each plant. The maximum on-peak capacity of a
generating unit is the total amount of electricity it can safely generate with its existing
configuration and state of the facility. This capacity is often higher than the nameplate
rating for hydroelectric developments because of plant upgrades and favorable head or
flow conditions. The nameplate, or installed capacity, is the capacity of a plant as rated
by the manufacturer. All six hydroelectric developments on the Spokane River connect
directly to the Avista electrical system.

Post Falls
Post Falls is the facility furthest upstream on the Spokane River. It is located several
miles east of the Washington/Idaho border. It began operating in 1906 and during
summer months maintains the elevation of Lake Coeur d’Alene. Post Falls has a 14.75-
MW nameplate rating and is capable of producing up to 18.0 MW with its six generating
units.

Upper Falls
The Upper Falls development sits within the boundaries of Riverfront Park in downtown
Spokane. It began generating in 1922. The project is comprised of a single 10.0-MW
nameplate unit with a 10.26-MW maximum capacity rating.

Monroe Street
Monroe Street was Avista’s first generation development. It began serving customers in
1890 in downtown Spokane near Riverfront Park. Rebuilt in 1992, the single generating
unit has a 14.8-MW nameplate rating and a 15.0-MW maximum capacity rating.

SC_PR_3-2 Attachment A Page 53 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-3
Monroe Street Development and Huntington Park, Downtown Spokane, WA

Nine Mile
A private developer built the Nine Mile development in 1908 near Nine Mile Falls,
Washington. Avista purchased the project in 1925 from the Spokane & Inland Empire
Railroad Company. Nine Mile has undergone recent substantial upgrades. The
development has two new 8-MW units and two 10-MW units for a total nameplate rating
of 36 MW.

Long Lake
The Long Lake development is located northwest of Spokane and maintains the Lake
Spokane reservoir, also known as Long Lake. The project’s four units have a nameplate
rating of 81.6 MW and 88.0 MW of combined capacity.

Little Falls
The Little Falls development, completed in 1910 near Ford, Washington, is the furthest
downstream hydroelectric facility on the Spokane River. The facility’s four units
generate 35.2 MW of on-peak capacity and have a 32.0 MW nameplate rating. Avista is
carrying out a series of upgrades to the Little Falls development. Much of the new
electrical equipment and the installation of a new generator excitation system are
complete. Projects include replacing station service equipment, updating the
powerhouse crane, and developing new control schemes and panels are complete.
Work is now ongoing to replace generators, turbines, and unit protection and control
systems on the four units will start.

SC_PR_3-2 Attachment A Page 54 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-4
Clark Fork River Hydroelectric Development
The Clark Fork River Development includes hydroelectric projects located near Clark
Fork, Idaho, and Noxon, Montana, 70 miles south of the Canadian border. The plants
operate under a FERC license through 2046. Both hydroelectric projects on the Clark
Fork River connect to the Avista transmission system.

Noxon Rapids
The Noxon Rapids development includes four generators installed between 1959 and
1960, and a fifth unit entered service in 1977. Avista completed major turbine upgrades
on units 1 through 4 between 2009 and 2012. The upgrades increased the capacity of
each unit from 105 MW to 112.5 MW and added 6.6 aMW of additional energy.

Cabinet Gorge
Cabinet Gorge started generating power in 1952 with two units, and added two
additional generators the following year. Upgrades to units 1 through 4 occurred in
1994, 2004, 2001, and 2007. The current maximum on-peak plant capacity is 270.5
MW; it has a nameplate rating of 265.2 MW.

Total Hydroelectric Generation
Avista’s hydroelectric plants have 1,080 MW of on-peak capacity. Table 4.1 summarizes
the location and operational capacities of Avista’s hydroelectric projects and the
expected energy output of each facility based on an 80-year hydrologic record.

Table 4.1: Avista-Owned Hydroelectric Resources

Monroe Street Spokane Spokane, WA 14.8 15.0 11.2
Post Falls Spokane Post Falls, ID 14.8 18.0 9.4
Nine Mile Spokane Nine Mile Falls, WA 36.0 32.0 15.7
Little Falls Spokane Ford, WA 32.0 35.2 22.6
Long Lake Spokane Ford, WA 81.6 89.0 56.0
Upper Falls Spokane Spokane, WA 10.0 10.2 7.3

Noxon Rapids Clark Fork Noxon, MT 518.0 610.0 196.5
Cabinet Gorge Clark Fork Clark Fork, ID 265.2 270.5 123.6

SC_PR_3-2 Attachment A Page 55 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-5
Thermal Resources
Avista owns seven thermal generation assets located across the Northwest. Based on
IRP analyses, Avista expects each plant to continue operation through the 20-year IRP
horizon. The resources provide dependable energy and capacity serving base- and
peak-load obligations. A summary of their capabilities is in Table 4.2.

Table 4.2: Avista-Owned Thermal Resources

Colstrip 3 (15%) Colstrip, MT Coal 1984 111.0 111.0 123.5
Colstrip 4 (15%) Colstrip, MT Coal 1986 111.0 111.0 123.5
Rathdrum Rathdrum, ID Gas 1995 176.0 130.0 166.5
Northeast Spokane, WA Gas 1978 66.0 42.0 61.2
Boulder Park Spokane, WA Gas 2002 24.6 24.6 24.6
Coyote Springs
21
Boardman, OR Gas 2003 317.5 286.0 287.3
Kettle Falls Kettle Falls, WA Wood 1983 47.0 47.0 50.7
Kettle Falls CT2 Kettle Falls, WA Gas 2002 11.0 8.0 7.5

Colstrip Units 3 and 4
The Colstrip plant, located in eastern Montana, consists of four coal-fired steam plants
connected to a double-circuit 500 kV BPA transmission line under a long-term wheeling
agreement. Talen Energy Corporation operates the facilities on behalf of the six owners.
Avista has no ownership interest in Units 1 or 2, but owns 15 percent of Units 3 and 4.
Unit 3 began operating in 1984 and Unit 4 was finished in 1986. Avista’s share of
Colstrip has a maximum net capacity of 222.0 MW, and a nameplate rating of 247.0
MW.

Rathdrum
Rathdrum consists of two simple-cycle combustion turbine (CT) units. This natural gas-
fired plant located near Rathdrum, Idaho connects to the Avista transmission system. It
entered service in 1995 and has a maximum capacity of 176.0 MW in the winter and
126.0 MW in the summer. The nameplate rating is 166.5 MW.

Northeast
The Northeast plant, located in Spokane, has two aero-derivative simple-cycle CT units
completed in 1978. It connects to Avista’s transmission system. The plant is capable of
burning natural gas or fuel oil, but current air permits preclude the use of fuel oil. The

1 For purposes of long-term transmission reservation planning for bundled retail service to native load
customers, replacement resources for Coyote Springs 2 is presumed and planned to be integrated via
Avista’s interconnection(s) to the Mid-Columbia region.
2 The Kettle Falls CT capacity quantities include output of the natural gas-fired turbine plus the benefit of
its steam to the main unit’s boiler.
SC_PR_3-2 Attachment A Page 56 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-6
combined maximum capacity of the units is 68.0 MW in the winter and 42.0 MW in the
summer, with a nameplate rating of 61.2 MW. The plant permit limits run hours to 100
per year.

Boulder Park
The Boulder Park project entered service in the Spokane Valley in 2002 and connects
directly to the Avista transmission system. The site uses six natural gas-fired internal
combustion reciprocating engines to produce a combined maximum capacity and
nameplate rating of 24.6 MW.

Coyote Springs 2
Coyote Springs 2 is a natural gas-fired combined cycle combustion turbine (CCCT)
located near Boardman, Oregon. The plant connects to the BPA 500 kV transmission
system under a long-term agreement. The plant began service in 2003; it has a
maximum capacity of 317.5 MW in the winter and 285 MW in the summer, with duct
burners. The nameplate rating of the plant is 287.3 MW. In 2016, the Advanced Hot Gas
Path is the latest upgrade to the plant increasing both the unit’s capacity and efficiency.

Kettle Falls Generation Station and Kettle Falls Combustion Turbine
The Kettle Falls Generating Station, a woody biomass facility, entered service in 1983
near Kettle Falls, Washington. It is among the largest biomass generation plants in
North America and connects to Avista on its 115 kV transmission system. The open-
loop biomass steam plant uses waste wood products from area mills and forest slash,
but can also burn natural gas. A 7.5 MW combustion turbine (CT), added to the facility
in 2002, burns natural gas and increases overall plant efficiency by sending exhaust
heat to the wood boiler.

The wood-fired portion of the plant has a maximum capacity of 50.0 MW, and its
nameplate rating is 50.7 MW. The plant typically operates between 45 and 47 MW
because of fuel conditions that change depending on the moisture content of the fuel.
The plant’s capacity increases to 55.0 to 58.0 MW when operated in combined-cycle
mode with the CT. The CT produces 8 MW of peaking capability in the summer and 11
MW in the winter. The CT resource can be limited in the winter when the natural gas
pipeline is capacity constrained. For IRP modeling, the CT does not run when
temperatures fall below zero. This operational assumption reflects natural gas
availability limits on the plant when local natural gas distribution demand is highest.

Power Purchase and Sale Contracts
Avista uses purchase and sale arrangements of varying lengths to meet a portion of its
load requirements. Contracts provide many benefits, including environmentally low-
impact and low-cost hydroelectric and wind power. This chapter describes the contracts
in effect during the timeframe of the 2017 IRP. Tables 4.3 through 4.5 summarize
Avista’s contracts.

SC_PR_3-2 Attachment A Page 57 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-7
Mid-Columbia Hydroelectric Contracts
During the 1950s and 1960s, Public Utility Districts (PUDs) in central Washington
developed hydroelectric projects on the Columbia River. Each plant was large
compared to loads then served by the PUDs. Long-term contracts with public,
municipal, and investor-owned utilities throughout the Northwest assisted with project
financing and ensured a market for the surplus power. The contract terms obligate the
PUDs to deliver power to Avista points of interconnection.

Avista originally entered into long-term contracts for the output of four of these projects
“at cost.” Avista now competes in capacity auctions to retain the rights of these expiring
contracts. The Mid-Columbia contracts in Table 4.3 provide energy, capacity and
reserve capabilities; in 2017, the contracts provide approximately 154 MW of capacity
and 101 aMW of energy. Recently, Avista successfully negotiated an extension of the
Chelan PUD contract. However, there are no guarantees to extend contract rights
beyond this term. Due to the uncertainty around future availability and cost, the IRP
does not include these contracts in the resource mix beyond their current expiration
dates. Avista was also able to extend its legacy Douglas PUD contract set to expire in
2018. The new contract provides capacity and energy through September 2028 at a
decreasing portion each year until it expires.

The timing of the power received from the Mid-Columbia projects is a result of
agreements including the 1961 Columbia River Treaty and the 1964 Pacific Northwest
Coordination Agreement (PNCA). Both agreements optimize hydroelectric project
operations in the Northwest U.S. and Canada. In return for these benefits, Canada
receives return energy under the Canadian Entitlement. The Columbia River Treaty and
the PNCA manage storage water in upstream reservoirs for coordinated flood control
and power generation optimization. On September 16, 2024, the Columbia River Treaty
may end. Studies are underway by U.S. and Canadian entities to determine possible
post-2024 Columbia River operations. Federal agencies are soliciting feedback from
stakeholders and soon negotiations will begin to determine the future of the treaty. This
IRP does not model alternative outcomes for the treaty negotiations, because it will not
likely affect long-term resource acquisition and we cannot speculate on future wholesale
electricity market impacts of the treaty.

Lancaster Power Purchase Agreement
Avista acquired output rights to the Lancaster CCCT, located in Rathdrum, Idaho, after
the sale of Avista Energy in 2007. Lancaster directly interconnects with the Avista
transmission system at the BPA Lancaster substation. Under the tolling contract, Avista
pays a monthly capacity payment for the sole right to dispatch the plant through October
2026. In addition, Avista pays a variable energy charge and arranges for all of the fuel
needs of the plant.

SC_PR_3-2 Attachment A Page 58 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-8
Table 4.3: Mid-Columbia Capacity and Energy Contracts3

Counter
Party
Project(s) Percent
Share
(%)
Start Date End Date Estimated
On-Peak
Capability
(MW)
Annual
Energy
(aMW)
Grant PUD Priest Rapids 3.7 Dec-2001 Dec-2052 34.8 19.5
Grant PUD Wanapum 3.7 Dec-2001 Dec-2052 34.5 18.7
Chelan PUD Rocky Reach 5.0 Jan-2016 Dec-2030 58.1 35.8
Chelan PUD Rock Island 5.0 Jan- 2016 Dec-2030 20.1 18.4
Douglas PUD Wells 3.34 Feb-1965 Sep-2028 27.9 14.3
Canadian Entitlement -10.1 -5.7
2018 Total Net Contracted Capacity and Energy 165.3 101.0

Public Utility Regulatory Policies Act (PURPA)
The passage of PURPA by Congress in 1978 required utilities to purchase power from
resources meeting certain size and fuel criteria. Avista has many PURPA contracts, as
shown in Table 4.4. The IRP assumes renewal of these contracts after their current
terms end.

Table 4.4: PURPA Agreements

Meyers Falls Hydro Kettle Falls, WA 12/2019 1.30 1.05
Spokane Waste to Energy Waste Spokane, WA 12/2017 18.00 16.00
Spokane County Digester Biomass Spokane, WA 8/2021 0.26 0.14
Plummer Saw Mill Wood Waste Plummer, ID 12/2019 5.80 4.00
Deep Creek Hydro Northpoint, WA 12/2017 0.41 0.23
Clark Fork Hydro Hydro Clark Fork, ID 12/2017 0.22 0.12
Upriver Dam5 Hydro Spokane, WA 12/2019 17.60 6.17
Big Sheep Creek Hydro Hydro Northpoint, WA 6/2021 1.40 0.79
Ford Hydro LP Hydro Weippe, ID 6/2022 1.41 0.39
John Day Hydro Hydro Lucille, ID 9/2022 0.90 0.25
Phillips Ranch Hydro Northpoint, WA n/a 0.02 0.01

3 For purposes of long-term transmission reservation planning for bundled retail service to native load
customers, replacement resources for each of the resources identified in Table 4.3 are presumed and
planned to be integrated via Avista’s interconnection(s) to the Mid-Columbia region.
4 The share from Wells is dependent on Douglas PUD’s load growth.
5 Energy estimate is net of the city of Spokane’s pumping load.
SC_PR_3-2 Attachment A Page 59 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-9
Bonneville Power Administration – WNP-3 Settlement
Avista signed settlement agreements with BPA and Energy Northwest on September
17, 1985, ending its nuclear plant construction delay claims against both parties. The
settlement provides an energy exchange through June 30, 2019, with an agreement to
reimburse Avista for WPPSS – Washington Nuclear Plant No. 3 (WNP-3) preservation
costs and an irrevocable offer of WNP-3 capability under the Regional Power Act.

The energy exchange portion of the settlement contains two basic provisions. The first
provision provides approximately 42 aMW of energy to Avista from BPA through 2019,
subject to a contract minimum of 5.8 million megawatt-hours. Avista is obligated to pay
BPA operating and maintenance costs associated with the energy exchange as
determined by a formula that ranges from $16 to $29 per megawatt-hour in 1987-year
constant dollars. The second provision provides BPA approximately 32 aMW of return
energy at a cost equal to the actual operating cost of Avista’s highest-cost resource. A
discussion of this obligation, and how Avista plans for it, is in Chapter 6.

Palouse Wind – Power Purchase Agreement
Avista signed a 30-year power purchase agreement in 2011 with Palouse Wind for the
entire output of its 105-MW project. Avista has the option to purchase the project after
10 years. Commercial operation began in December 2012. The project is EIA-qualified
and directly connected to Avista’s transmission system.

Table 4.5: Other Contractual Rights and Obligations

Contract Type Fuel
Source
End
Date
Winter
Capacity
(MW)
Summer
Capacity
(MW)
Annual
Energy
(aMW)
Douglas Settlement Purchase Hydro 9/2018 2 2 3
Energy America Sale CEC RECs6 12/2019 50 50 50
WNP-3 Purchase System 6/2019 82 0 42
Lancaster Purchase Natural Gas 10/2026 283 233 218
Palouse Wind Purchase Wind 12/2042 0 0 40
Nichols Pumping Sale System n/a -1 -1 -1
Total 416 284 352

Customer-Owned Generation
A small but growing number of customers install their own generation systems. In 2007
and 2008, the average number of new net-metering customers added was 10 yearly;
and between 2009 and 2014, the average is 41 per year, but over the last two years, an
increasing amount, 76 in 2015, and 112 in 2016. The recent increase likely driven by
solar price reductions and the near term expiring of the generous federal and new state
tax incentives. Certain renewable projects qualify for the federal government’s 30
percent tax credit and Washington tax incentives of up to $5,000 per year through July

6 CEC RECs are renewable resources based on approval of the California Energy Commission. Kettle
Falls, Palouse Wind, Nine Mile Falls, Post Falls, Monroe Street, and Upper Falls are CEC certified.
SC_PR_3-2 Attachment A Page 60 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-10
2020. The Washington utility taxes credit finances these incentives that rise to as much
as $1.08 per kWh.

Avista had 490 customer-installed net-metered generation projects on its system in
early June 2017 representing a total installed capacity of 3.5 MW. Eighty-eight percent
of installations are in Washington, with most located in Spokane County. Figure 4.2
shows annual net metering customer additions through 2016. Solar is the primary net
metered technology; the remaining is a mix of wind, combined solar and wind systems,
and biogas. The average annual capacity factor of the solar facilities is 13 percent.
Small wind turbines typically produce at less than a 10 percent capacity factor,
depending on location. Given the current tax incentives when the IRP modeling
occurred were nearing optimal payback, the number of new net-metered systems rose
significantly in 2016. The signature of SB 5939 on July 7, 2017 established a new solar
incentive program from October 1, 2017 through 2029 at a lower rate than the current
subsidy. If the number of net-metering customers continues to increase, Avista may
need to adjust rate structures for customers who rely on the utility’s infrastructure, but
do not contribute financially for infrastructure costs.

Figure 4.2: Avista’s Net Metering Customers

-
583
1,166
1,749
2,332
2,915
3,498
0
20
40
60
80
100
120
19
9
9
20
0
0
20
0
1
20
0
2
20
0
3
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
In
s
t
a
l
l
e
d

C
a
p
a
c
i
t
y

(
k
W
)
An
n
u
a
l

N
e
w

C
u
s
t
o
m
e
r
s
ID
WA
Cumulative
SC_PR_3-2 Attachment A Page 61 of 205
Chapter 4: Existing Supply Resources
Avista Corp 2017 Electric IRP 4-11
Solar
As solar equipment and installation prices have decreased, the nation’s interest and
development of the technology has increased dramatically. Avista has three small
projects of its own and is working with a developer to construct a fourth. The first was
three kilowatts on its corporate headquarters as part of the Solar Car initiative. The solar
production helped power two electric vehicles in the corporate fleet. Avista installed a
15-kilowatt solar system in Rathdrum, Idaho to supply Buck-A-Block, a voluntary
program allowing customers to purchase green energy. The 423-kW Avista Community
Solar project entered service in 2015. The project takes advantage of federal and state
subsidies. The $1,080/MWh Washington solar subsidy allows customers to purchase
individual solar panels within the facility and receive payments that more than offset
their upfront investment. The program utilizes approximately $600,000 each year in
state tax incentives. SB 5939, signed by Governor Inslee on July 7, 2017, updates the
solar incentive program for residential, commercial and shared commercial projects
starting on or after October 1, 2017. The new solar program pays an incentive for eight
years with projects starting later receiving a smaller incentive.

In April 2017, the company released a Request for Proposals to develop up to a 15 MW
(DC) solar facility for the company’s new Solar Select Program. This project will
voluntarily allow commercial and industrial customers to assign the solar costs and
production of the facility to their bill as a substitute for the utility’s regular power supply
cost. The participating customer will continue to pay their regular bill, but get a rate
credit for the variable power supply portion of their rate and then substitute a “lock-in”
solar rate for up to 20 years and the rate will not increase beyond its rate schedule for
the term. This new rate schedule once approved by state Commissions will allow
participating customers to acquire renewable energy and hedge power supply costs
from future increases. Avista plans to file this tariff by the end of 2017.

Boulder Park Community Solar Project

SC_PR_3-2 Attachment A Page 62 of 205

SC_PR_3-2 Attachment A Page 63 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
5. Energy Efficiency & Demand Response

Introduction
Avista began offering energy efficiency programs to its customers in 1978. These
programs pursue all cost-effective energy efficiency and operate within the prevailing
market and economic conditions. Recent programs with the highest impacts on energy
savings include residential and non-residential prescriptive lighting, residential fuel
efficiency, site-specific lighting, and small business projects. In addition, the Oracle
(formerly Opower) Home Energy Report program began sending peer-comparison
reports to participating customers every two months beginning in June 2013.
Conservation programs regularly meet or exceed regional shares of the energy efficiency
gains outlined by the Northwest Power and Conservation Council (NPCC).

Figure 5.1 illustrates Avista’s historical electricity conservation acquisitions. Avista has
acquired 219 aMW of energy efficiency since 1978; however, the 18-year average
measure life of the conservation portfolio means some measures no longer are reducing
load. The 18-year measure life accounts for the difference between the cumulative and
online trajectories in Figure 5.1. Currently 145 aMW of conservation serves customers,
representing nearly 12.3 percent of 2016 load.

Avista energy efficiency programs provide conservation and education options to the
residential, low income, commercial, and industrial customer segments. Program delivery
includes prescriptive, site-specific, regional, upstream, behavioral, market transformation,
and third-party direct install options. Prescriptive programs, or standard offerings, provide
cash incentives for standardized products such as the installation of qualifying high-
efficiency heating equipment. Prescriptive programs work in situations where uniform
products or offerings are applicable for large groups of homogeneous customers and
primarily occur in programs for residential and small commercial customers.

Site-specific programs, or customized offerings, provide cash incentives for any cost-
effective energy saving measure or equipment with an economic payback greater than
one year and less than eight years for non-LED lighting projects, or less than 13 years for
Section Highlights





SC_PR_3-2 Attachment A Page 64 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
all other end uses and technologies. Other delivery methods build off these approaches
but may include upstream buy downs of low cost measures, free-to-customer direct install
programs, and coordination with regional entities for market transformation efforts.

Figure 5.1: Historical Conservation Acquisition (system)

Efficiency programs with economic paybacks of less than one year are not eligible for
incentives, although Avista assists in educating and informing customers about these
types of efficiency measures. Site-specific programs require customized services for
commercial and industrial customers because of the unique characteristics of each of
their premises and processes. In some cases, Avista uses a prescriptive approach where
similar applications of energy efficiency measures result in reasonably consistent savings
estimates in conjunction with a high achievable savings potential. An example is
prescriptive lighting for commercial and industrial applications.

The Conservation Potential Assessment
Avista retained Applied Energy Group (AEG) as an independent third party to assist in
developing a Conservation Potential Assessment (CPA) for this IRP. The study forms the
basis for the conservation portion of this plan. The CPA identifies the 20-year potential
for energy efficiency and provides data on resources specific to Avista’s service territory
for use in the resource selection process in the PRiSM model, in accordance with the
EIA’s energy efficiency goals. The energy efficiency potential considers the impacts of
existing programs, the influence of known building codes and standards, technology
developments and innovations, changes to the economic influences, and energy prices.

AEG implemented several changes to its current study including a regionally specific
categorization of savings potential. In the 2015 IRP, AEG provided three levels of
aMW
20 aMW
40 aMW
60 aMW
80 aMW
100 aMW
120 aMW
140 aMW
160 aMW
180 aMW
200 aMW
220 aMW
aMW
2 aMW
4 aMW
6 aMW
8 aMW
10 aMW
12 aMW
14 aMW
16 aMW
18 aMW
20 aMW
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
20
0
4
20
0
6
20
0
8
20
1
0
20
1
2
20
1
4
20
1
6
Cu
m
u
l
a
t
i
v
e

S
a
v
i
n
g
s
An
n
u
a
l

S
a
v
i
n
g
s
SC_PR_3-2 Attachment A Page 65 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
potential: technical, economic, and achievable. This approach first considered the
economic screening of measures in the CPA then applied ramp rates in order to arrive at
the achievable potential. For the 2015 plan, Avista compared using this methodology
versus its new methodology utilizing a technical and achievable technical approach and
using PRiSM to select measures. Both methodologies arrived at similar results in the
2015 study, but the inclusion in the PRiSM model allows conservation to dynamically
reduce portfolio risk. In the 2015 IRP Washington acknowledgement, Washington agreed
Avista should make the methodology change. In the new method, AEG first develops
estimates of technical potential, reflecting the adoption of all conservation measures,
regardless of cost-effectiveness. Achievable Technical Potential modifies the technical
potential by accounting for customer adoption constraints, using the Council’s Seventh
Plan ramp rates. The estimated achievable technical potential for each individual
measure, along with associated costs, feed into the PRiSM model to select the cost-
effective measures on a measure-by-measure basis rather than by bundling. AEG took
the following steps to assess and analyze energy efficiency and potential within Avista’s
service territory. Figure 5.2 illustrates the steps of the analysis.

Figure 5.2: Analysis Approach Overview

1. Characterize the Market: Categorizes energy consumption in the residential
(including low-income customers), commercial, and industrial sectors. This
assessment uses utility and secondary data to characterize customers’ electricity
usage behavior in Avista’s service territory. AEG uses this assessment to develop
energy market profiles describing energy consumption by market segment, vintage
(existing or new construction), end use, and technology.
SC_PR_3-2 Attachment A Page 66 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
2. Baseline Projection: Develops a projection of energy and demand for electricity,
absent the effects of future conservation by sector and by end use for the entire 20-
year study.
3. Measure Assessment: Identifies and characterizes energy efficiency measures
appropriate for Avista, including regional savings from energy efficiency measures
acquired through Northwest Energy Efficiency Alliance efforts.
4. Potential: Analyzes measures to identify technical and achievable technical
conservation potential.

Market Segmentation
The CPA divides Avista customers by state and class. The residential class segments
include single-family, multi-family, manufactured home, and low-income customers.1 AEG
incorporated information from the Commercial Building Stock Assessment to break out
the commercial sector by building type. Avista analyzed the industrial sector as a whole
for each state. AEG characterized energy use by end use within each segment in each
sector, including space heating, cooling, lighting, water heat or motors; and by
technology, including heat pump and resistance-electric space heating.

The baseline projection is the “business as usual” metric without future utility conservation
programs. It estimates annual electricity consumption and peak demand by customer
segment and end use absent future efficiency programs. The baseline projection includes
the impacts of known building codes and energy efficiency standards as of 2016 when
the study began. Codes and standards have direct bearing on the amount of energy
efficiency potential existing beyond the impact of these efforts. The baseline projection
accounts for market changes including:

 customer and market growth;
 income growth;
 retail rates forecasts;
 trends in end use and technology saturations;
 equipment purchase decisions;
 consumer price elasticity;
 income; and
 persons per household.

For each customer class, AEG compiled a list of electrical energy efficiency measures
and equipment, drawing from the NPCC’s Seventh Power Plan, the Regional Technical
Forum, and other measures applicable to Avista. The 3,400 individual measures included
in the CPA represent a wide variety of end use applications, as well as devices and
actions able to reduce customer energy consumption. The AEG study includes measure
costs, energy and capacity savings, and estimated useful life.

1 The low-income threshold for this study is 200 percent of the federal poverty level. Low-income information
is available from census data and the American Community Survey data.
SC_PR_3-2 Attachment A Page 67 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Avista, through its PRiSM model, considers other performance factors for the list of
measures and performs an economic screening on each measure for every year of the
study to develop the economic potential of Avista’s service territory. Many measures
initially do not pass the economic screen of supply side resource options, but some
measures may become part of the energy efficiency program as contributing factors
evolve during the 20-year planning horizon.

Avista supplements energy efficiency activities by including potentials for distribution
efficiency measures consistent with EIA conservation targets and the NPCC Seventh
Power Plan. Details about the distribution efficiency projects are in Chapter 8 –
Transmission and Distribution Planning.

Overview of Energy Efficiency Potential
AEG’s approach adhered to the conventions outlined in the National Action Plan for
Energy Efficiency Guide for Conducting Potential Studies.2 The guide represents the most
credible and comprehensive national industry standard practice for specifying energy
efficiency potential. Specifically, two types of potential are in this study, as discussed
below. Table 5.1 shows the CPA results for technical and achievable technical potential.

Table 5.1: Cumulative Potential Savings (Across All Sectors for Selected Years)

2018 2019 2022 2027 2037
Cumulative (GWh)
Achievable Technical Potential 88.0 186.8 468.3 927.1 1,516.3
Technical Potential 190.1 376.7 771.7 1,370.9 1,937.0

Cumulative (aMW)
Achievable Technical Potential 10.0 21.3 53.5 105.8 173.1
Technical Potential 21.7 43.0 88.1 156.5 221.1

Technical Potential
Technical potential finds the most energy-efficient option commercially available for
each purchase decision regardless of its cost. This theoretical case provides the
broadest and highest definition of savings potential because it quantifies savings if all
current equipment, processes, and practices in all market sectors were replaced by
the most efficient and feasible technology. Technical potential in the CPA is a “phased-
in technical potential,” meaning only the current equipment stock at the end of its
useful life is considered and changed out with the most efficient measures available.
Non-equipment measures, such as controls and other devices (e.g., programmable
thermostats) phase-in over time, just like the equipment measures.

Achievable Technical Potential
Achievable Technical Potential is a subset of technical potential representing the
portion of technically feasible reductions in load associated with applicable end-uses.

2 National Action Plan for Energy Efficiency (2007). National Action Plan for Energy Efficiency Vision for
2025: Developing a Framework for Change. www.epa.gov/eeactionplan.
SC_PR_3-2 Attachment A Page 68 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
It refines technical potential by applying customer participation rates to account for
market barriers, customer awareness and attitudes, program maturity, and other
factors that may affect market penetration of efficiency measures. The customer
participation rates use the NPCC Seventh Plan ramp rates.

PRiSM Co-Optimization
Avista’s identifies achievable economic conservation potential by concurrently evaluating
supply side and over 8,700 demand side resources in PRiSM. This methodology was the
result of a 2013 IRP Action Item to streamline the process of selecting conservation in
conjunction with the Efficient Frontier. The 2015 IRP tested this method by comparing the
traditional methodology with the co-optimization. The co-optimization resulted in similar
savings, and portfolios further down the Efficient Frontier selected additional energy
efficiency to reduce risk at a higher cost. The Washington 2015 IRP acknowledgement
asked Avista to make this change for the 2017 IRP. Now in PRiSM, the individual energy
efficiency resources compete with supply- and demand response options to meet
resource deficits, although, energy efficiency measures benefit by receiving 10 percent
more value compared to the supply-side resources. This methodology does not change
the amount of conservation selected in the PRS, but provides information regarding
conservation selection if Avista choses different portfolios in the Efficient Frontier analysis
or other scenario analysis. Each program’s winter and summer peak contribution
(including line loss benefit), plus the value of its energy savings are considered. Figure
5.3 shows the combined Washington and Idaho CPA for 2018 through 2037.3

Figure 5.3: Achievable Conservation Potential Assessment (20-Year Cumulative)

3 The achievable conservation does not include savings from T&D losses. Chapter 11 conservation totals
include losses.
1
7
15
21
28
4
21
42
57
74
-
10
20
30
40
50
60
70
80
2018 2022 2027 2031 2037
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Idaho Washington
SC_PR_3-2 Attachment A Page 69 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Conservation Targets
The IRP process provides conservation targets for the Washington EIA Biennial
Conservation Plan. Other components, including conservation from distribution and
transmission efficiency improvements, combine with energy efficiency targets to arrive at
the full Biennial Conservation Plan target for Washington. Pursuant to requirements in
Washington, the biennial conservation target must be no lower than a pro rata share of
the utility’s ten-year conservation potential. In setting the Company’s target, both the two-
year achievable potential and the ten-year pro rata savings are determined with the higher
value used to inform the EIA Biennial target.

Figure 5.4: Washington Annual Achievable Potential Energy Efficiency (Megawatt Hours)

For the 2018-2019 CPA, the two-year achievable potential is 69,899 MWh for Washington
Electric operations. The pro rata share of the utility’s ten-year conservation potential of
73,636 MWh is the basis for calculating the biennial target. Table 5.2 contains achievable
conservation potential for 2018-2019 using the PRiSM methodology.

In addition to traditional efficiency programs, Avista is replacing approximately 21,640
high-pressure sodium fixtures in Washington and Idaho with comparable LED fixtures.
The expected completion of this project is late 2019; efficiency savings are not available
at this time to include in the achievable target. Also included is the energy savings from
feeder upgrade projects. These projects, described in Chapter 8 – Transmission and
Distribution Planning, reduce system losses.

2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
PRiSM 32,394 69,899 103,108 141,532 181,537 219,967 256,765 296,068 332,472 368,181
10-yr Prorata 36,818 73,636 110,454 147,272 184,091 220,909 257,727 294,545 331,363 368,181
32,394
69,899
368,181
-
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000
Ac
h
i
e
v
a
b
l
e

C
o
n
s
e
r
v
a
t
i
o
n
SC_PR_3-2 Attachment A Page 70 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Table 5.2: Annual Achievable Potential Energy Efficiency (Megawatt Hours)

2018-2019 Biennial Conservation Target Savings
(MWh)
Pro Rata Share of CPA 73,636
Behavioral Program 15,386
Less: NEEA (21,812)
End-Use Efficiency Measures Subtotal 67,210
Plus: Distribution Efficiency 714
Plus: Generation Efficiency 151
Total 68,075
Plus: Decoupling Commitment 3,404
Proposed Biennial Conservation Target + Decoupling (EIA)
(Subject to Penalties) 71,479
Plus: NEEA Projection 21,812
Total Conservation Commitment 93,291

Table 5.3: Annual Achievable Potential Energy Efficiency (Megawatt Hours)

2018 Feeder Upgrades 233 TBD 233
2019 Feeder Upgrades 481 472 953

2018 LED Street Lighting TBD TBD TBD
2019 LED Street Lighting TBD TBD TBD

2018 Facility Efficiencies 0 300 300
2019 Facility Efficiencies 151 0 151

For conservation efforts in Idaho, the Idaho Public Utilities Commission asked Avista to
pursue cost effective measures and set conservation goals based on the Utility Cost Test
(UCT). While the conservation identified in this IRP uses the Total Resource Cost (TRC)
in terms of power planning over twenty years, the amount of conservation the Company
will pursue in Idaho beginning in 2018 will use the UCT.

Using the UCT as the basis for conservation, Avista identifies achievable potential
conservation in Idaho of 15,370 MWh in 2018. The company determined this savings
amount by applying an adjustment factor of 1.28 to Avista’s TRC goal of 12,008 MWh.
The 1.28 adjustment factor is the ratio of the TRC to the UCT from the Company’s 2016
Idaho DSM Annual Report. In this report, Avista obtained a TRC of 2.13 and a UCT of
2.73 with the UCT being 1.28 times higher than the TRC.

NPCC’s Seventh Power Plan Benchmarking
Figure 5.5 illustrates the comparison between Avista’s CPA Achievable Conservation and
its estimated allocation of the Seventh Power Plan’s regional savings. Commercial and
Industrial sectors have been combined into a single category titled “non-residential.”
SC_PR_3-2 Attachment A Page 71 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
It is important to note that the value for from the Seventh Power Plan represents a single
point within a range of values. The comparison relies on the assumption that Avista’s
share of the region is 3.5 percent (Sixth Power Plan assumption). A 0.5 percent variance
in this allocation would increase or decrease Avista’s allocation of the Seventh Power
Plan by approximately 12 aMW.

Comparing Avista’s CPA to the Seventh Power Plan
The Washington 2015 IRP acknowledgement asked Avista to compare its IRP
conservation and demand response (DR) results to the Seventh Power Plan. Avista’s
Washington Electric CPA identifies 42 aMW of savings for the 2018-2027 period with 13
aMW from Residential and 29 aMW from Non-Residential saving. Avista’s allocation of
the Seventh Power Plan’s regional savings is approximately 61 aMW, with 24 aMW from
Residential and 37 aMW from Non-Residential. See Figure 5.5.

The comparison of Avista’s CPA and its share of the Seventh Power Plan considers
several factors. Avista’s avoided cost is lower than the costs used to calculate average
regional energy costs. Because avoided cost is a primary factor in determining cost-
effectiveness, some regional portfolio measures are not cost effective in Avista’s CPA.

Avista calculated the 61 aMW using the highest Levelized Cost Bins for Conservation4.
While information that is more granular is available, complications exist depending on end
use customers and the type of individual measures considered. For consistency, the
comparison uses the highest Cost Bin in calculating Avista’s share of the Seventh Power
Plan. This approach provides the most conservative estimates on cost.

Figure 5.5: 2017 Avista CPA / Seventh Power Plan Benchmark Comparison

4 Seventh Power Plan Appendix G, Table G-7: Levelized Cost Bins for Conservation.
13
29
42
24
37
61
0
10
20
30
40
50
60
70
Residential Non-Residential Total
Av
e
r
a
g
e

M
a
g
a
w
a
t
t
s
Avista CPA 7th Power Plan
19-28
30-44
49-72
SC_PR_3-2 Attachment A Page 72 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Consistency with the Seventh Power Plan
AEG’s methodology to develop the electric CPA is consistent with the Council’s Seventh
Power Plan methodology and fulfills the requirements of the utility analysis option as
specified in WAC 194-37-070 subsection (6),(a)(i) through (xv).5 This CPA, like the
Seventh Plan, uses an end-use model to distinctly consider and account for the following:

 Building characteristics that reflect Avista’s service territory;
 Fuel and equipment saturations based on the knowledge of Avista’s customers;
 Measure life;
 Stock accounting;
 Existing and new construction;
 Lost-and non-lost opportunities;
 Measure saturation and applicability;
 Measure savings, including contribution to system peak;
 Customer growth; and
 Federal and state standards for appliances and technologies.

Like the Seventh Plan, the Avista CPA uses a frozen-efficiency approach assuming
equipment efficiency purchase decisions are fixed, with the exception of changes due to
the phase-in of new codes and standards.

For this CPA, AEG develops estimates of Technical Potential and Achievable Technical
Potential.6 The Economic Achievable Potential was determined by running the Achievable
Technical Potential through PRiSM. The Power Act’s 10 percent adder for conservation
is added to the avoided energy costs within the PRiSM model.

In terms of conservation measures, the CPA includes all measures incorporated in the
Seventh Plan, as well as additional measures. However, the CPA analyzes each measure
individually, whereas the Seventh Plan bundles measures in some cases. All measures
were characterized using data from the Seventh Plan and RTF workbooks, when
available. If a measure was not characterized using the Seventh Plan or RTF workbooks,
AEG relied upon its database of energy efficiency measures (DEEM) that is developed
by incorporating measures encountered throughout the country and characterized using
sources typically cited by the NPCC in its analyses. Similar to the Council’s approach,
AEG removes measures with market saturation, such as LED TVs, while at the same time
includes and updates commercially available technologies.

To develop Technical Potential, AEG’s LoadMAP model includes all technically feasible
potential conservation. The model choses the most efficient option at the time of
equipment turnover. The market acceptance rates used to develop Achievable Technical
potential are based upon the new, simplified Seventh Plan ramp rates. AEG mapped each
of the individual measures to a Seventh Plan ramp rate and compared the results to
historical achievements. AEG then adjusted the 2018 achievable technical potential for

5 http://apps.leg.wa.gov/WAC/default.aspx?cite=194-37&full=true
6 AEG provided estimates of Technical Potential, Economic Potential, and Achievable Potential in previous
CPAs. For this study, the ramp rates were applied to the Technical Potential and provided to Avista to run
through PRiSM to estimate the cost-effective conservation potential.
SC_PR_3-2 Attachment A Page 73 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
those specific measures to line up with 2018. This provided a starting point for 2018
potential aligned to historic results. AEG provided the individual measure characteristics
at the Achievable Technical level to Avista to run through PRiSM to determine which
measures are cost-effective and included in the Economic Achievable Potential or targets.

Energy Efficiency-Related Financial Impacts
The EIA requires utilities with over 25,000 customers to obtain a fixed percentage of their
electricity from qualifying renewable resources and acquire all cost-effective and
achievable energy conservation.7 For the first 24-month period under the law, 2010-2011,
this equaled a ramped-in share of the regional 10-year conservation target identified in
the Seventh Power Plan. Penalties of at least $50 per MWh exist for utilities not achieving
Washington EIA targets.

The EIA requirement to acquire all cost-effective and achievable conservation may pose
significant financial implications for Washington customers. Based on CPA results, the
projected 2018 conservation acquisition cost to electric customers is $14.5 million. This
amount grows by 200% to $29 million by 2027, a total of $214 million over this 10-year
period. Costs continue increasing after 2027 to more than $40 million in 2037.

Integrating Results into Business Planning and Operations
The CPA and IRP energy efficiency evaluation processes provide high-level estimates of
conservation cost-effectiveness and acquisition opportunities. Results establish baseline
goals for continued development and enhancement of energy efficiency programs, but
the results are not detailed enough to form an actionable plan. Avista uses both
processes’ results to establish a budget for energy efficiency measures, help determine
the size and skill sets necessary for future operations, and identify general target markets
for energy efficiency programs. This section provides an overview of recent operations of
the individual sectors, as well as energy efficiency business planning.

The CPA is useful for implementing energy efficiency programs in the following ways:

 Identifying conservation resource potentials by sector, segment, end use, and
measure of where energy savings may come from. Energy efficiency staff uses
CPA results to determine the segments and end uses/measures to target.
 Identifying measures with the highest TRC benefit-cost ratios, resulting in the
lowest cost resources, brings the greatest amount of benefits to the overall
portfolio.
 By identifying measures with great adoption barriers based on the economic
versus achievable results by measure, staff can develop effective programs for
measures with slow adoption or significant barriers.
 By improving the design of current program offerings, staff can review the measure
level results by sector and compare the savings with the largest-saving measures
currently offered. This analysis may lead to the addition or elimination of programs.

7 The EIA defines cost effective as 10 percent higher than the cost a utility would otherwise spend on energy
acquisition.
SC_PR_3-2 Attachment A Page 74 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Additional consideration for lost opportunities can lead to offering greater
incentives on measures with higher benefits and lower incentives on measures
with lower benefits.

The CPA illustrates potential markets and provides a list of cost-effective measures to
analyze through the on-going energy efficiency business planning process. This review
of both residential and non-residential program concepts, and their sensitivity to more
detailed assumptions, feeds into program planning.

Residential Sector Overview
The Company’s residential portfolio is composed of several approaches to engage and
encourage customers to consider energy efficiency improvements within their home.
Prescriptive rebate programs are the main component of the portfolio, but augment
variety of other interventions. These include: upstream buy-down of low-cost lighting and
water saving measures, select distribution of low-cost lighting and weatherization
materials, direct-install programs and a multi-faceted, multichannel outreach and
customer engagement effort.

Washington and Idaho residential customers received over $10.2 million in rebates to
offset the cost of implementing these energy efficiency measures. All programs within the
residential portfolio contributed over 83,400 MWh and over 669,800 therms to the 2016
annual energy savings.

Avista launched a Home Energy Reports program in June 2013, targeting 73,501 Idaho
and Washington customers with high electric use. As of December 2015, Avista had
48,800 customers still in the Home Energy Reports program. In January of 2016, Avista
‘refilled’ their existing Home Energy Reports Program by 24,706 customers bringing total
distribution to approximately 73,506 electric customers in Idaho and Washington.
Eligibility for treatment includes several criteria such as sufficient (two year) billing history,
enough peers to build comparison group, not in the control group, not a ‘do not solicit’
customer and high enough electric use to be cost-effectively treated. In an effort to reduce
energy usage through behavioral changes, Home Energy Reports show personalized
usage insights and energy saving tips. Customers also see a ranking of similar homes,
comparison to themselves and a personal savings goal on the Reports. In addition to
closely matching usage curves, the similar home comparisons use the following four
criteria: square footage, home type, heat type and proximity. The Oracle Home Energy
Report contributed 12,131 MWh of savings in 2016.

Low-Income Sector Overview
The Company leverages the infrastructure of six network Community Action Program
(CAP) agencies and one tribal weatherization organization to deliver energy efficiency
programs for the Company’s low-income residential customers in the Washington service
territory. CAP agencies have resources to income qualify, prioritize and treat client homes
based upon a number of characteristics. In addition to the Company’s annual funding, the
agencies have other monetary resources to leverage when treating a home with
weatherization or other energy efficiency measures. The agencies either have in‐house
or contract crews to install many of the efficiency measures of the program. The low-
SC_PR_3-2 Attachment A Page 75 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
income energy efficiency programs contributed 830 MWh of electricity savings and
19,183 therms of natural gas savings in 2016 to Avista’s system.

The general outreach programs provide energy management information and resources
at events (such as resource fairs) and through partnerships to reach target populations.
These programs also include bill payment options and assistance resources in senior and
low-income publications. In 2016, Avista participated in 193 events in Idaho and
Washington including workshops, energy fairs, mobile outreach events, and general
outreach partnerships and events reaching over 16,500 individuals.

Non-Residential Sector Overview
The non-residential energy efficiency market delivers through a combination of
prescriptive and site-specific offerings. Any measure not offered through a prescriptive
program is automatically eligible for treatment through the site-specific program, subject
to the criteria for program participation. Prescriptive paths for the non-residential market
are preferred for small and uniform measures.

In 2016, more than 2,900 prescriptive and site-specific nonresidential projects received
funding. Additionally, the Small Business program installed over 27,500 measures. Avista
contributed more than $14.8 million for energy efficiency upgrades in nonresidential and
small business applications. Non-residential programs realized over 73,900 MWh and
196,875 therms in annual first‐year energy savings.

Program changes made at the beginning of 2016 to the non-residential programs include
the addition of new program offerings and changes to eligibility or incentive levels. Avista
communicates the majority of program changes after the Business Plan is final and the
changes become effective at the beginning of the year. In addition, some program’s
change throughout the year as necessary but these are less typical.

For non-residential programs, changes effective January 1, 2016 to rebates reflect new
information regarding new unit energy savings (UES) and cost values. Avista accepted
rebate applications through March 31, 2016 for 2015 measures and amounts. This 90-
day grace period allows for a smooth transition when rebate programs change to allow
enough time for customers in the pipeline to complete their projects yet close out changes
in a timely but balanced approach.

After years of review, Avista began converting a large portion of its high-pressure sodium
(HPS) street light system to LED units in 2015. Advancements in LED technology and
lower product costs make early replacements cost effective. LEDs consume about half of
the energy as their conventional counterparts for the same light output. Other non-energy
benefits include improved visibility and color rendering relative to HPS lighting, and longer
product life. The initial focus of the program is replacing 26,000 100-watt cobra-head style
streetlights.

Conservation’s T&D Deferral Analysis
Cost-effective energy efficiency programs require a review of cost versus its potential
benefits. One benefit is the generation and delivery system investments avoided or
SC_PR_3-2 Attachment A Page 76 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
deferred. Generation avoided investments are fairly straightforward, but avoided
transmission and distribution (T&D) system components tend to be less straightforward
as the investments are lumpy, location specific, and may or may not be reduced by energy
efficiency due to the thermal limitations of the system.

Utilities use a number of methods to estimate avoided T&D costs and there is no one
“best” approach to developing these estimates. There is a wide range of estimates for
avoided T&D, underscoring the diverse nature of the methods used to calculate avoided
costs. For the past several IRPs, Avista used $10 per kW-yr (2006 dollars), based on a
study for the 2007 IRP, this out of date study is driving the need for a new methodology
as part of the 2015 IRP action plan.

For this IRP, Avista chose to value these benefits using the current values approach. The
current values approach considers the amount of current investment in both T&D from a
revenue requirement reference point, then divided by the peak load of the system, to
estimate a $/kW-yr. value (see Table 5.2). This method’s strength is its simplicity, lending
itself to frequent updates, but does not accurately portray the amount of deferred future
T&D investment due to new conservation programs. Avista will consider moving to
another methodology to account for this benefit in the next IRP. Further, in Chapter 8,
there is a discussion of a storage facility’s benefit to the distribution system by deferring
new capital investment using three feeders as case studies. Given, T&D deferments
importance, Avista will evaluate alternative methods to value these benefits to future
investment.

Table 5.4: Transmission and Distribution Benefit

Transmission Net
Book Value
Distribution Net
Book Vale
Washington $294,988,593 $675,072,411
Idaho $153,799,772 $348,486,297
Total $448,788,365 $1,023,558,708
Revenue Requirement $448,859,497 $1,099,186,748

Peak Load (MW) 1,693 1,693
Current $/kW $265 $649
Levelized Cost $13.77 $15.95
Total Levelized cost $29.72

Generation Efficiency Audits of Avista Facilities
Avista engineers performed energy efficiency audits at all of Avista’s hydroelectric dams
and most of thermal generation facilities where Avista wholly owns or is a partial owner,
excluding Colstrip Generating station in Colstrip, Montana. The scoping audits focused
on lighting, shell, HVAC and motor controls on processes. Table 5.5 shows efficiency
potential and Table 5.6 shows the efficiency projects for Avista generation facilities
planned for 2017 and 2018.

SC_PR_3-2 Attachment A Page 77 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Table 5.5: Preliminary Generation Facility Efficiency Upgrade Potential

Facility Description Measure
Life
(years)
Electric
Savings
(kWh)
Boulder Park

Control Room Lighting 15 3,931
Generating Floor Lighting High Bays 15 16,099
Replacing Engine Bay Lights 15 6,736
Replace Exterior Wall Packs 15 16,054
Instrument Air Cycling Air-Dryers 12 10,074
Oil Reservoir Heater Fuel Conversion8 15 525,600
Coyote Springs

Control Room Lighting 15 6,368
Generating Floor Lighting High Bays 15 85,778
Roadway Lighting 15 1,085
Air-Compressor VFD 12 130,000
Retrofit Air-Dryer with Dew-Point Controls 12 25,000
Kettle Falls

Plant Lighting 15 150,190
Plant Lighting Controls 15 183,058
Yard Lighting 15 48,180
Forced Draft Boiler Fan VSD 12 700,000
Little Falls Speed Controls Cooling/Exhaust Fans 12 247,909
Long Lake Variable Speed Stator Cooling Blowers 12 135,000
Exterior Wall Packs 15 2,084
Northeast CT Halogen Pole Lights 15 5,146
Noxon Rapids Full LED Lighting Upgrade (Completed) 15 382,115
Post Falls

Control Room T12s 15 1,776
Generating Floor HPS 15 3,312
Upper Falls

Utility Men Break Room Lighting 15 2,151
Control Room Lighting 15 4,340
Network Feeder Tunnel Lighting 15 8,344
Rathdrum CT

Roadway Lighting 15 16,273
Halogen Pole Lights 15 3,200

Lighting Projects
The facilities have a mixture of T12, T8 and some T5 linear fluorescent fixtures as well as
many incandescent bulbs. The proposed replacement fixtures from the lighting audits are
primarily linear, high bay, and screw in LED fixtures. Noxon Rapids is the only facility with
a completed a lighting retrofit.

Shell Projects
No shell measures are cost effective due to negligible savings and cost prohibitive nature
of the measure due to the size of the facilities and large internal heat gain of the
equipment in the facilities. However, small maintenance weatherization are available to
improve occupant comfort.

8 Also saves 23,911 therms of natural gas per year.
SC_PR_3-2 Attachment A Page 78 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
HVAC Projects
There are no recommendations to replace current HVAC equipment but there are
recommendations to replace equipment with more efficient technology when each unit
reaches the end of its’ useful life.

Controls on Process Motors
There are a number of air compressors, fans and pumps driven by electric motors in
Avista’s facilities. These motors could use variable speed drives to match the current
process needs and reduce the energy consumption of the motors as opposed to the
current control systems.

Table 5.6: Planned Generation Facility Efficiency Upgrades 2017 – 2018

Facility Description Measure
Life
(years)
Electric
Savings
(kWh)
Cabinet Gorge Lighting Retrofit 15 300,000
Little Falls Lighting Retrofit 15 62,266
Long Lake Lighting Retrofit 15 17,441
Nine Mile Lighting Retrofit 15 71,455

Demand Response
Over the past decade, demand response or DR gained attention as an alternative to new
generation to meet peak load growth. DR reduces load to specific customers during peak
demand periods until the load event is over or the customer has met its commitment.
Enrolling customers allows the utility to modify customer usage in exchange for bill
discounts. National attention focuses on residential programs to control water heaters,
space heating, and air conditioners. A 2013 IRP Action Item suggested further study of
the DR potential based on its selection as a PRS resource from 2022 to 2027. Avista
retained AEG to study the potential of future commercial and industrial programs for both
the 2015 IRP and 2017 IRPs.

Previous Demand Response Programs
Avista’s first DR experience began in the 2001 Energy Crisis. Avista responded with an
all-customer and irrigation customer buy-back programs and bi-lateral agreements with
its largest industrial customers. These programs, along with enhanced commercial and
residential energy efficiency programs, reduced the need for purchases in very high-cost
wholesale electricity markets. A July 2006 multi-day heat wave again led Avista to request
DR through a media request for customers to conserve and short-term agreements with
large industrial customers. During the 2006 event, Avista estimates DR reduced loads by
50 MW.

Avista conducted a two-year residential load control pilot between 2007 and 2009 to study
specific technologies and examine cost-effectiveness and customer acceptance. The
pilot tested scalable Direct Load Control (DLC) devices based on installation in
approximately 100 volunteer households in Sandpoint and Moscow, Idaho. The sample
allowed Avista to test DR with the benefits of a larger-scale project, but in a controlled
SC_PR_3-2 Attachment A Page 79 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
and customer-friendly manner. Avista installed DLC devices on heat pumps, water
heaters, electric forced-air furnaces, and air conditioners to control operation during 10
scheduled events at peak times ranging from two to four hours. A separate group within
the same communities participated in an in-home-display device study as part of the pilot.
The program provided Avista and its customers experience with “near-real time” energy-
usage feedback equipment. Information gained from the pilot is in the report filed with the
Idaho Public Utilities Commission.

Avista engaged in a DR program as part of the Northwest Regional Smart Grid
Demonstration Project (SGDP) with Washington State University (WSU) and
approximately 70 residential customers in Pullman and Albion, Washington. Residential
customer assets including forced-air electric furnaces, heat pumps, and central air-
conditioning units received a Smart Communicating Thermostat provided and installed by
Avista. The control approach was non-traditional in several ways. First, the DR events
were not prescheduled, but Avista controlled customer loads defined by pre-defined
customer preferences (no more than a two degree offset for residential customers and an
energy management system at WSU with a console operator). More importantly, the
technology used in the DR portion of the SGDP predicted if equipment was available for
participation in the control event. Lastly, value quantification extended beyond demand
and energy savings and explored bill management options for customers with whole
house usage data analyzed in conjunction with smart thermostat data.

Inefficient homes identified through this analysis prompted customer engagement. For
example, an operational anomaly prompted an investigation uncovering a control board
in a customer’s heat pump causing the system to draw warm air from inside the home
during the heating season. This in turn caused the auxiliary heat to come on prematurely
and cycle too frequently, resulting in high customer bills. The repair saved the customer
money and allowed them to be more comfortable in their home. Lessons learned from the
SGDP program helped craft Avista’s new Smart Thermostat rebate program (an
efficiency-only program) implemented in October 2014. The Smart Grid demonstration
project concluded December 31, 2014.

Experiences from both residential DLC pilots (North Idaho Pilot and the SGDP) show
participating customer engagement is high; however, recruiting participants is
challenging. Avista’s service territory has high natural gas penetration for typical DLC
space and water heat applications. Customers who have interest may not have qualifying
equipment, making them ineligible for participation in the program. Secondly, customers
did not seem overly interested in the DLC program offerings. BPA has found similar
challenges in gaining customer interest in their recent regional DLC programs. Finally,
Avista is unable at this time to offer pricing strategies other than direct incentives to
compensate customers for participation in the program.

Avista is committed to evaluating and considering DR to meet future load requirements if
it cost effective compared to other alternatives and does not influence the customer’s
reliability or satisfaction with service. To fulfill this commitment, Avista will determine if a
study is needed to evaluate the residential DR potential for the next IRP to meet its winter
and summer peak requirements as part of this IRP’s action plan.

SC_PR_3-2 Attachment A Page 80 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Demand Response Comparison to the Seventh Power Plan
For DR, Avista reviewed the NPCC’s Seventh Plan and found some differences between
Avista’s DR analysis and the NPCC’s including 1) the NPCC’s analysis includes
residential and agricultural programs, 2) specific summer and winter programs, and 3) the
NPCC excludes standby generator programs. Further, the NPCC models these programs
in bins, rather than specific programs. Avista will determine if it is necessary to include
residential DR programs in the 2019 IRP, but agricultural programs will be limited due to
Avista’s limited irrigation pumping load, although other agricultural process were included
in the industrial portion of the existing study. Avista only includes winter C&I programs in
its study, as at the time of the analysis Avista’s capacity requirements are winter peaking
rather than summer peaking.

The NPCC estimates 600 MW9 of DR for the region; using Avista’s 3.5 percent share of
the region10, equates to 21 MW of DR. Avista’s PRS, as described in Chapter 11, includes
9 MW of winter C&I DR and 35 MW of standby generation, for 44 MW11 of total peak load
reduction. This more than doubles the amount of DR the NPCC includes as cost effective
in the Seventh Power Plan.

Demand Response Potential Assessment Study
Avista retained AEG to study the potential for commercial and industrial DR in Avista’s
service territory for the 20-year planning horizon of 2018–2037. It primarily sought to
develop reliable estimates of the magnitude, timing, and costs of DR resources likely
available to Avista for meeting winter peak loads. The study focuses on resources
assumed achievable during the planning horizon, recognizing known market dynamics
may hinder acquisition. Avista includes in the DR analysis savings from avoiding T&D
losses, but does not include T&D capital deferral benefits as it is not determined whether
or not a system peak DR program will actually defer any specific T&D investment.

The IRP incorporates DR study results, and the study will affect subsequent DR planning
and program development efforts. A full report outlining the DR potential for commercial
and industrial customers is in Appendix C from the 2015 IRP. AEG updated the costs and
savings for this IRP, but the report showing the amount of DR in Avista’s service territory
is the same. Table 5.3 details achievable DR potential for the programs studied by AEG.

Table 5.7: Commercial and Industrial Demand Response Achievable Potential (MW)

Program 2018 2019 2020 2037 2037
Direct Load Control 0.4 1.1 2.2 3.9 4.2
Firm Curtailment 5.8 11.6 17.5 17.7 18.2
Opt-in Critical Peak Pricing 0.1 0.4 0.9 4.4 4.6

9 NPCC’s Seventh Power Plan, page 3-4.
10 Avista’s estimate share of the region per the NPCC Sixth Power Plan, this calculation is not available for
the Seventh Power Plan at this time.
11 The 44 MW figure does not include additional savings from transmission and distribution loses.
SC_PR_3-2 Attachment A Page 81 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
Direct Load Control
A DLC program targeting Avista General and Large General Service customers in
Washington and Idaho would directly control electric space heating load in winter, and
water heating load throughout the year, through a load control switch or programmable
thermostat. Central electric furnaces, heat pumps, and water heaters would cycle on
and off during high-load events. Typically, DLC programs take five years to ramp up
to maximum participation levels.

Firm Curtailment
Customers participating in a firm curtailment program agree to reduce demand by a
specific amount or to a pre-specified consumption level during the event. In return,
they receive fixed incentive payments. Customers receive payments even if they never
receive a load curtailment request. The capacity payment typically varies with the firm
reliability-commitment level. In addition to fixed capacity payments, participants
receive compensation for reduced energy consumption. Because the program
includes a contractual agreement for a specific level of load reduction, enrolled loads
have the potential to replace a firm generation resource. Penalties are a possible
component of a firm curtailment program.

Industry experience indicates customers with loads greater than 200 kW participate in
firm curtailment programs. However, there are a few programs where customers with
100-kW maximum demand participate. In Avista’s case, the study lowered the demand
threshold level to include Large General Service customers with an average demand
of 100 kW or more.

Customers with operational flexibility are attractive candidates for firm curtailment
programs. Examples of customer segments with high participation possibilities include
large retail establishments, grocery chains, large offices, refrigerated warehouses,
water- and wastewater-treatment plants, and industries with process storage (e.g. pulp
and paper, cement manufacturing). Customers with operations requiring continuous
processes, or with obligations such as schools and hospitals, generally are not good
candidates.

Third parties generally administer firm curtailment programs for utilities and are
responsible for all aspects of program implementation, including program marketing
and outreach, customer recruitment, technology installation and incentive payments.
Avista could contract with a third party to deliver a fixed amount of capacity reduction
over a certain specified timeframe. The contracted capacity reduction and the actual
energy reduction during DR events is the basis of payment to the third party.

Critical Peak Pricing
Critical peak pricing programs set prices much higher during short critical peak periods
to encourage lower customer usage at those times. Critical peak pricing is usually offered
in conjunction with time-of-use rates, implying at least three periods: critical peak, on-
peak and off-peak. Utilities offer heavy discounts to participating customers during off-
peak periods, even relative to a standard time-of-use rate program. Event days generally
are a day ahead or even during the event day. Over time, establishment of event-trigger
criteria enables customers to anticipate events based on hot weather or other factors.
SC_PR_3-2 Attachment A Page 82 of 205
Chapter 5- Energy Efficiency & Demand Response
Avista Corp 2017 Electric IRP
System contingencies and emergencies are candidates for Critical peak pricing. Critical
peak pricing differentials between on-peak and off-peak in the AEG study are 6:1, and
available to all three commercial and industrial classes.

There are two ways to offer critical peak pricing. The opt-in rate allows voluntary
enrollment in the program or the utility enrolls all customers in an opt-out program,
requiring them to select another rate program if they do not want to participate. Avista is
only modeling the opt-in program. The success of the critical peak pricing program will
vary according to whether customers have enabling technology to automate their
response. For General and Large General Service customers, the enabling technology is
a programmable communicating thermostat. For Extra Large General Service customers,
the enabling technology is automated DR implemented through energy management and
control systems.

Critical peak pricing programs require formal rate design based on customer billing data
to specify peak and off-peak price levels and periods the rates are available. Rate design
was outside the scope of the AEG study. Further, new metering technology is required.
Given these requirements, critical peak pricing was not an option for the IRP.

Standby Generation Partnership
Few utilities have contracted with large industrial customers to use their standby
generation resources during peak load events or to provide non-spinning reserves. Avista
studied a standby generation option similar to the Portland General Electric program
where existing customers use their standby generation. Portland General Electric
dispatches, tests, and maintains the customer generation resources in exchange for
control of the resource in non-emergency situations. It uses customer generators for
limited hours for peak requirements, operating reserves, and potentially for voltage
support on certain distribution feeders.

Environmental regulations limit the use of backup generation facilities unless they meet
strict emission guidelines. To provide more operating hours a program could introduce
natural gas blending to improve the emissions and operating costs. Avista estimates
approximately 40 MW12 of standby generation resources are available for utility use over
20-year acquisition period. The IRP assumes a standby generation program would cost
$50 per kW in upfront investments, and $10 to $15 per kW-year in O&M costs.

In May 2015, the federal courts overturned Reciprocating Internal Combustion Engine
(RICE) rule limiting the availability of standby generation resources. The RICE rule was
remanded to EPA and remains in its 2013 form the former rule. Under clarification of this
rule, the EPA allows generators to dispatch 50 hours per year in non-emergency
conditions. Local air authorities may further restrict qualifying generators to new
technologies. In the event this program is part of Avista’s plans to meet resource deficits,
additional environmental and potential studies will begin.

12 The AEG DR study included standby generation in its firm curtailment section, in the event both programs
are cost effective, firm curtailment will include a 50 percent reduction in its capability.

SC_PR_3-2 Attachment A Page 83 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-1
6. Long-Term Position

Introduction & Highlights
This chapter describes the analytical framework used to develop Avista’s net resource
position. It describes reserve margins held to meet peak loads, risk-planning metrics
used to meet hydroelectric variability, and plans to meet renewable goals set by
Washington’s Energy Independence Act (EIA).

Avista has unique attributes affecting its ability to meet peak load requirements. It
connects to several neighboring utility systems, but is only 5 percent of the total regional
load. Annual peaks can occur either in the winter or in the summer; but Avista is winter
peaking on a planning basis using extreme weather conditions. The winter peak
generally occurs in December or January, but may happen in November or February
when extreme weather events may occur. As described in Chapter 4 – Existing Supply
Resources, Avista’s resource mix contains roughly equal amounts of hydroelectric and
thermal generation. Hydroelectric resources meet most of Avista’s flexibility
requirements for load and intermittent generation, though thermal generation is playing
a larger role as load growth and intermittent generation increase flexibility demands.

Reserve Margins
Planning reserves accommodate situations when load exceeds and/or resource output
falls below expectations due to adverse weather, forced outages, poor water conditions,
or other unplanned events. Reserve margins, on average, increase customer rates
when compared to resource portfolios without reserves because of the cost of carrying
rarely used generating capacity. Reserve resources have the physical capability to
generate electricity, but most have high operating costs that limit their dispatch and
revenue.

There is no industry standard reserve margin level; standardization across systems with
varying resource mixes, system sizes, and transmission interconnections, is difficult.
NERC defines reserve margins as follows:

Section Highlights
Avista’s first long

Avista’s peak hour planning margin is 14

SC_PR_3-2 Attachment A Page 84 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-2
Generally, the projected demand is based on a 50/50 forecast. Based on
experience, for Bulk Power Systems that are not energy-constrained, reserve
margin is the difference between available capacity and peak demand,
normalized by peak demand shown as a percentage to maintain reliable
operation while meeting unforeseen increases in demand (e.g. extreme weather)
and unexpected outages of existing capacity. Further, from a planning
perspective, planning reserve margin trends identify whether capacity additions
are keeping up with demand growth. As this is a capacity based metric, it does
not provide an accurate assessment of performance in energy limited systems,
e.g., hydro capacity with limited water resources. Data used here is the same
data that is submitted to NERC for seasonal and long-term reliability
assessments. Figures above shows forecast net capacity reserve margin in US
and Canada from 2008 to 2017.

NERC's Reference Reserve Margin is equivalent to the Target Reserve Margin
Level provided by the Regional/subregional’s own specific margin based on load,
generation, and transmission characteristics as well as regulatory requirements.
If not provided, NERC assigned 15 percent Reserve Margin for predominately
thermal systems and 10 percent for predominately hydro systems. As the
planning reserve margin is a capacity based metric, it does not provide an
accurate assessment of performance in energy limited systems, e.g., hydro
capacity with limited water resources.1

Avista and the region’s hydroelectric system is energy constrained, so the 10 or 15
percent metrics from NERC do not adequately define our planning margin. Beyond
planning margins, as defined by NERC, a utility must maintain operating reserves to
cover forced outages on the system. Avista includes operating reserves in addition to a
planning margin. Per Western Electric Coordinating Council (WECC) requirements,
Avista must maintain 1.5 percent of control area load and 1.5 percent of on-line control
area generation as spinning reserves.2 Then an additional 1.5 percent of control area
load and 1.5 percent of on-line control area generation as non-spinning reserves. Avista
must also maintain reserves to meet load following and regulation requirements of
within-hour load and generation variability, this amount equals 16 MW at the peak hour.
Recently, the WECC began experimenting with changing the reserve rules. The current
proposal is to keep three percent of load and three percent of generation as operating
reserves, but to remove the requirement to hold half the reserves as spinning reserve.
In lieu of spinning reserves is a requirement to hold 24 MW (for Avista) as Frequency
Response Reserves (FRR). FRR can instantaneously respond to changes in frequency.
Avista has sufficient FRR resource capability; but will require operational changes to
insure the units with this capability are available. Avista will not acquire additional
capacity until its expected peak loads plus reserve margins exceed resources beyond
2026 either on a single-hour or on a sustained three-day basis.

1 http://www.nerc.com/pa/RAPA/ri/Pages/PlanningReserveMargin.aspx.
2 Spinning reserves synchronize to the system while stand-by reserves must be available within 10
minutes.
SC_PR_3-2 Attachment A Page 85 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-3
Planning Margin
Utility capacity planning begins with identifying the broader regional capacity position,
as regional surpluses can offset high planning margins and utility investments. The
Northwest has a history of capacity surpluses and energy deficits because of its
hydroelectric generation base. Since the 2000-2001 energy crisis, the Northwest added
nearly 6,400 MW of natural gas-fired generation. During this same time, Oregon and
Washington added 7,890 MW of wind generation. Northwest load growth projections are
lower as compared to history, but with announced coal plant retirements and wind’s lack
of on-peak capacity contribution, the region is approaching load-resource capacity
balance, while retaining an energy surplus.

Given the interconnected landscape of the Northwest power market, selecting a
planning margin target is not straightforward. One approach is to conduct a regional
loss of load probability (LOLP) study calculating the amount of capacity required to meet
a five percent LOLP threshold. Five percent LOLP means a utility meets all customer
demand in all hours of the year in 19 of 20 years; this allows one loss-of-load event in a
20-year period. Regional LOLP analysis is beyond the scope of an IRP. Fortunately, the
NPCC conducts regional LOLP studies.

The NPCC analyzes northwest resource adequacy. Based on their work, the northwest
begins to fail the five percent LOLP measure in the winter of 2021-22 when major coal
generators retire.3 The NPCC identifies a loss of load probability after conservation is
7.2 percent, assuming the region can import 2,500 MW of power from southern
neighbors. The projected shortages occur primarily in the winter, but now the summer
as well, the same periods when Avista would expect its peak loads to occur. The
summer LOLP is new to the Council’s analysis prompting Avista to consider a summer
planning margin. In prior studies during the 2015 IRP cycle, the Council concluded the
region had enough capacity to meet summer demand. The recent change is due to
additional coal plant retirement announcements.

Avista is an interconnected utility, an advantage over its sister utility Alaska Electric
Light & Power (AELP). AELP is an electrical island and must meet all loads
instantaneously using its own resources without relying on its neighbors. AELP retains
large reserve margins to account for avalanche danger – typically 115 percent of peak
load. Avista, as an interconnected utility, can rely on its neighbors (and the neighbors
can rely on Avista) to lower planning margins. The harder question is how much
reliance it should place on the wholesale market. Wholesale markets are important to
help meet load when controlled resource dispatch is not available from factors such as
economic dispatch, forced or planned outages, low renewable energy production (such
as wind/hydro), or higher than normal loads. In the 2013 IRP, Avista found a 30 percent
planning margin (in addition to operating reserves) would be required to meet the 5
percent LOLP without connecting to the wholesale market. This higher planning margin
is due to Avista’s large resources as compared to its load. Since Avista is an
interconnected utility, a lower planning margin of 14 percent (winter) and seven percent
(summer) is included in the plan to balance the reliance on the marketplace when large

3https://www.nwcouncil.org/media/7491213/2017-5.pdf.
SC_PR_3-2 Attachment A Page 86 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-4
resources have forced outages or other combination of events. This difference results in
Avista requiring 270 MW less winter peak generation in 2018 than if Avista was an
electrical island, a similar amount to its largest contingencies. The total requirement for
planning margin and other reserves equates to a 22.6 percent planning margin.

Avista studied planning margins used by transmission organizations and utilities across
the country as part of the 2015 IRP. The results varied depending on the amount and
size of their interconnections and the resource mix within their systems. One challenge
in comparing planning margins across utilities is determining if they include ancillary
service, or operating reserve, obligations in their planning margins. Utilities with minimal
interconnections or a large hydroelectric system have higher planning margins than
better-interconnected and/or thermal-based systems. Avista and its neighbors generally
meet a large portion of their load obligations with hydroelectric resources, implying that
their planning margins might need to be higher than NERC’s 15 percent
recommendation.

Another consideration when selecting the appropriate planning margin is the utility’s
largest single contingency relative to peak load. Avista’s largest single unit contingency
is Coyote Springs 2. This plant is 18.8 percent of weather-adjusted peak load in 2018, a
high statistic relative to Western Interconnect peers. Some resource planners argue
planning margins should be no smaller than a utility’s single largest contingency on the
basis that if the largest resource fails, other resources may not be able to replace it.
Given the Northwest’s contingency reserve sharing agreement, lower reserve levels are
required for the first hour following a qualifying generation outage. Signatories to the
contingency reserve sharing agreement can call on assistance from neighboring utilities
for up to 60 minutes to help meet shortages. Beyond the first hour, utilities are
responsible for replacing the lost power themselves, either from other utility resources,
from purchases from other generators, or from load reductions.

In Avista’s prior LOLP studies, both summer and winter capacity shortages are possible
due to high peak loads. Past IRPs planned to utilize the wholesale market for summer
capacity due to the amount of available surplus market capacity. As this capacity
surplus shrinks, Avista is changing its summer planning margin to seven percent plus
operating reserves and regulation. Avista chose the seven percent planning margin by
comparing the standard deviation of potential loads in the summer (69 MW) to winter
peak load standard deviation (138 MW).4 Avista concluded the summer planning margin
should be half of the winter planning margin because the standard deviation of summer
potential peak loads is half of the winter peak loads. Avista will continue to analyze
planning margins using its loss of load model to validate or update this requirement as
part of the 2019 IRP. Avista will monitor the summer market depth and may revise the
planning margin standard from after reviewing work by the NPCC. The addition of a
seven percent summer planning margin for this IRP does not add additional resources
requirements above the winter peak requirement due to our dual peaking load profile,
but it will require the selection of resources than can provide both winter and summer

4 Peak winter loads can occur from the last two weeks of November through the second week of
February. The standard deviation of all the monthly peak loads in this period is 138 MW.
SC_PR_3-2 Attachment A Page 87 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-5
peaking capabilities. Avista intends on meeting this requirement using owned resources
or power purchase agreements (PPAs) as identified in Chapter 11 – Preferred
Resource Strategy. Avista does not plan to use short-term market purchases to meet
the 14 and seven percent planning margin requirements.

Northwest Power and Conservation Council Operating Reserve Planning Data
The NPCC’s Seventh Plan and the Washington Commission’s 2015 IRP
acknowledgment letters request utilities to provide additional documentation regarding
reserves:

Utilities should include their planning assumptions for the provision of operating
reserves in their Integrated Resource Plans and Bonneville in its Resource Program.
These assumptions should emphasize reliability ahead of economic operations, that
is, reasonable estimates for times of power system stress. The following should also
be included:

 An estimate of the utility’s or Bonneville’s requirement for operating reserves
 Reasonable planning assumptions for the amount of the reserve requirement
estimated to be held on hydropower generation and which projects should be
assigned in power system models to provide these reserves
 Reasonable planning assumptions for the amount of the reserve requirement
estimated to be held on thermal plants and which plants should be assigned
in power system models to provide these reserves
 Reasonable planning assumptions for any third-party provision of reserves5

In response to this request, Avista provides the following:
 Avista includes operating reserves as part of its planning criteria; these
operating reserves are not included in the 14 percent winter or the seven
percent summer planning margin calculations. For the 2018 winter peak hour
estimated load, the operating reserves sum to 122 MW.6 An additional 16
MW7 of capacity is for within hour requirements such as regulation.
Regulation is typically met with Avista’s hydroelectric facilities. Avista tends to
hold out of the money thermal resources as non-spinning reserve resources
and the remaining requirements at its hydroelectric facilities. The amounts
held at the hydroelectric system versus thermal facilities depends on water
conditions and plant economics. For example, it is possible to hold all these
reserves on the hydroelectric system in summer months due to lower flows
and Avista’s storage at both the Noxon Rapids and Mid-Columbia projects.
 Avista has several hydroelectric units with the ability to provide operating
reserves; these include Noxon Rapids, Cabinet Gorge, Long Lake and
contracted Mid-Columbia projects. These facilities provide both spinning and

5 Northwest Power and Conservation Council’s Seventh Power Plan, Chapter 4, Page 7, REG-4
6 Avista holds operating reserves for the entire control area, including non-Avista generation and loads.
7 Avista typically holds 20 MW for both increases and decreases during normal operating conditions (non-
peak event), but may vary depending on wind forecasts.
SC_PR_3-2 Attachment A Page 88 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-6
non-spinning reserves. Under the new FRR rules, only four units at Noxon
Rapids and one of Cabinet Gorge’s units can provide this capacity.
 Avista can also provide operating reserves with its thermal fleet. Rathdrum
CT, and Northeast CT can provide non-spinning reserves. Coyote Springs 2
and Lancaster can provide non-spin, spinning, and FRR reserves when the
units are not at full capacity.
 Avista on occasion will contract to sell reserves to other control areas under
short-term agreements, but this information is proprietary.

Energy Imbalance Market
Avista recently participated in a regional effort to evaluate the viability of an intra-hour
Energy Imbalance Market (EIM) in the Northwest Power Pool area. The Market
Coordination (MC) Initiative officially launched on March 19, 2012 to explore alternatives
to address the growing operational and commercial challenges to integrate variable
energy resources affecting the regional power system. In December 2015, the MC
evaluation effort concluded. The agreement ended after the group could not agree to a
final market design and several participants decided to join the California Independent
System Operator (CAISO) Western EIM.

Avista is conducting a cost/benefit analysis associated with joining the CAISO EIM. This
analysis will be complete in the fall of 2017. Avista is also evaluating other factors
influencing the decision to join the CAISO EIM. These include the reduction of near term
market liquidity as other utilities join the EIM and the additional integration of renewable
resources in our service territory. Avista will use the cost/benefit analysis and evaluation
of other market factors to inform its decision to participate in the Western EIM.

Balancing Loads and Resources
Both single-hour and sustained-peaking requirements compare future load and
resource projections to identify any shortages. The single peak hour is a larger concern
in the winter than the three-day sustained 18-hour peak. During winter months, the
hydroelectric system can sustain generation levels for longer periods than in the
summer due to higher inflows. Figure 6.1 illustrates the winter balance of loads and
resources. The first year Avista has a significant winter capacity deficit is November
2026 when including future conservation acquisitions. If all conservation programs
ended, the first capacity deficit would occur in January 2022. Until recently, the capacity
position was short beginning in 2022, but the extension of a PPA from the Mid-Columbia
PUDs filled this deficiency.

Avista plans to meet its summer peak load with a smaller planning margin than in the
winter. During summer months, operating reserve and regulation obligations are
included in addition to a seven percent planning margin. Market purchases in the deep
regional market should satisfy any weather-induced load variation or generation forced
outage that otherwise would be included in the planning margin as is the case in the
higher 14 percent winter planning margin. Resource additions to serve winter peaks
meet smaller summer deficits as well. Figure 6.2 shows Avista’s summer resource
SC_PR_3-2 Attachment A Page 89 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-7
balance. Like the winter, Avista expects its first summer deficit in 2027 after the
expiration of the Lancaster PPA in October 2026.

Figure 6.1: Winter One-Hour Capacity Load and Resources

Figure 6.2: Summer One-Hour Capacity Load and Resources

0
500
1,000
1,500
2,000
2,500
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
g
a
w
a
t
t
s
Existing Resources & Rights
Load w/o Conservation + Cont.
Load w/ Conservation + Cont.
0
500
1,000
1,500
2,000
2,500
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
g
a
w
a
t
t
s
Existing Resources & Rights
Load w/o Conservation + Cont.
Load w/ Conservation + Cont.
SC_PR_3-2 Attachment A Page 90 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-8
Energy Planning
For energy planning, resources must be adequate to meet customer requirements even
when loads are high for extended periods, or a sustained outage limits the contribution
of a resource. Where generation capability is not adequate to meet these variations,
customers and the utility must rely on the short-term electricity market. In addition to
load variability, Avista holds energy-planning margins accounting for variations in
month-to-month hydroelectric generation.

As with capacity planning, there are differences in regional opinions on the proper
method for establishing energy-planning margins. Many utilities in the Northwest base
their energy planning margins on the amount of energy available during the “critical
water” period of 1936/37.8 The critical water year of 1936/37 is low on an annual basis,
but it does not represent a low water condition in every month. The IRP could target
resource development to reach a 99 percent confidence level on being able to deliver
energy to its customers, and it would significantly decrease the frequency of its market
purchases. However, this strategy requires investments in approximately 200 MW of
generation in addition to the capacity planning margins included in the Expected Case
of the 2017 IRP to cover a one-in-one-hundred year event. Investments to support this
high level of reliability would increase pressure on retail rates for a modest benefit.
Avista plans to the 90th percentile for hydroelectric generation. Using this metric, there is
a one-in-ten-year chance of needing to purchase energy from the market in any given
month over the IRP timeframe.

Beyond load and hydroelectric variability, Avista’s legacy WNP-3 contract with BPA
contains supply risk. The contract includes a return energy provision in favor of BPA
that can equal 32 aMW annually. Under adverse market conditions, BPA almost
certainly would exercise this right, as it did during the 2001 Energy Crisis. To account
for this contract risk, the energy contingency increases by 32 aMW until the contract
expires in 2019. With the addition of WNP-3 contract contingency to load and
hydroelectric variability, the total energy contingency amount equals 231 aMW in 2018.
See Figure 6.3 for the summary of the annual average energy load and resource net
position.

8 The critical water year represents the lowest historical generation level in the streamflow record.
SC_PR_3-2 Attachment A Page 91 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-9
Figure 6.3: Annual Average Energy Load and Resources

Washington State Renewable Portfolio Standard
Washington’s EIA requires utilities with more than 25,000 customers to source 9
percent of their energy from qualified renewables through 2019 and 15 percent by 2020.
Utilities also must acquire all cost effective conservation as explained in Chapter 5 –
Energy Efficiency and Demand Response. In 2011, Avista signed a 30-year PPA with
Palouse Wind to help meet the EIA goal. In 2012, an amendment to the EIA allowed
Avista’s 50-MW Kettle Falls project to qualify for the EIA goals beginning in 2016.

Table 6.1 shows the forecast amount of RECs Avista needs to meet the EIA renewable
requirement and the amount of qualifying resources already in Avista’s generation
portfolio. Without the ability to roll RECs from previous years, Avista would require
additional renewables in 2026. With this ability, Avista does not need additional EIA
resources over the planning horizon of this IRP. The company may have surplus RECs
depending upon the qualifying output of Kettle Falls and Palouse Wind. Kettle Falls
qualifying output varies depending upon the availability of qualifying fuel and the
economics of the facility. Given its expected renewables surplus until 2020, Avista will
market the excess RECs until 2019. Beginning in 2019, surplus RECs will roll into 2020,
allowing the banking provision to delay additional renewable resource investment.

0
500
1,000
1,500
2,000
2,500
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Existing Resources & Rights
Load w/o Conservation + Cont.
Load w/ Conservation + Cont.
SC_PR_3-2 Attachment A Page 92 of 205
Chapter 6: Long-Term Position
Avista Corp 2017 Electric IRP 6-10
Table 6.1: Washington State EIA Compliance Position Prior to REC Banking (aMW)

2018 2020 2025 2030 2035
Percent of Washington Sales 9% 15% 15% 15% 15%
Two-Year Rolling Average Washington Retail
Sales Estimate 644 658 683 699 720

Renewable Goal -58 -99 -103 -105 -108
Incremental Hydroelectric 22 22 22 22 22
Net Renewable Goal -36 -77 -81 -83 -86

Other Available REC's
Palouse Wind with Apprentice Credits 48 48 48 48 48
Kettle Falls 33 33 33 33 33
Net Renewable Position (before rollover RECs) 45 4 0 -2 -5

SC_PR_3-2 Attachment A Page 93 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
7. Policy Considerations

Public policy affects Avista’s current generation resources and the resources it can
pursue. Each resource option presents different cost, environmental, operational,
political, regulatory, and siting challenges. Regulatory environments regarding energy
topics such as renewable energy and greenhouse gas regulation continue to evolve since
publication of the last IRP. Current and proposed regulations by federal and state
agencies, coupled with political and legal efforts, have implications for the development
and continued use of coal and natural gas-fired generation. This chapter discusses
pertinent public policy issues relevant to the IRP.

Environmental Issues
The evolving and sometimes contradictory nature of environmental regulation from state
and federal perspectives creates challenges for resource planning. The IRP cannot add
renewables or reduce emissions in isolation from topics such as system reliability, least
cost requirements, price mitigation, renewable portfolio standards, financial risk
management, and meeting changing environmental requirements. Each generating
resource has distinctive operating characteristics, cost structures, and environmental
regulatory challenges that can change significantly based on timing and location. All
resource choices have costs and benefits requiring careful consideration of the utility and
customer needs being fulfilled, their location, and the regulatory and policy environment
at the time of procurement.

Traditional thermal generation technologies, like coal and natural gas-fired plants, provide
reliable capacity and energy. New coal plants as compared to natural gas-fired resources
have environmental and economic disadvantages. It is unlikely without major
technological improvements any new coal-fired resources will be developed in the U.S.
Existing coal-fired resources are also under increasing pressure from lower-cost
resources and increasing regulatory constraints and costs.

Natural gas-fired plants have relatively low capital costs, can typically be located closer
to load centers, have relatively short construction time frames, lower emissions and fewer
waste issues than coal, and are often the only available utility-scale baseload resource.
On the other hand, higher fuel price volatility historically affected natural gas-fired plant
economics. In addition, their performance decreases in hot weather, they are difficult to
site with sufficient water rights for their efficient operation, and they emit greenhouse
gases.

Chapter Highlights

Avista’s Climate Policy Council monitors greenhouse gas legislation and

does not directly impact any of Avista’s generating fleet.
SC_PR_3-2 Attachment A Page 94 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
Renewable energy technologies such as wind, biomass, geothermal, and solar have
different benefits and challenges. Renewable resources have low or no fuel costs and
few, if any, direct emissions. However, solar and wind-based generation have limited or
no capacity value, their own unique siting limitations, and their variable output can present
integration challenges requiring additional capacity investments. Renewable resources
are often located to maximize capability rather than proximity to load centers. The need
to site renewable resources in remote locations often requires significant investments in
transmission and capacity expansion, as well as mitigating possible wildlife and aesthetic
issues. Distributed resources may alleviate some of these issues, but the price
differentials of distributed resources make them more difficult to develop at utility scale.
Unlike fossil fuel-fired plants, the fuel for non-biomass renewables may not be
transportable to utilize existing transmission or to minimize opposition to project
development. Dependence on the health of the forest products industry and access to
biomass materials, often located in publicly owned forests, poses challenges to biomass
facilities. Transportation costs and logistics also complicate the location of biomass
plants.

The long-term economics of renewable resources also faces some uncertainties. Federal
investment and production tax credits are set to expire. The extension credits and grants
may not be sustainable given their impact on government finances and the maturity of
wind and solar technologies. Many relatively unpredictable factors affect renewables,
such as renewable portfolio standards (RPS), construction and component prices,
international trade issues and currency exchange rates. Decreasing capital costs for wind
and solar may slow or stop.

The design and scope of greenhouse gas regulation is in a state of flux due to legal
challenges and evolving political realities. As a result, greenhouse gas policy-making is
shifting from the federal to the state and local level. Since the 2015 IRP publication,
changes in the approach to greenhouse gas emissions regulation and supporting
programs, include:

 The EPA proposed actions to regulate greenhouse gas emissions under the Clean
Air Act (CAA) through the proposed Clean Power Plan (CPP) were stayed by the
U.S. Supreme Court on February 9, 2016;
 The President signaled a shift in federal priorities through Executive Orders as well
as proposed budgets.
 EPA plans to reevaluate the CPP and submit a new CPP proposal to the Office of
Management and Budget;
 California failed to pass an extension to its cap-and-trade program beyond 2020,
but did raise its RPS to 50 percent and expanded energy storage requirements;
and
 The State of Washington implemented the Clean Air Rule

Natural Gas System Emissions
The physical makeup of the natural gas system includes extraction rigs, pipelines and
storage; each of these facilities have fugitive emissions. Fugitive emissions are the
SC_PR_3-2 Attachment A Page 95 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
unintended or irregular releases of natural gas as part of the production cycle. The EPA
introduced the Natural Gas STAR Program in 1993 in response to these emissions
concerns. This Natural Gas STAR Program is a voluntary program allowing the self-
reporting of emission reduction technologies and practices and includes all of the major
industry sectors. In May 2016, the EPA finalized rules to reduce methane emissions from
wells under the CAA. The program requires natural gas well owners to find and repair
leaks at the well site no less than twice per year and four times per year at compressor
stations. The EPA placed a 90-day delay on portions of the rule to allow additional
comments.

Natural gas wells utilizing shale deposits have a high production curve at the beginning
of the extraction process and then dramatically levels off. If not constructed properly, there
is a risk of leakage that may lower the return on investment. In addition, risk of increased
regulation incentivizes producers to manage emissions as effectively as possible as more
regulations generally increase costs and reduce return on investments. Over time a
smaller return on investment could mean the difference in survival outcomes for each
producer.

Avista’s Climate Change Policy Efforts
Avista’s Climate Policy Council is an interdisciplinary team of management and other
employees that:
 Facilitates internal and external communications regarding climate change issues;
 Analyzes policy impacts, anticipates opportunities, and evaluates strategies for
Avista Corporation; and
 Develops recommendations on climate related policy positions and action plans.

The core team of the Climate Policy Council includes members from Environmental
Affairs, Government Relations, External Communications, Engineering, Energy
Solutions, and Resource Planning groups. Other areas participate for topics as needed.
The meetings for this group include work for both immediate and long-term concerns.
Immediate concerns include reviewing and analyzing proposed or pending state and
federal legislation and regulation, reviewing corporate climate change policy, and
responding to internal and external requests about climate change issues. Longer-term
issues involve emissions measurement and reporting, different greenhouse gas policies,
actively participating in legislation, and benchmarking climate change policies and
activities against other organizations.

Membership in the Edison Electric Institute is Avista’s main vehicle to engage in federal-
level climate change dialog, supplemented by other industry affiliations. Avista monitors
regulations affecting hydroelectric and biomass generation through its membership in
other associations.

State and Federal Environmental Policy Considerations
The CPP was the focus of federal greenhouse gas emissions policies in the 2015 IRP
and the starting point for this IRP emission reduction assumptions. Details about
greenhouse gas emissions modeling are in Chapter 10 – Market Analysis. As explained
SC_PR_3-2 Attachment A Page 96 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
above, the application and form of the future CPP is uncertain as this IRP is being written.
However, a form of federal regulation will be put in place. As explained in Chapter 10, this
IRP does not include specific carbon pricing with the exception of states and provinces
with existing carbon trading and taxing regulations. This IRP does include regional
emission reduction goals leading to a shadow price of carbon pricing, rather than an
arbitrary carbon price. If a carbon tax or cap and trade program develops in the future, it
will require alternative analysis in a later IRP.

EPA Regulations
EPA regulations, or the States’ authorized versions, directly, or indirectly, affecting
electricity generation include the CAA, along with its various components, including the
Acid Rain Program, the National Ambient Air Quality Standard, the Hazardous Air
Pollutant rules, and Regional Haze Programs. The U.S. Supreme Court ruled the EPA
has authority under the CAA to regulate greenhouse gas emissions from new motor
vehicles and the EPA has issued such regulations. When these regulations became
effective, carbon dioxide and other greenhouse gases became regulated pollutants under
the Prevention of Significant Deterioration (PSD) preconstruction permit program and the
Title V operating permit program. Both of these programs apply to power plants and other
commercial and industrial facilities. In 2010, the EPA issued a final rule, known as the
Tailoring Rule, governing the application of these programs to stationary sources, such
as power plants. EPA proposed a rule in early 2012, and modified in 2013, setting
standards of performance for greenhouse gas emissions from new and modified fossil
fuel-fired electric generating units and for existing sources through the draft CPP in June
2014. The EPA released the final CPP rules and the Carbon Pollution Standards (CPS)
as published in the Federal Register on October 23, 2015, when they were both
challenged thorough a series of lawsuits. Standards under Section 111(d) of the CAA are
currently stayed by the Supreme Court. The EPA also finalized new source performance
standards (NSPS) for new, modified and reconstructed fossil fuel-fired generation under
CAA section 111(b).

Promulgated PSD permit rules may affect Avista’s thermal generation facilities in the
future. These rules can affect the amount of time to obtain permits for new generation,
major modifications to existing generating units, and the final limitations contained in
permits. The promulgated and proposed greenhouse gas rulemakings mentioned above
have been legally challenged in multiple venues so we cannot fully anticipate the outcome
or extent our facilities may be impacted, nor the timing of rule finalization.

Clean Air Act Operating Permits
The CAA, originally adopted in 1970 and modified significantly since, intends to control
covered air pollutants to protect and improve air quality. Avista complies with the
requirements under the CAA in operating our thermal generating plants. Title V operating
permits are required for our largest generation facilities and are renewed every five years.
Title V operating permit renewal applications are in process for Colstrip Units 3 and 4,
Coyote Springs 2 and Kettle Falls. Boulder Park, Northeast CT, and other small facilities
require only minor source operating or registration permits based on their limited
operation and emissions. Discussion of some major CAA programs follows.

SC_PR_3-2 Attachment A Page 97 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
New Source Proposal
After receiving over 2.5 million comments on the April 2012 proposal for new resources
under section 111(b) of the CAA, the EPA withdrew that proposal and submitted a new
proposal on September 20, 2013. This proposal covers new fossil fuel-fired resources
larger than 25 MW for the following resource types:

 Natural gas-fired stationary combustion turbines: 1,000 pounds CO2 per MWh for
units burning greater than 850 mmBtu/hour and 1,100 pounds CO2 per MWh units
burning less than or equal to 850 mmBtu/hour.
 Fossil fuel-fired utility boilers and integrated gasification combined cycle (IGCC)
units: 1,100 pounds CO2 per MWh over a 12-operating month period or 1,000–
1,500 pounds CO2 per MWh over a seven-year period.

The EPA finalized the new source standard on August 3, 2015. The final rule differs from
the proposal, which was the basis for the development of this IRP. The final rule guided
modeling assumptions for the 2017 IRP.

Acid Rain Program
The Acid Rain Program is an emission-trading program for reducing nitrous dioxide by
two million tons and sulfur dioxide by 10 million tons below 1980 levels from electric
generation facilities. Avista manages annual emissions under this program for its
ownership interest in Colstrip Units 3 and 4, Coyote Springs 2, and Rathdrum.

National Ambient Air Quality Standards
EPA sets National Ambient Air Quality Standards for pollutants considered harmful to
public health and the environment. The CAA requires regular court-mandated updates to
occur for nitrogen dioxide, ozone, and particulate matter. Avista does not anticipate any
material impacts on its generation facilities from the revised standards at this time.

Hazardous Air Pollutants (HAPs)
HAPs, often known as toxic air pollutants or air toxics, are pollutants that may cause
cancer or other serious health effects. EPA regulates toxic air pollutants from a published
list of industrial sources referred to as "source categories". These pollutants must meet
control technology requirements if they emit one or more of the pollutants in significant
quantities. EPA finalized the Mercury Air Toxic Standards (MATS) for the coal and oil-
fired source category in 2012. Colstrip Units 3 & 4’s existing emission control systems
should be sufficient to meet mercury limits. For the remaining portion of the rule that
utilized Particulate Matter as a surrogate for air toxics (including metals and acid gases),
the Colstrip owners reviewed recent stack testing data and expected that no additional
emission control systems would be needed for Units 3 & 4 MATS compliance.

Regional Haze Program
EPA set a national goal to eliminate man-made visibility degradation in Class I areas by
the year 2064. Individual states are to take actions to make “reasonable progress” through
10-year plans, including application of Best Available Retrofit Technology (BART)
requirements. BART is a retrofit program applied to large emission sources, including
electric generating units built between 1962 and 1977. In the absence of state programs,
SC_PR_3-2 Attachment A Page 98 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
EPA may adopt Federal Implementation Plans (FIPs). On September 18, 2012, EPA
finalized the Regional Haze FIP for Montana. The FIP includes both emission limitations
and pollution controls for Colstrip Units 1 and 2. Colstrip Units 3 and 4 are not currently
affected, although the units will be evaluated for Reasonable Progress at the next review
period in September 2017. Avista does not anticipate any material impacts on Colstrip
Units 3 and 4 at this time. In November 2012, several groups petitioned the U.S. Court of
Appeals for the Ninth Circuit for review of Montana’s FIP. The Court vacated portions of
the Final Rule and remanded back to EPA for further proceedings on June 9, 2015.

EPA Mandatory Reporting Rule
Any facility emitting over 25,000 metric tons of greenhouse gases per year must report
its emissions to EPA. Colstrip Units 3 and 4, Coyote Springs 2, and Rathdrum currently
report under this requirement. The Mandatory Reporting Rule also requires greenhouse
gas reporting for natural gas distribution system throughput, fugitive emissions from
electric power transmission and distribution systems, fugitive emissions from natural gas
distribution systems, and from natural gas storage facilities. Washington requires
mandatory greenhouse gas emissions reporting similar to the EPA requirements and
Oregon has similar reporting requirements.

Coal Ash Management and Disposal
The EPA issued a final rule regarding coal combustion residuals (CCR) in 2014. This
affects Colstrip since it produces CCR. The rule establishes technical requirements for
CCR landfills and surface impoundments under Subtitle D of the Resource Conservation
and Recovery Act, the nation’s primary law for regulating solid waste. The CCR rule
became effective October 2015. The owners of Colstrip are developing a multi-year plan
to comply with the new CCR standards. Any financial or operational impacts to Colstrip
from the CCR are still estimates, but are included in this IRP.

Particulate Matter
Particulate Matter (PM or particle pollution) is the term for a mixture of solid particles and
liquid droplets found in the air. Some particles, such as dust, dirt, soot, or smoke, are
large or dark enough to be seen with the naked eye. Others are so small they can only
be detected using an electron microscope. Particle pollution includes:

 PM10: inhalable particles, with diameters that are generally 10 micrometers and
smaller; and
 PM2.5: fine inhalable particles, with diameters that are generally 2.5 micrometers
and smaller.

There are different standards for PM10 and PM2.5. Limiting the maximum amount of PM
to be present in outdoor air protects human health and the environment. The CAA
requires EPA to set National Ambient Air Quality Standards (NAAQS) for PM, as one of
the six criteria pollutants considered harmful to public health and the environment. The
law also requires EPA to periodically review the standards to ensure that they provide
adequate health and environmental protection, and to update those standards as
necessary.

SC_PR_3-2 Attachment A Page 99 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
Avista has ownership and/or operational control for the following thermal electric
generating stations: Boulder Park, Colstrip, Coyote Springs, Kettle Falls, Lancaster,
Northeast and Rathdrum that produce PM. Table 7.1 shows each of these generating
stations, location, status of the surrounding area with NAAQS for PM2.5 and PM10,
operating permit and PM pollution controls.

Table 7.1: Avista Owned and Controlled PM Emissions

Thermal
Generating
Station
Location
County, City,
State
PM2.5
NAAQS
Status
PM10
NAAQS
Status
Air
Operating
Permit
PM Pollution
Controls
Boulder
Park
Spokane Co.,
Spokane, WA
Attainment Maintenance Minor
Source
Pipeline Natural
Gas
Colstrip Rosebud Co.,
Lame Deer, MT
Attainment Non-
Attainment
Major
Source
Title V OP
Fluidized Bed
Wet Scrubber
Coyote
Springs
Morrow Co.,
Boardman, OR
Attainment Attainment Major
Source
Title V OP
Pipeline Natural
Gas, Air filters
Kettle Falls Lincoln Co.,
Kettle Falls, WA
Attainment Attainment Major
Source
Title V OP
Multi-clone
collector,
Electrostatic
Precipitator
Lancaster Kootenai Co.,
Rathdrum, ID
Attainment Attainment Major
Source
Title V OP
Pipeline Natural
Gas, Air filters
Northeast Spokane Co.,
Spokane, WA
Attainment Maintenance Minor
Source
Pipeline Natural
Gas, Air filters
Rathdrum Kootenai Co.,
Rathdrum, ID
Attainment Attainment Major
Source
Title V OP
Pipeline Natural
Gas, Air filters

Our generating stations are issued air quality operating permits from the appropriate EPA
delegated air quality agency under the authority of the Federal CAA. These operating
permits require annual compliance certifications and are fully renewed every five years to
incorporate any new standards including any updated NAAQS status. If warranted, EPA
would issue specific requirements to protect human health and the environment at that
time.

State and Regional Level Policy Considerations
The lack of a comprehensive federal greenhouse gas policy encouraged states, such as
California, to develop their own climate change laws and regulations. Climate change
legislation takes many forms, including economy-wide regulation under a cap and trade
system, a carbon tax, and emissions performance standards for power plants.
Comprehensive climate change policy can include multiple components, such as
renewable portfolio standards, energy efficiency standards, and emission performance
standards. Washington enacted all of these components, but other Avista jurisdictions
have not. Individual state actions produce a patchwork of competing rules and regulations
for utilities to follow and may be particularly problematic for multi-jurisdictional utilities
SC_PR_3-2 Attachment A Page 100 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
such as Avista. There are 29 states, plus the District of Columbia, with active renewable
portfolio standards, and eight additional states have adopted voluntary standards.1

Idaho Policy Considerations
Idaho does not regulate greenhouse gases or have an RPS. There is no indication Idaho
is moving toward regulation of greenhouse gas emissions beyond federal regulations.

Montana Policy Considerations
Montana’s RPS law requires covered utilities to meet 15 percent of their load with qualified
renewables since 2015. Montana implemented a mercury emission standard under Rule
17.8.771 in 2009. The standard exceeds the most recently adopted federal mercury limit.
Avista’s generation at Colstrip Units 3 and 4 have emissions controls currently meeting
Montana’s mercury emissions goal.

Oregon Policy Considerations
The State of Oregon has a history of greenhouse gas emissions and renewable portfolio
standards legislation. The Legislature enacted House Bill 3543 in 2007, calling for, but
not requiring, reductions of greenhouse gas emissions to 10 percent below 1990 levels
by 2020 and 75 percent below 1990 levels by 2050. Compliance is expected through a
combination of the RPS and other complementary policies, like low carbon fuel standards
and energy efficiency measures. The state has been working towards the adaptation of
comprehensive requirements to meet these goals. Oregon’s SB 1547, enacted in March
2016, ends the use of coal to serve Oregon loads by 2030 and increases the RPS to 50
percent by 2040. HB 2135, or the cap and trade bill, is under consideration at the time
this chapter is being written. This bill would repeal the greenhouse gas emissions goals
stated above and would require the Environmental Quality Commission to adopt
greenhouse gas emissions goals for 2025, and set limits for years 2035 and 2050.

These reduction goals are in addition to a 1997 regulation requiring fossil-fueled
generation developers to offset carbon dioxide (CO2) emissions exceeding 83 percent of
the emissions of a state-of-the-art gas-fired combined cycle combustion turbine by
funding offsets through the Climate Trust of Oregon.

Washington State Policy Considerations
The State of Washington has enacted several fossil-fueled generation emissions and
resource diversification measures. A 2004 law requires new fossil-fueled thermal electric
generating facilities of more than 25 MW of generation capacity to offset CO2 emissions
through third-party mitigation, purchased carbon credits, or cogeneration. An agreement
with the State of Washington requires the Centralia Coal Plant to shut down one unit by
December 2020 and the other unit by December 2025.

Washington’s EIA requires utilities with more than 25,000 retail customers to use qualified
renewable energy or renewable energy credits to serve nine percent of retail load by 2012
and 15 percent by 2020. Failure to meet RPS requirements results in at least a $50 per
MWh fine. The initiative also requires utilities to acquire all cost-effective conservation

1 http://www.dsireusa.org/resources/detailed-summary-maps/
SC_PR_3-2 Attachment A Page 101 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
and energy efficiency measures up to 110 percent of avoided cost. Additional details
about the energy efficiency portion of the EIA are in Chapter 6 – Long-Term Position.

In 2012, Senate Bill 5575 amended the EIA to define Kettle Falls Generating Station and
other legacy biomass facilities commencing operation before March 31, 1999 as EIA-
qualified resources beginning in 2016. A 2013 EIA amendment allows multistate utilities
to import RECs from outside the Pacific Northwest to meet renewable goals and allows
utilities to acquire output from the Centralia Coal Plant without jeopardizing alternative
compliance methods.

Avista will meet or exceed its renewable requirements in this IRP planning period through
a combination of qualified hydroelectric upgrades, wind generation from the Palouse Wind
PPA, and output from its Kettle Falls generation facility. The 2017 IRP Expected Case
ensures that Avista meets all EIA RPS goals.

Former Governor Christine Gregoire signed Executive Order 07-02 in February 2007
establishing the following GHG emissions goals:
 1990 levels by 2020;
 25 percent below 1990 levels by 2035;
 50 percent below 1990 levels by 2050 or 70 percent below Washington’s expected
emissions in 2050;
 Increase clean energy jobs to 25,000 by 2020; and
 Reduce statewide fuel imports by 20 percent.

The Washington Department of Ecology adopted regulations to ensure that its State
Implementation Plan comports with the requirements of the EPA's regulation of
greenhouse gas emissions. We will continue to monitor actions by the Department as it
may proceed to adopt additional regulations under its CAA authorities. In 2007, Senate
Bill 6001 prohibited electric utilities from entering into long-term financial commitments
beyond five years for fossil-fueled generation creating 1,100 pounds per MWh or more of
greenhouse gases. Beginning in 2013, the emissions performance standard is lowered
every five years to reflect the emissions profile of the latest commercially available CCCT.
The emissions performance standard effectively prevents utilities from developing new
coal-fired generation and expanding the generation capacity of existing coal-fired
generation unless they can sequester emissions from the facility. The Legislature
amended Senate Bill 6001 in 2009 to prohibit contractual long-term financial
commitments for electricity deliveries that include more than 12 percent of the total power
from unspecified sources. The Department of Commerce filed a rule adopting a standard
of 970 pounds per MWh for greenhouse gas emissions on March 6, 2013, with rules
becoming effective on April 6, 2013.2 Commerce announced that work for the next update
would begin in the summer of 2017.

2 http://www.commerce.wa.gov/Programs/Energy/Office/Utilities/Pages/EmissionPerfStandards.aspx
SC_PR_3-2 Attachment A Page 102 of 205
Chapter 7–Policy Considerations
Avista Corp 2017 Electric IRP
April 29, 2014, Washington Governor Jay Inslee issued Executive Order 14-04,
“Washington Carbon Pollution Reduction and Clean Energy Action.” The order created a
“Climate Emissions Reduction Task Force” tasked with providing recommendations to the
Governor on designing and implementing a market-based carbon pollution program to
inform possible legislative proposals in 2015. The order also called on the program to
“establish a cap on carbon pollution emissions, with binding requirements to meet our
statutory emission limits.” The order also states that the Governor’s Legislative Affairs
and Policy Office “will seek negotiated agreements with key utilities and others to reduce
and eliminate over time the use of electrical power produced from coal.” The Task Force
issued a report summarizing its efforts, which included a range of potential carbon-
reducing proposals. Subsequently, in January 2015, at Governor Inslee’s request, the
Carbon Pollution Accountability Act was introduced as a bill in the Washington legislature.
The bill includes a proposed cap and trade system for carbon emissions from a wide
range of sources, including fossil-fired electrical generation, “imported” power generated
by fossil fuels, natural gas sales and use, and certain uses of biomass for electrical
generation. The bill was not enacted during the 2015 legislative session. After the
conclusion of the 2015 legislative sessions, Governor Inslee directed the Department of
Ecology to commence a rulemaking process to impose a greenhouse gas emission
limitation and reduction mechanism under the agency’s CAA authority to meet the future
emissions limits established by the Legislature in 2008. This resulted in Washington’s
Clean Air Rule (CAR).

The CAR imposes new compliance obligations on sources identified by Ecology. The rule
imposes caps and requirements to reduce or offset emissions on large emitting facilities,
fuel providers and natural gas distribution companies. It initially applies to 29 entities.
Compliance obligations for energy-intensive trade-exposed industries, including pulp and
paper manufacturers, steel and aluminum manufacturers and food processors, are
deferred for three years. When fully implemented, the CAR could cover as many as 70
emitters who account for about two-thirds of Washington’s emissions. The CAR caps
emissions for facilities emitting more than 100,000 metric tons per year, and reduces the
emissions threshold by 5,000 metric tons per year, until covering all entities emitting over
70,000 metric tons by 2035. The Washington Commission may implement rules regarding
RCW 70.235, from the Executive Order 07-02. The CAR became effective January 1,
2017 and is currently under legal challenge. Avista does not have any generating facilities
under the CAR rule.
SC_PR_3-2 Attachment A Page 103 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-1

8. Transmission & Distribution Planning

Introduction
This chapter introduces the Avista Transmission and Distribution systems and provides a
brief description of how Avista studies these systems and recommends projects that keep
the systems functioning reliably. Avista’s Transmission System is only one part of the
networked Western Interconnection, so a discussion of regulations and regional planning
is also provided. This chapter includes a brief summary of planned transmission projects
and generation interconnection requests currently under study, and provides links to
documents describing these studies in more detail. Further, this section describes how
distribution planning is now playing a role in the IRP.

Avista Transmission System
Avista owns and operates a system of over 2,200 miles of electric transmission facilities
including approximately 660 miles of 230 kV transmission lines and 1,550 miles of 115
kV transmission lines (see Figure 8.1).

Figure 8.1: Avista Transmission System

Section Highlights





SC_PR_3-2 Attachment A Page 104 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-2

230 kV Backbone
The backbone of the Avista Transmission System functions at 230 kV. Figure 8.2 shows
a station-level drawing of Avista’s 230 kV Transmission System including
interconnections to neighboring utilities. Avista’s 230 kV Transmission System is
interconnected to the BPA 500 kV transmission system at the Bell, Hot Springs and
Hatwai Stations.

Figure 8.2: Avista 230 kV Transmission System

Transmission System Areas
Avista separates its Transmission System into five geographical study areas:

1. Big Bend
2. Coeur d’Alene
3. Lewiston-Clarkston
4. Palouse
5. Spokane

Figure 8.3 shows the approximate boundaries of the study areas and these areas are
referenced individually in Avista’s Local Planning Report.

SC_PR_3-2 Attachment A Page 105 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-3

Figure 8.3: Avista Transmission System Planning Regions

Transmission Planning Requirements and Processes
Avista coordinates its transmission planning activities with neighboring interconnected
transmission operators. Avista complies with FERC requirements related to both regional
and local area transmission planning. This section describes several of the processes
and forums important to Avista transmission planning.

Western Electricity Coordinating Council
The Western Electricity Coordinating Council (WECC) is the group responsible for
promoting bulk electric system reliability, compliance monitoring, and enforcement in the
Western Interconnection. This group facilitates development of reliability standards and
helps coordinate operating and planning among its membership. WECC is the largest
geographic territory of the regional entities with delegated authority from the NERC and
the FERC. It covers all or parts of 14 Western states, the provinces of Alberta and British
Columbia, and the northern section of Baja, Mexico.1 See Figure 8.4 for the map of
WECC.

Peak Reliability
Peak Reliability (Peak) performs the federally mandated reliability coordinator function for
a majority of the Western Interconnection. While each transmission operator within the
Western Interconnection operates its respective transmission system, Peak has the
authority to direct specific actions to maintain reliable operation of the overall transmission
grid.

1 https://www.wecc.biz/Pages/About.aspx
SC_PR_3-2 Attachment A Page 106 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-4

Figure 8.4: NERC Interconnection Map

Northwest Power Pool
Avista is a member of the Northwest Power Pool (NWPP), an organization formed in 1942
when the federal government directed utilities to coordinate operations in support of
wartime production. The NWPP serves as a northwest electricity reliability forum, helping
to coordinate present and future industry restructuring, promoting member cooperation to
achieve reliable system operation, coordinating power system planning, and assisting the
transmission planning process. NWPP membership is voluntary and includes the major
generating utilities serving the Northwestern U.S., British Columbia and Alberta. The
NWPP operates a number of committees, including its Operating Committee, the Reserve
Sharing Group Committee, the Pacific Northwest Coordination Agreement (PNCA)
Coordinating Group, and the Transmission Planning Committee (TPC).

ColumbiaGrid
ColumbiaGrid formed on March 31, 2006. Its membership includes Avista, BPA, Chelan
County PUD, Grant County PUD, Puget Sound Energy, Seattle City Light, Snohomish
County PUD, and Tacoma Power. ColumbiaGrid aims to enhance and improve the
operational efficiency, reliability, and planned expansion of the Pacific Northwest
transmission grid. Consistent with FERC requirements issued in Orders 890 and 1000,
ColumbiaGrid provides an open and transparent process to develop sub-regional
transmission plans, assess transmission alternatives (including non-wires alternatives),
and provides a decision-making forum and cost-allocation methodology for new
transmission projects.

Northern Tier Transmission Group
The Northern Tier Transmission Group (NTTG) formed on August 10, 2007. NTTG
members include Deseret Power Electric Cooperative, Idaho Power, Northwestern
Energy, PacifiCorp, Portland General Electric, and Utah Associated Municipal Power
Systems. These members rely upon the NTTG committee structure to meet FERC’s
coordinated transmission planning requirements. Avista’s transmission network has a
number of strong interconnections with three of the six NTTG member systems. Due to
the geographical and electrical positions of Avista’s transmission network related to NTTG
SC_PR_3-2 Attachment A Page 107 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-5

members, Avista participates in the NTTG planning process to foster collaborative
relationships with our interconnected utilities.

Annual Transmission Planning Report
Avista’s Local Planning Report is the end product of both the Local Transmission Planning
Process and the annual Planning Assessment. The Local Transmission Planning Process
(Process) is outlined in Attachment K to Avista’s Open Access Transmission Tariff, FERC
Electric Volume No. 8. The Process identifies single system projects needed to mitigate
future reliability and load-service requirements for the Avista Transmission System. The
Planning Assessment is outlined in the NERC Reliability Standard TPL-001-4.

The Planning Assessment determines where the Transmission System may not meet
performance requirements as defined in the NERC Reliability Standards, and identifies
Corrective Action Plans addressing how the performance requirements will be met. The
Planning Assessment includes steady state contingency analysis, analysis of potential
voltage collapse, and transient technical studies. Development of the Local Planning
Report supports compliance with applicable NERC Reliability Standards as well as
satisfying necessary steps in the Local Transmission Planning Process.

The Local Planning Report provides a 10-year Transmission System expansion plan by
including all Transmission System facility improvements. The following sections
summarize information from this report and other studies done by the Transmission
Planning group in the 2016 Assessment.

Transmission System Study Results

Big Bend Area
The Big Bend area transmission system performance will significantly improve upon
completion of the Benton – Othello Station 115 kV Transmission Line Rebuild project.
Improvements are made with reconductor projects, the Saddle Mountain 230 kV Station
project, and the addition of communication aided protection schemes.

Coeur d’Alene Area
Completion of the Coeur d’Alene – Pine Creek 115 kV Transmission Line Rebuild project
and Cabinet – Bronx – Sand Creek 115 kV Transmission Line Rebuild project will improve
transmission system performance in the near and long term planning horizons. The
Sandpoint Reinforcement Project and installation of capacitor banks at the St. Maries
Substation are part of the long range plan for the area.

Lewiston/Clarkston Area
The transmission system in the Lewiston/Clarkston area performs well. Issues are limited
primarily to N-1-1 outages2 on the 230 kV system and voltage exceeding facility ratings

2 Failure of two separate facilities.
SC_PR_3-2 Attachment A Page 108 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-6

during light loading conditions. Installation of shunt reactors is recommended to mitigate
these issues.

Palouse Area
Completion of the Moscow 230 Station Rebuild project in 2014 mitigated several
performance issues. The remaining issue is a potential outage of both the Moscow and
Shawnee 230/115 kV transformers. An operational and strategic long term plan is under
development to best address a possible double transformer outage.

Spokane Area
Several performance issues exist with the present state of the transmission system in the
Spokane area and worsen with additional load growth. The staged construction of new
230 kV facilities at the Garden Springs 230 kV and Ninth and Central 230 kV Stations to
reinforce the area will be required. Dependency on Beacon Station leaves the system
susceptible to performance issues for outages related to the station.

Short Circuit Study
This study identified six undersized 230 kV breakers at Noxon and two undersized 115
kV breakers at Sunset. A list of corrective actions plans developed to mitigate
performance issues observed during the assessment are in the 2016 Annual Assessment
document.3

IRP Generation Interconnection Options
Table 8.1 shows the projects and cost information for each of the IRP-related studies
where Avista evaluated new generation options. These studies provide a high-level view
of generation interconnection costs, and are similar to third-party feasibility studies
performed under Avista’s generator interconnection process. In the case of third-party
generation interconnections, FERC policy requires a sharing of costs between the
interconnecting transmission system and the interconnecting generator. Accordingly, it is
anticipated that all identified generation integration transmission costs will not be directly
attributable to a new interconnected generator.

Large Generation Interconnection Requests
Third-party generation companies may request transmission studies to understand the
cost and timelines for integrating potential new generation projects. These requests follow
a strict FERC process, including three study steps to estimate the feasibility, system
impact, and facility requirement costs for project integration. After this process is
completed, a contract offer to integrate the project may occur and negotiations can begin
to enter into a transmission agreement if necessary. Table 8.2 lists major projects
currently in Avista’s interconnection queue.4

3 http://www.oasis.oati.com/AVAT/AVATdocs/2016_Avista_System_Planning_Assessment.pdf
4 http://www.oasis.oati.com/AVAT/AVATdocs/GIP_Queue-V83.pdf
SC_PR_3-2 Attachment A Page 109 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-7

Table 8.1: 2017 IRP Generation Study Transmission Costs

Project Size (MW) Cost Estimate ($ Millions)5
Kootenai County 100 2
Kootenai County 350 100
Rathdrum Station (115 kV) 26 <1
Rathdrum Station (115 kV) 50 <1
Rathdrum Station (115 kV) 200 55
Rathdrum Station (230 kV) 50 <1
Rathdrum Station (230 kV) 200 56
Thornton Station 100 <1
Othello Station 25 <1
Northeast Station (Spokane) 10 <1
Kettle Falls Station 10 <1
Long Lake 68 33
Monroe Street 80 2
Post Fall 10 <1
Post Falls 20 <1

Table 8.2: Third-Party Large Generation Interconnection Requests

Project Size
(MW)
Type Interconnection
Location
Proposed Date
#46 126 Wind Big Bend (WA) December 2018
#47 750 Wind Colstrip 500kV (MT) September 2018
#49 144 Wind Big Bend (WA) September 2018
#50 450 Pumped Hydro Colstrip 500kV (MT) December 2020
#51 300 Solar Broadview (MT) December 2020
#52 100 Solar Big Bend (WA) July 2020
#53 12 Solar Big Bend (WA) October 2018
#54 40 Solar Big Bend (WA) January 2019

Distribution Planning
Avista continually evaluates its distribution system. The distribution system consists of
approximately 347 feeders covering 30,000 square miles, ranging in length from three to
73 miles. For rural distribution, feeder lengths vary widely to meet electrical loads resulting
from the startup and shutdown of the timber, mining, and agriculture industries. The goals
of the distribution evaluation are to determine if there are capacity limitations on the
system to serve current and future projected load for each individual feeder. The analysis
also includes whether or not the system meets reliability and level of service requirements
including voltage and power quality. When a potential constraint is identified an action
plan is prepared and compared against other options, and then the best course of action
is budgeted.

The primary role of electric distribution planning is to identify system capacity and service
reliability constraints, and subsequently identify the best and lowest life-cycle cost

5 Cost estimates are in 2017 dollars and use engineering judgment with a 50 percent margin for error.
SC_PR_3-2 Attachment A Page 110 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-8

solution. Traditionally this solution has centered on infrastructure upgrades such as poles,
wire, and cable. New technologies are emerging that may impact system analysis,
including storage, photovoltaic (solar) and demand response. As these alternatives
mature and evolve they are likely to play a role in our investment portfolio either as primary
solutions or capital deferment solutions. Avista has deployed several pilot projects with
the intent of determining how best to meet customer needs and maintain a high degree
of reliability now and in the future.

To properly evaluate each feeder for new technologies, load data and system data is
required. Quality load data is not available for all Avista feeders beyond monthly data logs
recording peak load and energy. Without detailed load data, evaluating new technologies
is limited to portions of the system with the available data. Detailed data is required to
validate whether new technologies solves current system constraint or just defers the
constraint to a different time.

Currently, 195 of 347 feeders have three-phase SCADA (Supervisory Control and Data
Acquisition) data available. We currently improve circuits as resource and budgeting allow
within our substation work schedule. As more demands beyond traditional capacity
constraints and level of service requirements are placed on the grid, an increased amount
of data is required to analyze and enhance the electric distribution system.

Further, new load forecasting techniques such as spatial load forecasting will be required.
This new forecasting method uses account GIS information regarding the feeder location
and can help forecast specific feeder load growth taking into account zoning,
demographics, land availability, and specific parcel information. With additional
investment in both technology and human capital, Avista will be prepared to quickly study
and implement new technologies on its system.

Deferred Capital Investment Analysis
New technologies such as storage, photovoltaics, and demand response programs could
help the electrical system by deferring or eliminating other investments. This is dependent
on the new technology to solve system constraints and meet customer expectations for
reliability. An advantage in using these technologies may be additional benefits
incorporated into the overall power system. For example, storage can help meet overall
power supply peak load needs, but it may also improve local reliability by providing
voltage support and deferring capital investment at the substation.

This section discusses the analysis for determining the capital investment deferment
value for distributed energy resources (DERs). Unfortunately, capital investment
deferment is not the same for all locations on the system. Feeders differ by whether they
are summer or winter peaking, the time of day the peaks occur, whether they are near
capacity or not, and how fast loads are growing in the area. It is not practical to have an
estimate for each feeders in an IRP, but it is prudent to have a representative estimate to
include in the resource selection analysis.

For this analysis, Avista uses three representative feeders on three substations; 1) Barker
Road, 2) Liberty Lake, and 3) Hallet & White. Each of these substations need capital
investment due to growth in the next several years. Each location was fitted with an
SC_PR_3-2 Attachment A Page 111 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-9

applicable storage device to determine how long the next investment could be deferred.
Then a financial analysis estimates the financial value to customers for deferring the
investment. The value of deferred investment is determined by comparing the present
value of the revenue requirement of the current plan versus the revenue requirement of
the alternative investment need when the storage device is installed. See Table 8.3 for
the results of the analysis.

The value of the deferment is a range as it depends when the storage device is installed.
The storage device has the greatest value when installed right before the investment is
needed rather than years before. For this plan, $10 per kW-year is assumed for the IRP
analysis. If distribution planning has a specific application for storage to meet distribution
needs, the IRP group can provide the power supply benefits to add to the specific capital
deferment analysis.

Table 8.3: Capital Deferment Analysis

Substation Storage
Capacity
(MW)
Storage
Energy
(MWh)
Deferment
Time
(Years)
Value
Range
($/kW-yr)
Barker Road 3.4 9.0 16 $5 - $16
Liberty Lake 6.0 43.0 21 $1 - $10
Hallet & White 1.7 10.5 9 $10 - $19

Grid Modernization
In 2008, an Avista system efficiencies team of operational, engineering, and planning staff
developed a plan to evaluate potential energy savings from transmission and distribution
system upgrades. The first phase summarized potential energy savings from distribution
feeder upgrades. The second phase, beginning in summer 2009, combined transmission
system topologies with right sizing distribution feeders to reduce system losses, improve
system reliability, and meet future load growth.

The system efficiencies team evaluated several efficiency programs to improve urban and
rural distribution feeders. The programs consisted of the following system enhancements:

 Conductor losses;
 Distribution transformers;
 Secondary districts; and
 Volt-ampere reactive compensation.

The analysis combined energy losses, capital investments, and reductions in O&M costs
resulting from the individual efficiency programs under consideration on a per feeder
basis. This approach provided a means to rank and compare the energy savings and net
resource costs for each feeder.

Building on the 2009 effort, a 2013 study assessed the benefits of distribution feeder
automation for increased efficiency and operability. The Grid Modernization Program
(GMP) combines the work from these system performance studies and provides Avista’s
SC_PR_3-2 Attachment A Page 112 of 205
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2017 Electric IRP 8-10

customers with refreshed system feeders with new automation capabilities across the
company’s distribution system. Table 8.4 shows the feeders currently planned for rebuild
and their associated energy savings. The total energy savings from both re-conductor
and transformer efficiencies for all completed feeders is approximately 1,930 MWh
annually.

The GMP charter ensures a consistent approach to how Avista addresses each project.
This program integrates work performed under various Avista operational initiatives,
including the Wood Pole Management Program, the Transformer Change-Out Program,
the Vegetation Management Program, and the Feeder Automation Program. The
Distribution Grid Modernization Program includes replacing undersized and deteriorating
conductors, and replacing failed and end-of-life infrastructure materials including wood
poles, cross arms, fuses, and insulators. It addresses inaccessible pole alignment, right-
of-way, under-grounding, and clear-zone compliance issues for each feeder section, as
well as regular maintenance work including leaning poles, guy anchors, unauthorized
attachments, and joint-use management. This systematic overview enables Avista to
cost-effectively deliver a modernized and robust electric distribution system that is more
efficient, easier to maintain, and more reliable for our customers.

Table 8.4: Planned Feeder Rebuilds

Feeder Area Year
Complete
Annual Energy
Savings (MWh)
MIL12F2 Millwood, WA 2017 186
ORO1280 Orofino, ID 2017 112
PDL1201 Clarkston, WA 2017 189
TUR112 Pullman, WA 2018 233
HOL1205 Lewiston, ID 2018 TBD
RAT233 Rathdrum, ID 2019 472
SPI12F1 Northport, WA (Spirit) 2019 115
SPR761 Sprague, WA 2019 106
F&C12F1 Spokane, WA (Francis & Cedar) 2019 260
MIS431 Kellogg, ID 2023 257
Total 1,930

SC_PR_3-2 Attachment A Page 113 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
9. Generation Resource Options

Introduction
Several generating resource options are available to meet future resource deficits. Avista
can upgrade existing resources, build new facilities, or contract with other energy
companies to meet its load obligations. This section describes resources Avista
considered in the 2017 IRP to meet future needs. They mostly are generic, as actual
resources identified through a competitive process may differ in size, cost, and operating
characteristics due to siting, engineering, or financial requirements.

Assumptions
Avista models only commercially available resources with well-known costs, availability,
and generation profiles priced as if Avista developed and owned the generation.
Resource options include natural gas-fired combined cycle combustion turbines (CCCT),
simple cycle combustion turbines (SCCT), natural gas-fired reciprocating engines, large-
scale onshore wind, energy storage, photovoltaic solar, hydroelectric upgrades, and
thermal unit upgrades. Several other resource options described later in the chapter are
not included in the PRS analysis, but discussed as potential resource options to respond
to a future resource acquisition. The IRP excludes potential contractual arrangements
with other energy companies as an option in the plan, but such arrangements may
actually offer a lower customer cost when a competitive acquisition process is completed.

The costs of each resource option include the transmission expenses described in
Chapter 8 – Transmission & Distribution Planning. Levelized costs result from discounting
nominal cash flows by a 6.46 percent-weighted average cost of capital approved by the
Idaho and Washington Commissions in recent rate case filings. All costs in this section
are in 2018 nominal dollars unless otherwise noted.

Many renewable resources are eligible for federal and state tax incentives. Federal solar
tax benefits begin to reduce beginning in 2020; federal production tax credits (PTCs) are
no longer available unless meeting certain provisions. Incentives, to the extent they are
available, are included in IRP modeling.
Section Highlights



Upgrades to Avista’s


SC_PR_3-2 Attachment A Page 114 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Avista relies on several sources including the NPCC, press releases, regulatory filings,
internal analysis, developer estimates, and Avista’s experience with certain technologies
for its resource assumptions. The natural gas-fired plants use operating characteristics
and cost information obtained from Thermoflow design software.

Levelized resource costs illustrate the differences between generator types. The values
show the cost of energy if the plants generate electricity during all available hours of the
year. In reality, plants do not operate to their maximum generating potential because of
market and system conditions. Costs are separated between energy in $/MWh, and
capacity in $/kW-year, to better compare technologies1. Without this separation of costs,
resources operating very infrequently during peak-load periods would appear more
expensive than base-load CCCTs, even though peaking resources are lower cost when
operating only a few hours each year. By allowing the expected costs to be divided by the
expected amount of energy deliveries, levelized energy costs fairly compare non-
dispatchable renewable resources to the energy component of natural gas-fired
resources because renewable technologies are typically not dispatchable. It is more
difficult to estimate levelized costs for dispatchable resources because the amount of
MWh to levelize the costs over is debatable, such as its potential energy or economic
dispatch.

The levelized cost calculations include the following cost items for both the capacity and
energy cost components.

 Capital Recovery and Taxes: Depreciation, return of and on capital, federal and
state income taxes, property taxes, insurance, and miscellaneous charges such
as uncollectible accounts and state taxes for each of these items pertaining to a
generation asset investment.
 Allowance for Funds Used During Construction (AFUDC): The cost of money
associated with construction payments made on a generation asset during
construction.
 Federal Tax Incentives: The federal tax incentive in the form of a PTC, or
investment tax credit (ITC), available to qualified generation.
 Fuel Costs: The average cost of fuel such as natural gas, coal, or wood per MWh
of generation. Additional fuel price details are included in the Market Analysis
section.
 Fuel Transport: The cost to transport fuel to the plant, including pipeline capacity
charges.
 Fixed Operations and Maintenance (O&M): Costs related to operating the plant
such as labor, parts, and other maintenance services not based on production
levels.
 Variable O&M: Costs per MWh related to incremental generation.

1 Storage technologies use a $ per kWh rather than $ per kW due to the resource is both energy and
capacity limited.
SC_PR_3-2 Attachment A Page 115 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
 Transmission: Includes depreciation, return on capital, income taxes, property
taxes, insurance, and miscellaneous charges such as uncollectible accounts and
state taxes for each of these items pertaining to transmission asset investments
needed to interconnect the generator and/or third party transmission charges.
Further information regarding interconnection cost are in Chapter 8.
 Other Overheads: Includes miscellaneous charges for non-capital expenses such
as un-collectibles, excise taxes, and commission fees.

Tables at the end of this section show incremental capacity, heat rates, generation capital
costs, fixed O&M, variable costs, and peak credits for each resource option.2 Table 9.1
compares the levelized costs of different resource types over a 30-year asset life.

Table 9.1: Natural Gas-Fired Plant Levelized Costs per MWh

Advanced Large Frame CT $54 $156 220
Modern Large Frame CT $53 $154 186
Advanced Small Frame CT $60 $142 102
Frame/Aero Hybrid CT $43 $154 106
Small Reciprocating Engine Facility $38 $230 47
Modern Small Frame CT $55 $174 49
Aero CT $50 $187 45
1 on 1 Advanced CCCT $35 $230 362
1 on 1 Modern CCCT $34 $233 306

Natural Gas-Fired Combined Cycle Combustion Turbine
Natural gas-fired CCCT plants provide reliable capacity and energy for a relatively modest
capital investment. The main disadvantage of a CCCT is generation cost volatility due to
reliance on natural gas, unless utilizing hedged fuel prices. CCCTs modeled in the IRP
are “one-on-one” (1x1) configurations, using hybrid air/water cooling technology and zero
liquid discharge. The 1x1 configuration consists of a single gas turbine with a heat
recovery steam generator (HRSG) and a duct burner to gain more generation from the
steam turbine. The plants have nameplate ratings between 250 MW and 350 MW each
depending on configuration and location. A two-on-one (2x1) CCCT plant configuration is
possible with two turbines and one HRSG, generating up to 650 MW. Avista would need
to share a 2 x 1 plant to take advantage of the modest economies of scale and efficiency
of a 2x1-plant configuration due to its large size relative to Avista’s needs.

Cooling technology is a major cost driver for CCCTs. Depending on water availability,
lower-cost wet cooling technology could be an option, similar to Avista’s Coyote Springs
2 plant. However, if no water rights are available, a more capital-intensive and less
efficient air-cooled technology may be used. For this IRP, Avista assumes water is

2 Peak credit is the amount of capacity a resource contributes at the time of system one hour peak load.
SC_PR_3-2 Attachment A Page 116 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
available for plant cooling based on its internal analysis, but only enough for a hybrid
system utilizing the benefits of combined evaporative and convective technologies.

This IRP models two types of CCCT plants, first a smaller 285 MW machine, and a larger
advanced 341 MW plant. Avista reviewed many CCCT technologies and sizes, and
selected these plants due to their use in the Northwest. If Avista pursues a CCCT, a
competitive acquisition process will allow analysis of other CCCT technologies and sizes.
The most likely location is in Idaho, mainly due to Idaho’s lack of an excise tax on natural
gas consumed for power generation, a lower sales tax rate relative to Washington, and
no state taxes on the emission of carbon dioxide.3 CCCT site or sites likely would be on
or near our transmission system to avoid third-party wheeling costs. Another advantage
of siting a CCCT resource in Avista’s Idaho service territory is access to relatively low-
cost natural gas on the GTN pipeline.

The smaller CCCT’s heat rate is 6,720 Btu/kWh in 2016.4 The larger machine is 6,631
Btu/kWh. The plants include duct firing for seven percent of rated capacity at a heat rate
of 7,912 and 7,843 Btu/kWh, respectively.

The IRP includes a three percent forced outage rate for CCCTs and 14 days of annual
plant maintenance. The smaller plant can back down to 62 percent of nameplate capacity,
while the larger plant can ramp down to 30 percent of nameplate capacity. The maximum
capability of each plant is highly dependent on ambient temperature and plant elevation.
The plan assumes a 30-year life absent capital upgrades for life extension.

The anticipated capital costs for the two CCCTs, located in Idaho on Avista’s transmission
system with AFUDC on a greenfield site, are $1,174 per kW for the smaller machine and
$1,122 per kW for the larger machine. These estimates exclude the cost of transmission
and interconnection. Table 9.1 shows levelized plant cost assumptions split between
capacity and energy. The costs include firm natural gas transportation, fixed and variable
O&M, and transmission. Table 9.2 summarizes key cost and operating components of
natural gas-fired resource options.

Natural Gas-Fired Peakers
Natural gas-fired SCCTs and reciprocating engines, or peaking resources, provide low-
cost capacity and are capable of providing energy as needed. Technological advances
and their simpler design relative to CCCT plants allow them to start and ramp quickly,
providing regulation services and reserves for load following and variable resources
integration.

The IRP models frame, hybrid-intercooled, reciprocating engines, and aero-derivative
peaking resource options. The peaking technologies have different load following abilities,
costs, generating capabilities, and energy-conversion efficiencies. Table 9.2 shows cost

3 Washington state applies an excise tax on all fuel consumed for wholesale power generation, the same
as it does for retail natural gas service, at approximately 3.875 percent. Washington also has higher sales
taxes and has carbon dioxide mitigation fees for new plants.
4 Heat rates shown are the higher heating value.
SC_PR_3-2 Attachment A Page 117 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
and operational characteristics based on internal engineering estimates. All peaking
plants assume 0.5 percent annual real dollar cost decrease and forced outage and
maintenance rates. The levelized cost for each of the technologies is in Table 9.1.

Table 9.2: Natural Gas-Fired Plant Cost and Operational Characteristics

Advanced Large
Frame CT
$654 $2.19 9,931 $3.73 1 203 203 $133
Modern Large
Frame CT
$684 $2.19 10,007 $2.67 1 170 170 $117
Advanced Small
Frame CT
$875 $3.28 11,265 $2.67 1 96 96 $84
Frame/Aero
Hybrid CT
$1,042 $3.28 8,916 $3.20 1 101 101 $105
Small
Reciprocating
Engine Facility
$1,229 $8.76 7,700 $3.20 5 9.3 47 $57
Modern Small
Frame CT
$1,349 $4.38 10,252 $2.67 1 45 45 $61
Aero CT $1,349 $6.57 9,359 $2.67 1 42 42 $57
1 x 1 Modern
CCCT
$1,148 $19.71 6,771 $4.00 1 341 341 $392
1 x 1 Advanced
CCCT
$1,207 $16.42 6,845 $3.20 1 286 286 $345

Firm natural gas fuel transportation is an electric reliability issue with FERC and the
subject of regional and extra-regional forums. For this IRP, Avista continues to assume it
will not procure firm natural gas transportation for peaking resources. Firm transportation
could be necessary where pipeline capacity becomes scarce during utility peak hours.
However, pipelines near evaluated sites are not presently full or expected to become full
in the near future. Where non-firm transportation options become inadequate for system
reliability, four options exist: contracting for firm natural gas transportation rights,
purchasing an option to exercise the rights of another firm natural gas transportation
customer during times of peak demand, on-site fuel oil, and liquefied natural gas storage.

Wind Generation
Governments promote wind generation with tax benefits, renewable portfolio standards,
carbon emission restrictions, and stricter controls on existing non-renewable resources.
In the Consolidated Appropriations Act 2016, HR 2029, section 301, passed December
2016, the U.S. Congress extended the PTC for wind through December 31, 2016, with
provisions allowing projects to qualify for a prorated credit after 2016 if commencing
construction prior to 2019. For projects commencing construction in 2017, the PTC is
SC_PR_3-2 Attachment A Page 118 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
reduced by 20 percent, 2018 is reduced by 40 percent, and 2019 reduced by 60 percent.
This IRP does not assume the PTC extends beyond this term, but does assume
preferential five-year tax depreciation remains.

Wind resources benefit from having no emissions profile or fuel costs, but they are not
typically dispatchable. On shore wind’s capital costs in 2018, including AFUDC, are
$1,798 per kW for Washington projects and $1,636 per kW in Montana, with annual fixed
O&M costs of $42.70 per kW-yr. Fixed O&M includes indirect charges to account for the
inherent variation in wind generation, oftentimes referred to as wind integration. The cost
of wind integration depends on the penetration of wind in Avista’s balancing authority and
the market price of power. Wind integration in this IRP is $4.40 per kW-year in 2018.
These estimates come from Avista’s experience in the market and results from Avista’s
2007 Wind Integration Study.

Wind capacity factors in the Northwest range between 25 and 40 percent depending on
location. This plan assumes Northwest wind has a 37 percent average capacity factor. A
statistical method, based on regional wind studies, derives a range of annual capacity
factors depending on the wind regime in each year (see stochastic modeling assumptions
for details). The expected capacity factor affects the levelized cost of a wind project. For
example, a 30 percent capacity factor site could be $30 per MWh higher than a 40 percent
capacity factor site holding all other assumptions equal.

As discussed above, levelized costs change substantially due to capacity factor, but can
change more from tax incentives. Figure 9.1 shows nominal levelized prices with different
start dates, capacity factors, and availability of the ITC. For a plant installed in 2018 with
utility ownership, the estimated “all-in” cost is $72 per MWh for 25 years, including the 20
percent REC apprenticeship adder for the EIA. Qualification for the adder requires 15
percent of construction labor by state-certified apprentices. It is possible for third party to
Independent Power Producers to develop a project at a lower cost for the PPA, depending
on turbine agreements, site conditions, and cost of capital. Typical PPA prices do not
include integration or transmission, and may reflect a different cost recovery period. If
Avista plans to acquire new wind generation, an RFP will help identify the least cost option
to meet customer needs.

This IRP includes analysis on wind projects located in Montana. Based on Avista’s
analysis, construction cost will be lower due to the absence of state sales tax and
indications of higher quality wind speeds. Sites in Montana will require third party
transmission wheeling. Adding Montana wind will be less costly to integrate due to its
different generation profile as compared to Palouse Wind, and it may add up to a 7.5
percent capacity contribution when combined with Palouse Wind’s expected output on to
meet the single-hour winter peak. For summer, the plan assumes the combined resources
would add 3 percent of its capability. Montana wind, with transmission to deliver it to
Avista’s system, costs $83 per MWh as compared to $72 per MWh with the same capacity
factor in the Northwest.

SC_PR_3-2 Attachment A Page 119 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Figure 9.1: Northwest Wind Project Levelized Costs per MWh

Photovoltaic Solar
Photovoltaic (PV) solar generation technology costs have fallen substantially in the last
several years partly due to low-cost imports and from demand driven by renewable
portfolio standards and tax incentives. Even with large cost reductions, IRP analyses
shows PV solar facilities still are uneconomic for winter-peaking utilities in the Northwest
compared to other renewable and non-renewable generation options. This is due to its
low capacity factor and lack of output during winter-peak periods. PV solar provides
predictable daytime generation complementing the loads of summer-peaking utilities,
though panels typically do not produce at full output during peak hours.

Adding a substantial amount of PV solar to a summer peaking utility system reduces the
peak hour recorded prior to the installation, but the peak hour shifts toward sundown when
PV solar output is lower. As more PV solar enters a system, the on-peak resource
contribution falls precipitously. Table 9.3 presents the peak credit by month with different
amounts of solar using output from the Rathdrum Solar Project. This table illustrates how
solar does not reduce Avista’s winter peak, reduces the summer peak, and is less
effective at reducing peak with additional solar installations.

Solar-thermal technologies can produce capacity factors as much as 30 percent higher
than PV solar projects and can store several hours of energy for later use in reducing
peak loads. However, solar thermal technologies do not lend themselves well to the
Northwest due to their lack of significant generation in the winter and higher overall
installation and operation costs; therefore, only PV solar systems are considered for this
IRP.
72
84
102
111
121
72 74
84
91
98
60 61
70 75
81
$0
$20
$40
$60
$80
$100
$120
$140
2018 2020 2025 2030 2035
No
m
i
n
a
l

L
e
v
e
l
i
z
e
d

$
/
M
W
h
Base Case Full PTC Full PTC + 40% CF
SC_PR_3-2 Attachment A Page 120 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Table 9.3: Solar Capacity Credit by Month

Month 5 MW 25 MW 50 MW 100 MW 150 MW 200 MW 300 MW
Jan 0% 0% 0% 0% 0% 0% 0%
Feb 0% 0% 0% 0% 0% 0% 0%
Mar 0% 0% 0% 0% 0% 0% 0%
Apr 28% 15% 11% 8% 6% 5% 3%
May 46% 46% 37% 26% 17% 13% 9%
Jun 39% 39% 36% 31% 25% 22% 19%
Jul 52% 49% 45% 43% 33% 27% 22%
Aug 40% 40% 40% 34% 32% 30% 24%
Sep 0% 0% 0% 0% 0% 0% 0%
Oct 0% 0% 0% 0% 0% 0% 0%
Nov 0% 0% 0% 0% 0% 0% 0%
Dec 0% 0% 0% 0% 0% 0% 0%

Utility-scale PV solar capital costs including AFUDC for a 50 MW (DC) system are $1,110
per kW for fixed panel and $1,165 per kW for single-axis tracking projects. A well-placed
utility-scale single-axis tracking PV system located in the Pacific Northwest would achieve
a first-year capacity factor of approximately 18 percent and a fixed panel system would
achieve 15 percent. PV solar output degrades over time; the IRP de-rates solar
generation output by one-half percent each year. The federal government’s 30 percent
tax credit begin phasing out after 2019. Projects starting construction in 2020 have a 26
percent ITC, 22 percent for 2021 projects, and 10 percent for any projects afterward.

Figure 9.2 shows the levelized costs of PV solar resources, including applicable federal
and state incentives, on-line dates, and capacity factors. Like wind projects, independent
power producers may have lower costs than utilities due to panel agreements, cost of
capital and the ability to using federal incentives to directly lower upfront costs, rather
than amortizing tax credits over the life of the asset. The costs in Figure 9.2 show the
price advantage of IPP development as far as transferring benefits from the ITC directly
to customers. IRP modeling in this IRP assumes the ITC would be a credit to the cost of
the project rather than amortized over the life of the asset.

The State of Washington offers a number of incentives for solar installations. Plants less
than five megawatts count double toward Washington’s EIA. The state also offers
substantial financial incentives for consumer-owned solar. Consumer-owned solar counts
in reductions in Avista’s retail load forecast.

SC_PR_3-2 Attachment A Page 121 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Figure 9.2: Solar Nominal Levelized Cost ($/MWh)

Energy Storage
Increasing solar and wind generation makes energy storage technologies attractive from
an operational perspective. Storage could smooth out renewable generation variability,
absorb oversupply, and assist in load following and regulation needs. The technology
could help meet peak demand, provide voltage support, relieve transmission congestion,
take power during oversupply events, and supply other non-energy needs for the system.
The IRP considered several storage technologies, including pumped hydroelectric, lead-
acid batteries, lithium-ion batteries, vanadium flow batteries, flywheels, compressed air,
liquefied air, and gravity systems. For modeling purposes, the IRP uses two plant types:
a 1x3-storage facility and a 1x6. Meaning, for each MW of capacity, it has three or six
MWh of storage.

Modeling each storage technology would not provide additional insight as a comparison
to other supply options because Avista’s capacity needs are not urgent, the technology
is changing rapidly, and each has different losses, lifespan and flexibility. Modeling of
storage’s non-power supply benefits is still in development. Although Avista is attempting
to estimate as many of these values as possible. For example, Chapter 8 discusses the
methodology to estimate the value of deferred distribution capital investment. The IRP
includes a value for market arbitrage and providing ancillary services such as regulation,
spinning, and non-spinning reserves. Avista is developing an evaluation for estimating
the storage benefit for network services such as reliability, voltage support and frequency
response (not all storage options can provide this service). Each of these benefits are
$0
$20
$40
$60
$80
$100
$120
2018 2020 2025 2030 2035
No
m
i
n
a
l

L
e
v
e
l
i
z
e
d

$
/
M
W
h
Tracked Base Case Fixed Base Case
Tracked 30% ITC Fixed 30% ITC
Tracked 30% ITC (IPP Method)Fixed 30% ITC (IPP Method)
SC_PR_3-2 Attachment A Page 122 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
part of the Clean Energy Funds/PNNL partnership to estimate values for storage. A report
will be available in the spring of 2018.

Storage may become an important part of the nation’s electricity grid if the technology
overcomes a number of physical, technical, and economic barriers. First, existing
technologies consume a significant amount of electricity relative to their output through
conversion losses. Second, equipment costs are still high, but falling, at nearly three times
the initial cost of a natural gas-fired peaking plant. Peaking plants provide many of the
same capabilities without the electricity consumption characteristics of storage. Storage
costs will decline over time and Avista will monitor the technologies as part of the IRP
process. Third, the current scale of most storage projects is relatively small, limiting their
applicability to utility-scale deployment.

Avista installed a vanadium flow battery in Pullman, Washington to learn more about
storage technology. The Turner Energy Storage Project provides insight about the
technology’s reliability, potential benefit to the transmission and/or distribution systems,
and potential power supply benefits including oversupply events. The battery has 1.2
megawatts of power capability and 3.5 megawatt-hours of energy storage. A Washington
State research and development grant partially funded this project.

Turner Energy Storage Project, Pullman, WA

As part of the Clean Energy Funds 2 grants, Avista proposes to develop two additional
storage projects in the University District of downtown Spokane. One 500 kW project with
two MWh of storage and the other project 100 kW with 0.5 MWh of storage. At the time
of this IRP’s drafting these projects are out to bid and expected to begin operation in late
2018.

The Northwest might be slower in adopting storage technology relative to other regions
in the country. The Northwest hydroelectric system already contains a significant amount
of storage relative to the rest of the country. However, as more capacity consuming
renewables enter the electric grid, new storage technologies might play a significant role
SC_PR_3-2 Attachment A Page 123 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
in meeting the need for additional operational flexibility if upfront capital costs and
operational losses continue to fall.

In addition to capital costs, storage project O&M costs are $20 per kWh-year levelized,
and recharge costs use off-peak Mid-Columbia energy prices. Levelized storage project
costs are inaccurate as storage projects do not create megawatt hours; in fact, they
consume megawatt hours with 15 to 20 percent or more of their charge being lost. Avista’s
experience with vanadium flow storage has losses from 30 to 50 percent. This IRP
assumes 17 percent losses over its 20 year expected life. Storage costs are typically
shown in $/kWh due to the energy limitation of the project rather than $ per kW. The
capital cost in 2018 dollars including AFUDC is $713 per kWh for the 1x3 project and
$642 for the 1x6 project. By 2025, the costs fall to $573 and $516 per kWh respectively.

Other Generation Resource Options
Many resources were not specifically included as resource options in this IRP. These
resources include biomass, geothermal, co-generation, nuclear, offshore wind, landfill
gas, and anaerobic digesters. This plan does not model these resource options explicitly,
but continues to monitor their availability; cost and operating characteristics to determine
if state policies change or the technology becomes more economically available.

Exclusion from the PRS does not necessarily exclude non-modeled technologies from
Avista’s future portfolio. The non-modeled resources can compete with resources
identified in the PRS through competitive acquisition processes. Competitive acquisition
processes identify technologies to displace resources otherwise included in the IRP
strategy. Another possibility is acquisition through PURPA mandates. PURPA provides
developers the ability to sell qualifying power to Avista at set prices and terms.5

Woody Biomass Generation
Woody biomass generation projects use waste wood from lumber mills or forest
management. In the generation process, a turbine converts boiler-created steam into
electricity. A substantial amount of wood fuel is required for utility-scale generation.
Avista’s 50 MW Kettle Falls Generation Station consumes over 350,000 tons of wood
waste annually, or 48 semi-truck loads of wood chips per day. It typically takes 1.5 tons
of wood to make one megawatt-hour of electricity; the ratio varies with the moisture
content of the fuel. The viability of another Avista biomass project depends on the
availability and cost of the fuel supply. Many announced biomass projects fail due to lack
of a long-term fuel source. If an RFP identifies a potential woody biomass project, Avista
will consider it for a future resource.

Geothermal Generation
Geothermal energy provides predictable capacity and energy with minimal carbon dioxide
emissions (zero to 200 pounds per MWh). Some forms of geothermal technology extract
steam from underground sources to run through power turbines on the surface while
others utilize an available hot water source to power an Organic Rankine Cycle
installation. Due to the geologic conditions of Avista’s service territory, no geothermal

5 Rates, terms, and conditions are available at www.avistautilities.com under Schedule 62.
SC_PR_3-2 Attachment A Page 124 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
projects are likely to develop. Geothermal energy struggles to compete economically due
to high development costs stemming from having to drill several holes thousands of feet
below the earth’s crust; each hole can cost over $3 million. Ongoing geothermal costs are
low, but the capital required locating and proving a viable site is significant. Further, there
are no good geothermal resource sites in or near Avista’s service territory or transmission
system.

Landfill Gas Generation
Landfill gas projects generally use reciprocating engines to burn methane gas collected
at landfills. The Northwest has developed many landfill gas resources. The costs of a
landfill gas project depend on the site specifics of a landfill. The Spokane area had a
project on one of its landfills, but was retired after the fuel source depleted to an
unsustainable level. Much of the Spokane area no longer landfills its waste and instead
uses the Spokane Waste to Energy Plant. Nearby in Kootenai County, Idaho, the
Kootenai Electric Cooperative developed the 3.2 MW Fighting Creek Project. Using
publically available costs and the NPCC estimates, landfill gas resources are
economically promising, but are limited in their size, quantity, and location. Further, due
to falling wholesale market pricing, many landfills are considering cleaning the gas to
create pipeline quality gas. This form of renewable gas has become an option for natural
gas utilities to offer a renewable gas alternative.

Anaerobic Digesters (Manure or Wastewater Treatment)
The number of anaerobic digesters is increasing in the Northwest. These plants typically
capture methane from agricultural waste, such as manure or plant residuals, and burn the
gas in reciprocating engines to power generators. These facilities tend to be significantly
smaller than utility-scale generation projects, at less than five megawatts. Most facilities
are located at large dairies and cattle feedlots. A survey of Avista’s service territory found
no large-scale livestock operations capable of implementing this technology.

Wastewater treatment facilities can host anaerobic digesting technology. Digesters
installed when a facility is initially constructed helps the economics of a project greatly,
though costs range greatly depending on system configuration. Retrofits to existing
wastewater treatment facilities are possible, but tend to have higher costs. Many projects
offset energy needs of the facility, so there may be little, if any, surplus generation
capability. Avista currently has a 260 kW wastewater system under a PURPA contract
with a Spokane County facility. Anaerobic digesters may opt to clean the gas to make to
pipeline quality to offer a clean gas alternative.

Small Cogeneration
Avista has few industrial customers with loads significantly large enough to support a
cogeneration project. If an interested customer was inclined to develop a small
cogeneration project, it could provide benefits including reduced transmission and
distribution losses, shared fuel, capital, and emissions costs, and credit toward
Washington’s EIA efficiency targets.

SC_PR_3-2 Attachment A Page 125 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Another potentially promising option is natural gas pipeline cogeneration. This technology
uses waste-heat from large natural gas pipeline compressor stations. In Avista’s service
territory few compressor stations exist, but the existing compressors in our service
territory have potential for this generation technology. Avista has discussed adding
cogeneration with pipeline owners, but no project has been determined feasible.

A big challenge in developing any new cogeneration project is aligning the needs of the
cogenerator with the utility need for power. The optimal time to add cogeneration is during
the retrofit of an industrial process, but the retrofit may not occur when the utility needs
new capacity. Another challenge to cogeneration within an IRP is estimating costs when
host operations drive costs for a particular project. The best method for the utility to
acquire this technology is through the PURPA process.

Nuclear
Avista does not include nuclear plants as a resource option in the IRP given the
uncertainty of their economics, regional political issues with the technology, U.S. nuclear
waste handling policies, and Avista’s modest needs relative to the size of modern nuclear
plants. Nuclear resources could be in Avista’s future only if other utilities in the Western
Interconnect incorporate nuclear power in their resource mix and offer Avista an
ownership share or if cost effective small-scale nuclear plants become commercially
available.

The viability of nuclear power could change as national policy priorities focus attention on
decarbonizing the nation’s energy supply. The limited amount of recent nuclear
construction experience in the U.S. makes estimating construction costs difficult. Cost
projections in the IRP are from industry studies, recent nuclear plant license proposals,
and the small number of projects currently under development. Modular nuclear design
could increase the potential for nuclear generation by shortening the permitting and
construction phase, and making these traditionally large projects a better fit the needs of
smaller utilities.

Offshore Wind
Avista does not include offshore wind resources in this IRP due to the current availability
of onshore wind resource options with lower prices and without third party transmission
services. Offshore wind is a proven technology outside of the US, so far only one project
is operational in the U.S. Avista will continue to monitor this technology as its cost and
efficiency change.

Coal
The coal generation industry is at a crossroads. In many states, like Washington, new
coal-fired plants are extremely unlikely due to emission performance standards and the
shortage of utility scale carbon capture and storage projects. Federal guidelines regarding
coal are uncertain given the current EPA administration’s review of section 111(b) of the
CAA and the CPP. The risks associated with future carbon legislation and projected low
natural gas costs make investments in this technology highly unlikely.

SC_PR_3-2 Attachment A Page 126 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Hydroelectric Project Upgrades and Options
Avista continues to upgrade its hydroelectric facilities. The latest hydroelectric upgrade
added ten megawatts to the Nine Mile Falls Development in 2016. Figure 9.3 shows the
history of upgrades to Avista’s hydroelectric system. Avista added 46.8 aMW of
incremental hydroelectric energy between 1992 and 2016. Upgrades completed after
1999 can qualify for the EIA, thereby reducing the need for additional renewable energy
options.

Figure 9.3: Historical and Planned Hydro Upgrades

Construction of the Spokane River hydroelectric project occurred in the late 1800s and
early 1900s, when the priority was to meet then-current loads. The developments
therefore do not capture a majority of river flows. In 2012, Avista reassessed its Spokane
River Project to evaluate opportunities to capture more of the streamflow. The goal was
to develop a long-term strategy and prioritize potential facility upgrades. Avista evaluated
five of the six Spokane River developments and estimated costs for generation upgrade
options. Each upgrade option should qualify for the EIA renewable energy goal. These
studies were part of the 2011 and 2013 IRP Action Plans and results appear below. Each
of these upgrades are major engineering projects, taking several years to complete and
requiring major changes to the FERC licenses and project water rights. Table 9.4
summarizes the upgrade options. The upgrades will compete against other renewable
options when more renewables are required.

0
10
20
30
40
50
0
2
4
6
8
10
19
9
2

-

M
o
n
r
o
e

S
t
r
e
e
t

U
n
i
t

1
19
9
4

-

N
i
n
e

M
i
l
e

U
n
i
t
s

3

&

4
19
9
4

-

C
a
b
i
n
e
t

U
n
i
t

1
19
9
4

-

L
o
n
g

L
a
k
e

U
n
i
t

4
19
9
4

-

L
i
t
t
l
e

F
a
l
l
s

U
n
i
t

3
19
9
6

-

L
o
n
g

L
a
k
e

U
n
i
t

1
19
9
7

-

L
o
n
g

L
a
k
e

U
n
i
t

2
19
9
9

-

L
o
n
g

L
a
k
e

U
n
i
t

3
20
0
1

-

C
a
b
i
n
e
t

U
n
i
t

3
20
0
1

-

L
i
t
t
l
e

F
a
l
l
s

U
n
i
t

4
20
0
4

-

C
a
b
i
n
e
t

U
n
i
t

2
20
0
7

-

C
a
b
i
n
e
t

U
n
i
t

4
20
0
9

-

N
o
x
o
n

U
n
i
t

1
20
1
0

-

N
o
x
o
n

U
n
i
t

2
20
1
1

-

N
o
x
o
n

U
n
i
t

3
20
1
2

-

N
o
x
o
n

U
n
i
t

4
20
1
6

-

N
i
n
e

M
i
l
e

U
n
i
t
s

1

&

2
Cu
m
u
l
a
t
i
v
e

A
v
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
SC_PR_3-2 Attachment A Page 127 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
At the time of this IRP, the company is developing a long-term strategy for Post Falls. The
current scope of the project is to replace the current generating equipment with newer
technology. Part of this IRP’s Action Plan will be to report on the redevelopment plan.

Table 9.4: Hydroelectric Upgrade Options

Resource Monroe
Street/Upper
Falls
Long
Lake
Cabinet
Gorge
Incremental Capacity (MW) 80 68 110
Incremental Energy (MWh) 237,352 202,592 161,571
Incremental Energy (aMW) 27.1 23.1 9.2
Peak Credit (Winter/ Summer) 31/0 100/100 0/0
Capital Cost ($2018 Millions) $196 $182 $290
Levelized Energy Cost ($2018/MWh) $93 $122 $200

Long Lake Second Powerhouse
Avista studied adding a second powerhouse at Long Lake over 30 years ago by using
the small arch or saddle dam located on the south end of the project site. This project
would be a major undertaking and require several years to complete, including major
changes to the Spokane River license and water rights. In addition to providing customers
with a clean energy source, this project could help reduce total dissolved gas levels by
reducing spill at the project and providing incremental capacity to meet peak load growth.

The 2012 study considered three alternatives. The first replaces the existing four-unit
powerhouse with four larger units totaling 120 MW, increasing capacity by 32 MW. The
other two alternatives develop a second powerhouse with a penstock beginning from a
new intake structure downstream of the existing saddle dam. One powerhouse option
was a single 68 MW turbine project. The second was a two-unit 152 MW project. The best
alternative in the study was the single 68 MW option. Table 9.4 shows upgrade costs and
characteristics.

Monroe Street/Upper Falls Second Power House
Avista replaced the powerhouse at its Monroe Street development on the Spokane River
in 1992. There are three options to increase its capacity. Each would be a major
undertaking requiring substantial cooperation with the City of Spokane to mitigate
disruption in Riverfront and Huntington parks and downtown Spokane during
construction. The upgrade could increase plant capacity by up to 80 MW. To minimize
impacts on the downtown area and the park, a tunnel drilled on the east side of Canada
Island could avoid excavation of the south channel. A smaller option would add a second
40 MW Upper Falls powerhouse, but this option would require south channel excavation.
A final option would add a second Monroe Street powerhouse for 44 MW.

Cabinet Gorge Second Powerhouse
Avista is exploring the addition of a second powerhouse at the Cabinet Gorge
development site to mitigate total dissolved gas and produce additional electricity. A new
SC_PR_3-2 Attachment A Page 128 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
110 MW underground powerhouse would benefit from an existing diversion tunnel around
the dam built during original construction. This resource does not add any peak capacity
credit due to the water right limitations of the license. The resource only creates additional
energy during spring runoff.

Thermal Resource Upgrade Options
The 2015 IRP identified several thermal upgrade options for Avista’s fleet. Some options,
such as the Cold Day Controls and Advanced Hot Gas Path at Coyote Springs 2, are
already in service. This plan contains new ideas to increase generating capability at
Avista’s thermal generating resources. No costs are presented in this section, as pricing
is sensitive to third-party suppliers.

Northeast CT Water Injection
This is a water injected NOx control system allowing the firing temperature to increase
and thereby increasing the capacity at the Northeast CT by 7.5 MW.

Rathdrum CT Supplemental Compression
Supplemental compression is a new technology developed by PowerPhase LLC that
increases airflow through a combustion turbine compressor increasing machine output.
This upgrade could increase Rathdrum CT capacity by 24 MW.

Rathdrum CT 2055 Uprates
By upgrading certain combustion and turbine components, the firing temperature can
increase to 2,055 degrees from 2,020 degrees corresponding to a five MW increase in
output.

Rathdrum CT Inlet Evaporation
Installing a new inlet evaporation system will increase the Rathdrum CT capacity by 17
MW on a peak summer day, but no additional energy is expected during winter months.

Kettle Falls Turbine Generator Upgrade
The Kettle Falls plant began operation in 1983. In 2025, the generator and turbine will be
42 years old and at the end of its expected life. At this time, Avista could spend additional
capital and upgrade the unit by five megawatts rather than replace it with in-kind
technology.

Kettle Falls Fuel Stabilization
The wood burned at Kettle Falls varies in moisture content, and dryer fuel burns more
efficiently. A fuel drying system added to the fuel handling system would allow the boiler
to operate at a higher efficiency point, increasing plant capability by three megawatts.

SC_PR_3-2 Attachment A Page 129 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Ancillary Services Valuation
IRPs traditionally model the value of resources using hourly models. This method
provides a good approximation of resource value, but it does not provide a value for the
intra-hour or ancillary services needs of a balancing area. Ancillary services modeled in
the IRP include spinning and non-spinning reserves, regulation, and load following.
Spinning and non-spinning reserve obligations together equal three percent of load and
three percent of on-line generation, as required by regional standards. Half of the
reserves must synchronize to the system and half must be capable of synchronizing
within ten minutes. Regulation meets instantaneous changes in load or resources with
plants responding to the change using automatic generating control. Load following
covers load changes within the hour, but for movements occurring across a timeframe
greater than ten minutes.

Avista developed a new tool, called the Avista Decision Support System (ADSS), for use
in operations and long-term planning. This model is a mixed-integer linear program
simulating Avista’s system. It optimizes a set of resources to meet system load and
ancillary services requirements using real-time information. The tool uses both actual and
forecasted information regarding the surrounding market and operating conditions to
provide dispatch decisions, but can also use historical data to simulate benefits of certain
system changes. ADSS uses historical data sets to estimate ancillary services values for
storage and natural gas-fired resources.

Storage
As intermittent resources grow in size, there is potential for the existing system not being
robust enough to integrate the resources and handle oversupply of renewable energy. To
address this concern, governments and utilities are promoting and investing in storage
technology. Today storage has a limited role due to cost and technology development.
This analysis uses the study competed for the 2015 IRP to determine the potential
financial value storage brings to Avista’s power supply costs. The study includes several
storage capacities with storage to peak ratio of three to one and 85 percent efficiency.
Table 9.5 shows the values brought to the power supply system for each storage capacity
size. These values are to the Avista system only and do not represent the value to other
systems or non-power supply benefits. Avista has a deep resource stack of flexible
resources and adding additional flexible resources do not necessarily add value unless
sold to third parties.

The values shown in Table 9.5 include margin from several value streams including
operating reserves, regulation, load following, and arbitrage. Arbitrage optimizes the
battery to charge in low price periods and discharge when prices are higher. Of the values
shown in Table 9.5, arbitrage represents the largest value stream. Figure 9.4 shows the
five value streams for power supply benefits. Load following and arbitrage represent 92
percent of the value to Avista.

SC_PR_3-2 Attachment A Page 130 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Table 9.4: Storage Power Supply Value

Storage
Capacity
(MW)
Annual
Value
Annual $/kW
Value
35 $1,201,590 $34
30 $1,024,569 $34
25 $923,291 $37
10 $381,407 $38
5 $189,000 $38
1 $36,862 $37

Figure 9.4: Storage’s Value Stream

Natural Gas-Fired Facilities
Natural gas-fired facilities can provide energy and ancillary services. This study looks at
their incremental ancillary services value to the system as prepared for the 2015 IRP. The
values do not represent the value for current resources of similar technology, but only the
incremental value of a new facility. This study assumes 100 MW resource increments in
2020. Table 9.6 shows the results of the analysis. The incremental values for these
resources are marginal due to the limited need for these types of services. The study
assumes each facility has different operating capabilities. For example, diesel back-up
can only provide non-spin reserves as it is for emergency use only, while the LMS 100
may provide non-spinning reserves, spinning reserves, regulation, and load following if
operating.
Arbitrage, 64%
Load Following,
28%
Spin & Non-Spin
Reserves, 5%
Regulation, 2%
SC_PR_3-2 Attachment A Page 131 of 205
Chapter 9- Generation Resource Options
Avista Corp 2017 Electric IRP
Table 9.5: Natural Gas-Fired Facilities Ancillary Service Value

Resource Type Capabilities Annual $/kW
Value
CCCT Load Following/ Spin6, Regulation $0.00
LMS 100 Load Following/ Spin, Non-Spin/ Regulation $1.12
Reciprocating Engines Load Following/Spin/Non-Spin $0.61
Diesel Back-Up Non-Spin $0.00

An action item from this IRP is to determine the intra hour valuation of these services for
both storage and natural gas-fired peakers using historical data closer to the 2019 IRP
release date and implementing new modeling techniques including intra hour modeling.
Avista’s DSS model at the time of the IRP is not capable of intra hour modeling, but it is
in process of adding this functionality.

6 Fast start CCCTs may have some non-spin reserve capability.
SC_PR_3-2 Attachment A Page 132 of 205

SC_PR_3-2 Attachment A Page 133 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
10. Market Analysis

Introduction
This section describes the electricity, natural gas, and other markets studied in the 2017
IRP. It contains price risks Avista considers to meet customer demands at the lowest
reasonable cost. The analytical foundation for the 2017 IRP is a fundamentals-based
electricity model of the entire Western Interconnect. The market analysis evaluates
potential resource options on their net value within the wholesale marketplace, rather
than the summation of their installation, operation, maintenance and fuel costs. The
Preferred Resource Strategy (PRS) analysis uses these net market values to select
future resource portfolios.

Understanding market conditions in the Western Interconnect is important because
regional markets are highly correlated due to large transmission linkages between load
centers. This IRP builds on prior analytical work by maintaining the relationships
between the sub-markets within the Western Interconnect and the changing energy
market values of company-owned and contracted-for resources. The backbone of the
analysis is an electricity market model. The model, AURORAXMP, emulates the dispatch
of resources to serve loads across the Western Interconnect given fuel prices,
hydroelectric conditions, and transmission and resource constraints. The model’s
primary outputs are electricity prices at key market hubs (e.g., Mid-Columbia) and
resource dispatch including the resources costs, market value and greenhouse gas
emissions.

Marketplace
AURORAXMP is a fundamentals-based modeling tool used by Avista to simulate the
Western Interconnect electricity market. The Western Interconnect includes states west
of the Rocky Mountains, the Canadian provinces of British Columbia and Alberta, and
the Baja region of Mexico as shown in Figure 10.1. The modeled area has an installed
resource base of approximately 240,000 MW.
Section Highlights






SC_PR_3-2 Attachment A Page 134 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.1: NERC Interconnection Map

The Western Interconnect is separate from the Eastern and ERCOT interconnects to
the east except for eight DC inverter stations. It follows operation and reliability
guidelines administered by the Western Electricity Coordinating Council (WECC). Avista
modeled the WECC electric system as 17 zones based on load concentrations and
transmission constraints. After extensive study in prior IRPs, Avista models the
Northwest region as a single zone because this configuration dispatches resources in a
manner consistent with historical operations. Table 10.1 describes the specific zones
modeled in this IRP.

Table 10.1: AURORAXMP Zones

Northwest- OR/WA/ID/MT Southern Idaho
COB- OR/CA Border Wyoming
Eastern Montana Southern California
Northern California Arizona
Central California New Mexico
Colorado Alberta
British Columbia South Nevada
North Nevada Baja, Mexico
Utah

Western Interconnect Loads
The 2017 IRP relies on a load forecast for each zone of the Western Interconnect.
Avista uses utility resource plans and regional plans to quantify load growth across the
west. These plans include estimates regarding energy efficiency, customer-owned
generation, plug-in electric vehicles and demand response reductions. Forecasting
future energy use is difficult because of large uncertainties with the long-term drivers of
future energy use.

SC_PR_3-2 Attachment A Page 135 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.2 shows regional load growth estimates. The total of the forecasts show
Western Interconnect loads rising nearly 0.85 percent annually over the next 20 years.
On a regional basis, the Northwest grows at 0.77 percent, California at 0.25 percent,
and the Rocky Mountain States at 1.63 percent. Canada is 1.5 percent. From a system
reliability perspective, regional peak loads grow at similar levels.

Figure 10.2: 20-Year Annual Average Western Interconnect Energy

Resource Retirements
The resource mix constantly changes as new resources start generating and older
resources retire. In prior IRPs, much of the existing fleet continued to serve loads in
combination with new resources. Many companies are now choosing to retire older
plants to comply with environmental regulations and economic changes. Most plant
closures are once-through-cooling (OTC) facilities in California and older coal
technology throughout North America.

Several states are developing rules to restrict or eliminate certain generation
technologies. In California, all OTC facilities require retrofitting to eliminate OTC
technology or the plant must retire. Over 14,200 MW of OTC natural gas-fired
generators in California likely will retire and need replacement in the IRP timeframe. The
IRP assumes the closure of OTC plants with identified shutdown dates from their utility
owners’ IRPs and announcements. Elimination of OTC plants in California will eliminate
older technology presently used for reserves and high demand hours. Replacement
plants will be expensive for California customers, but they will have a more modern,
efficient and flexible generation fleet.

Coal-fired facilities face increasing regulatory scrutiny. In the Northwest, the Boardman
and Centralia coal plants will retire by the end of calendar years 2020 and 2025
California
Northwest
Desert SW
Rocky Mountains
Canada
aGW
20 aGW
40 aGW
60 aGW
80 aGW
100 aGW
120 aGW
140 aGW
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
SC_PR_3-2 Attachment A Page 136 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
respectively. Recently Colstrip 1 & 2 announced closure by 2022, for a reduction total of
about 2,621 megawatts. Other coal-fired plants throughout the Western Interconnect
have announced plant closures, including Four Corners, Carbon, Arapahoe, San Juan,
Reid Gardner, Dave Johnson, North Valmy, and Intermountain. The Nevada legislature
successfully placed into law a plan to retire all in-state coal plants, and other utilities
appear poised to retire many plants as indicated in recent IRPs. Over the next 20 years,
roughly 43 percent of the US Western Interconnection coal fleet retires in the Expected
Case. In total, announced retirements for all generation technologies, as shown in
Figure 10.3, equal approximately 25 gigawatts between 2017 and 2037. Avista did not
forecast additional coal retirements beyond official announcements prior to development
of the Expected Case.

Figure 10.3: Resource Retirements (Nameplate Capacity)

New Resource Additions
New resource capacity is required to meet load growth and replace retiring power plants
over the next 20 years. The generation additions meet capacity, energy, ancillary
services and Renewable Portfolio Standards (RPS). Only natural gas-fired peaking and
CCCT plants, storage, solar, and wind facilities are in the plan. The IRP does not
include new nuclear or coal plants over the forecast horizon. The model objective is to
meet capacity and renewable energy targets, but actual resources constructed may
differ.

Many states have RPS requirements promoting renewable generation to reduce
greenhouse gas emissions, provide jobs, and diversify energy mixes. RPS legislation
generally requires utilities to meet a portion of their load with qualified renewable
resources. No federal RPS mandate exists presently; therefore, each state defines RPS
obligations differently. AURORAXMP now models RPS levels explicitly. The RPS
GW
5 GW
10 GW
15 GW
20 GW
25 GW
30 GW
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Nuclear
Oil
Coal
Natural Gas
SC_PR_3-2 Attachment A Page 137 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
requirements are loaded into the model and the model selects resources to satisfy state
laws. Figure 10.4 illustrates new capacity and RPS additions made in the modeling
process. Nearly 98 GW will be required to meet the renewable and capacity
requirements for the US system. Wind and solar facilities meet most renewable energy
requirements.

Geothermal, biomass, and hydroelectric resources provide limited RPS contributions;
given their large range in costs and availability, these resources are not included in the
capacity expansion study. Due to its low capacity factor, large quantities of solar
capacity are necessary to make a meaningful contribution.

Figure 10.4: Cumulative WECC Generation Resource Additions (Nameplate Capacity)

In total, 61,000 MW of new utility and consumer-owned renewable generation will
pressure afternoon peak pricing lower and move peak load requirements later in the
day. Potential for oversupply in shoulder months in California will increase imports to the
Northwest and other markets. The largest resource additions expected in the west are
solar and natural gas-fired generation. Solar is the largest driver of new resource
additions due to RPS requirements and the reduction in costs compared to alternative
renewable resources. Most natural gas-fired technology will be peakers to provide a low
cost flexible capacity to balance intermittent power generation and not burden
customers with high capacity costs. Given the large amount of future renewables on the
system, wholesale power prices will remain low and costs of larger baseload plants built
to meet peak capacity requirements will be difficult to extract from the wholesale market,
placing a burden on utility ratepayers or independent power producer’s shareholders.
Based on these market fundamentals and the requirement to have a reliable system
where peakers rather than combined cycle plants will play a larger role in the future.

GW
20 GW
40 GW
60 GW
80 GW
100 GW
120 GW
140 GW
160 GW
GW
2 GW
4 GW
6 GW
8 GW
10 GW
12 GW
14 GW
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
To
t
a
l
Cu
m
u
l
a
t
i
v
e

Storage Wind
Net Meter Solar
Natural Gas Cumulative
SC_PR_3-2 Attachment A Page 138 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
A new entrant into the resource forecast is storage technology. At the time of the IRP
analysis, the capacity expansion model cannot model the economic additions of
storage; current storage additions for the most part either are mandated or pilot
projects. This forecast cautiously includes 5,500 MW of new storage capacity over the
20-year period. Given the changes in storage costs and policy, Avista will continue to
monitor this technology to determine if a larger level of market penetration is likely.

The Northwest market needs new capacity resources in the 2021/22 period. This study
includes nearly 7,000 MW of new natural gas-fired generation to meet load growth and
replace retiring resources across the four Northwest states. As for renewable
requirements, new generation will continue to consist of wind, but Avista expects
movement to solar as costs decrease allowing solar to grow at a greater pace than wind
energy. Table 10.2 shows the amount of new renewables added to the Northwest by the
end of 2037. Also included in this analysis, is consumer driven renewables. These
additions, amounting to one percent of load meet customer demand for renewables as
part of a utility’s renewable energy offerings.

Table 10.2: Added Northwest Renewable Generation Resources

Resource Type Capacity (MW)
Wind 4,100
Utility- Solar 4,800
Customer- Solar 1,922

Fuel Prices and Conditions
Fuel cost and availability are some of the most important drivers of the wholesale
electricity marketplace and resource values. Some resources, including geothermal and
biomass, have limited fuel options or sources, while natural gas has greater potential.
Hydroelectric, wind, and solar resources benefit from free fuel, but are highly dependent
on weather and siting opportunities.

Natural Gas
The natural gas industry continues its fundamental shift towards hydraulic fracturing and
shale resources. New methods and technology continue to increase efficiency and
production from wells. Over the next 25 years, demand in the residential, commercial,
and industrial natural gas markets should slightly decline. At the same time, exports to
Mexico and for LNG will ramp up as demand for natural gas-fired generation in Mexico
and completion of LNG plants materialize.

Natural gas used for power generation is growing due to its ability to support variable
output from renewable energy and as a replacement for coal plant retirements. The fuel
of choice for new base-load and peaking generation continues to be natural gas.
Natural gas has a history of significant price volatility, generally attributed to weather
related demand and supply issues. The long-term forecasted supply for natural gas
shows the average daily supply will increase to meet new demand through 2050.

SC_PR_3-2 Attachment A Page 139 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Avista uses forward market prices and a forecast from a prominent energy industry
consultant to develop the natural gas price forecast for this IRP. Based on these
forecasts, the levelized nominal price is $4.20 per dekatherm (Dth) at Henry Hub
(shown in Figure 10.5 as the green bars). The pricing methodology used to create a
fundamental price forecast follows:

 2018-2019: 100 percent market;
 2020: 75 percent market, 25 percent consultant;
 2021: 50 percent market, 50 percent consultant;
 2022: 25 percent market, 75 percent consultant; and
 2023-2037: 100 percent consultant.

Figure 10.5: Henry Hub Natural Gas Price Forecast

Price differences across North America depend on demand at the major trading hubs
and pipeline constraints existing between them. Table 10.3 presents western natural
gas basin differentials from Henry Hub prices. Prices converge over the course of the
study as new pipelines and sources of natural gas materialize. To illustrate the
seasonality of natural gas prices, monthly Stanfield price shapes are in Table 10.4 for
selected forecast years.

$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
IRP Forecast
Consultant
Forwards (12/12/16)
SC_PR_3-2 Attachment A Page 140 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Table 10.3: Natural Gas Price Basin Differentials from Henry Hub

Basin 2018 2020 2025 2030 2035
Stanfield 93% 93% 96% 97% 100%
Malin 96% 96% 97% 99% 101%
Sumas 90% 89% 92% 97% 100%
AECO 73% 74% 84% 91% 92%
Rockies 95% 94% 96% 97% 99%
Southern CA 102% 103% 103% 102% 103%

Table 10.4: Monthly Price Differentials for Stanfield from Henry Hub

Month 2018 2020 2025 2030 2035
Jan 97% 95% 97% 98% 102%
Feb 97% 95% 97% 98% 102%
Mar 93% 94% 96% 98% 100%
Apr 91% 94% 96% 97% 99%
May 91% 91% 95% 96% 99%
Jun 91% 91% 94% 96% 98%
Jul 92% 91% 94% 95% 98%
Aug 92% 93% 95% 96% 98%
Sep 93% 94% 95% 98% 99%
Oct 92% 94% 96% 97% 100%
Nov 95% 95% 97% 99% 102%
Dec 97% 95% 97% 98% 101%

Coal
This IRP assumes no new coal plants in the Western Interconnect, but models existing
plants as part of the electric system unless scheduled for retirement. Existing coal
facilities typically have medium to long-term fuel contracts in place and many have ties
to oil prices due to transportation costs. These contracts are not publically available. For
each coal plant, Avista uses publically available coal prices filed with FERC, and then
uses an average annual price increase over the IRP timeframe of 1.2 percent for railed
coal and 1.4 percent for mine mouth coal based on data from the Energy Information
Administration1. For Colstrip Units 3 and 4, Avista used escalation rates based on
expectations from existing and expectations of future contracts.

Hydroelectric
The Northwest U.S., British Columbia and California have substantial hydroelectric
generation capacity. A favorable characteristic of hydroelectric power is its ability to
provide near-instantaneous generation up to and potentially beyond its nameplate
rating. Hydro is valuable for meeting peak load, following general intra-day load trends,
storing and shaping energy for sale during higher-valued hours, and integrating variable

1 Energy Information Administration’s Annual Energy Outlook 2016, reference case.
SC_PR_3-2 Attachment A Page 141 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
generation resources. The key drawback to hydroelectric generation is its variable and
limited fuel supply.

This IRP uses an 80-year hydroelectric data record from the 2014 BPA rate case. The
study provides monthly energy levels for the region over an 80-year hydrological record
spanning 1928 to 2009.

Many IRP studies use an average of the hydroelectric record, whereas stochastic
studies randomly draw from the record, as the historical distribution of hydroelectric
generation is not normally distributed. Avista does both. Figure 10.6 shows the average
hydroelectric energy of 15,720 aMW in the northwest, defined here as Washington,
Oregon, Idaho and western Montana. The chart also shows the range in potential
energy used in the stochastic study, with a 10th percentile water year of 12,489 aMW (-
21 percent) and a 90th percentile water year of 18,586 aMW (+18 percent).

AURORAXMP maps each hydroelectric plant to a load zone, creating a similar energy
shape for all plants in the load zone. For Avista’s hydroelectric plants, AURORAXMP
uses the output from its own proprietary software with a more accurate representation of
operating characteristics and capabilities. AURORAXMP represents hydroelectric plants
using annual and monthly capacity factors, minimum and maximum generation levels,
and sustained peaking generation capabilities. The model’s objective, subject to
constraints, is to shift hydroelectric generation into peak load hours to maximize the
value of the system consistent with actual operations.

Figure 10.6: Northwest Expected Energy

Wind
New wind resources satisfy a significant share of western renewable portfolio standards
over the IRP timeframe. These additions increase competition for the remaining higher-
quality wind sites. Similar to how AURORAXMP maps each hydroelectric plant to a load
0%
2%
4%
6%
8%
10%
12%
12
,
0
0
0
12
,
5
0
0
13
,
0
0
0
13
,
5
0
0
14
,
0
0
0
14
,
5
0
0
15
,
0
0
0
15
,
5
0
0
16
,
0
0
0
16
,
5
0
0
17
,
0
0
0
17
,
5
0
0
18
,
0
0
0
18
,
5
0
0
19
,
0
0
0
19
,
5
0
0
20
,
0
0
0
20
,
5
0
0
21
,
0
0
0
21
,
5
0
0
Pe
r
c
e
n
t

o
f

H
y
d
r
o

D
r
a
w
s
Average Megawatts
SC_PR_3-2 Attachment A Page 142 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
zone, the capacity factors in Figure 10.7 are averages for each zone. The IRP uses
capacity factors from a review of the BPA and the National Renewable Energy
Laboratory (NREL) wind data sets.

Figure 10.7: Regional Wind Expected Capacity Factors

Greenhouse Gas Emissions and the Clean Power Plan
Greenhouse gas, or carbon emissions, regulation is a significant uncertainty for the
electricity industry because of reliance on carbon-emitting generation and the potential
of regulation to increase wholesale prices. Regulation may require the reduction of
carbon emissions at existing power plants, the construction of low- and non-carbon-
emitting technologies, and for changes to existing operations. Between 2008 and 2015,
western states carbon emissions from generation dropped nearly 13 percent due to
reduced loads and less coal generation.

Future carbon emissions could fall due to fundamental market changes or regulation. In
2014, the EPA released the draft Clean Power Plan (CPP) under section 111(d) of the
CAA to reduce emissions from existing plants. With a new Federal Administration, the
future of regulation under 111(d) may change; at the same time, state-level emission
reduction policies may move forward. Washington’s Clean Air Rule (CAR) caps
emissions for facilities emitting more than 100,000 metric tons per year, and reduces
the emissions threshold by 5,000 metric tons per year, until covering all entities emitting
over 70,000 metric tons by 2035. The Washington Commission may implement rules
regarding RCW 70.235, from the Executive Order 07-02. Other states, such as Oregon,
are also considering carbon emissions limitations at the state legislature. Without final
or specific rules and regulations, modeling the impact of future policies is difficult, but
this plan includes specific assumptions. Due to uncertainty and the likelihood of
31%33%35%
31%
37%38%
29%28%
32%
0%
10%
20%
30%
40%
50%
NW BC AB CA MT WY SW UT CO
Ca
p
a
c
i
t
y

F
a
c
t
o
r
SC_PR_3-2 Attachment A Page 143 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
greenhouse gas regulations, this IRP used the CPP goals to guide the development of
the emission reduction forecast of this IRP.

For the Expected Case, the CAR limits plant level emissions in Washington. The Clean
Air Rule identifies specific reductions to plants over a glide path by 2035. As an
alternative to reductions, emission credits or RECs from Washington State may satisfy
compliance obligations. The CAR monitors compliance at three-year intervals.
Washington State may generate up to the cap each year based on the three-year
average generation between 2012 and 2015. Each year the cap declines. For covered
plants, the total allowance is for the group rather than the individual facility providing for
allowance trading. The Department of Ecology intends to set baseline emission levels
and reduction targets for new plants covered under the CAR.

The Oregon emissions policies, beyond the requirements in SB 1547 ending the use of
coal to serve Oregon loads by 2030 and an increased RPS reaching 50 percent by
2040, are in development at the writing of this IRP. However, emissions are not likely to
increase long term. This IRP assumes emissions fall by 30 percent compared to 2015
by 2025. The IRP assumes Idaho emissions follow the CPP emission intensity goals.
Additional details about the state-level emissions reductions programs are in Chapter 7-
Policy Considerations.

For the other states, outside of the current programs in California, carbon emissions will
likely fall under federal policies. The current form of the CPP used to develop this IRP is
not likely to remain in force under the current Federal Administration, but some form of
regulation may replace it. The EPA sent information regarding CPP intent to the Office
of Management and Budget on March 8, 2017, but had publically not released any
proposal. This will require additional review and analysis in the next IRP. To consider
this future affect to our facilities, Montana reduces electric generation carbon emissions
following existing CPP targets with new source complement, but the start of this effort is
delayed four years. For the remaining western states, an emission reduction goal is in
place allowing each of the states to trade between each other based on the CPP target
with new sources, but delayed until 2024.

This IRP does not include specific carbon pricing except for states and provinces with
existing carbon trading and tax regulations. By modeling emission goals and
constraints, the model estimates potential emission trading prices for each ton of
emissions. This methodology is in line with current policy discussions using cap and
trade markets rather than taxes or fees. Any future tax or price policy will require
alternative analysis in a later IRP. Avista uses a different carbon reduction methodology
in this IRP than in its prior plans. In this IRP, the model forces reductions in emissions
and the model estimates the shadow price of the emission reduction. Past IRPs used an
arbitrary carbon price not tied to a specific reduction level. Arbitrary carbon prices
without a correlation to market fundamentals may not achieve desired emissions
reductions. Without a tie to external factors, a “tax or fee” may not achieve a specific
emission goal due to changing external factors such as natural gas prices or
hydroelectric conditions. For example, a higher carbon price is required to reduce
SC_PR_3-2 Attachment A Page 144 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
emissions when natural gas prices are high or hydroelectric conditions are unfavorable,
as such, a lower carbon price will reduce emissions in a low natural gas price
environment or favorable hydroelectric conditions. This phenomenon is shown later in
this chapter regarding the Washington Clean Air Rule. Avista will monitor policy
directives regarding greenhouse gas emissions to determine if the methodology is
consistent with future policy objectives.

Risk Analysis
A stochastic analysis, using the variables discussed earlier in this chapter, evaluates the
market to account for future uncertainty. It is better to represent the electricity price
forecast as a range because point estimates are unlikely to reflect underlying
assumptions perfectly. Stochastic price forecasts develop more robust resource
strategies by accounting for tail risk. The IRP uses 500 distinct 20-year market futures,
providing a large distribution of the marketplace illustrating potential tail risk outcomes.
The next several pages discuss the input variables driving market prices, and describe
the methodology and the range in inputs used in the modeling process.

Natural Gas
Natural gas prices are a volatile commodity in relation to its historical prices. Daily
Stanfield prices ranged between $1.21 and $24.36 per Dth between 2004 and 2017.
Figure 10.8 shows average Stanfield monthly prices since January 2004. Prices
retreated from 2008 highs to a monthly price of $1.44 per Dth in March 2016. Prices
since 2009 are lower than the previous five years, but continue to show volatility.

Figure 10.8: Historical Stanfield Natural Gas Prices (2004-2015)

$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
$12/Dth
1/
1
/
2
0
0
4
8/
1
/
2
0
0
4
3/
1
/
2
0
0
5
10
/
1
/
2
0
0
5
5/
1
/
2
0
0
6
12
/
1
/
2
0
0
6
7/
1
/
2
0
0
7
2/
1
/
2
0
0
8
9/
1
/
2
0
0
8
4/
1
/
2
0
0
9
11
/
1
/
2
0
0
9
6/
1
/
2
0
1
0
1/
1
/
2
0
1
1
8/
1
/
2
0
1
1
3/
1
/
2
0
1
2
10
/
1
/
2
0
1
2
5/
1
/
2
0
1
3
12
/
1
/
2
0
1
3
7/
1
/
2
0
1
4
2/
1
/
2
0
1
5
9/
1
/
2
0
1
5
4/
1
/
2
0
1
6
11
/
1
/
2
0
1
6
SC_PR_3-2 Attachment A Page 145 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.9 shows Stanfield natural gas price duration curves for 2018, 2025 and 2035.
The chart illustrates a larger price range in the later years of the study, reflecting less
forecast certainty. Shorter-term prices are more certain due to additional market
information and the quantity of near term natural gas trading. Figure 10.10 shows
another view of the forecast. The mean price in 2018 is $2.80 per Dth, represented by
the horizontal bar, and the levelized price over the 20 years is $4.21 per Dth. The
bottom and top of the bars represent the 10th and 90th percentiles. The bar length
indicates price uncertainty. Figure 10.11 illustrates the difference in pricing between the
deterministic case and the mean and median of the 500 simulations. On a levelized
basis, the median and deterministic cases are $4.00 and $4.03 per Dth, while the mean
is higher at $4.20 per Dth2. Due to the methodology of the stochastic model, the mean
is greater than both the median and the starting deterministic. The model randomizes
prices based on the lognormal distribution of the change in the deterministic monthly
price forecast. Given a lognormal distribution, the mean prices trend higher than the
median prices given the skewed distribution curve.

Figure 10.9: Stanfield Annual Average Natural Gas Price Distribution

2 The 20-year levelized mean price at Henry Hub is $4.35 per dekatherm.
0
100
200
300
400
500
$/
D
t
h
$2
/
D
t
h
$4
/
D
t
h
$6
/
D
t
h
$8
/
D
t
h
$1
0
/
D
t
h
$1
2
/
D
t
h
$1
4
/
D
t
h
$1
6
/
D
t
h
$1
8
/
D
t
h
$2
0
/
D
t
h
It
e
r
a
t
i
o
n
2018 2025 2035
SC_PR_3-2 Attachment A Page 146 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.10: Stanfield Natural Gas Distributions

Figure 10.11: Stanfield Natural Gas Annual Price Statistical Comparison

$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
$12/Dth
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
20
1
8
-
3
7
$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
$12/Dth
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Mean
Deterministic
Median
SC_PR_3-2 Attachment A Page 147 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Regional Load Variation
Several factors drive load variability. The largest short-run driver is weather. Long-run
economic conditions, like the Great Recession, tend to have a larger impact on the load
forecast. IRP loads increase on average at the levels discussed earlier in this chapter,
but risk analyses emulate varying weather conditions and base load impacts.

Avista continues with its previous practice of modeling load variation using FERC Form
714 data from 2007 to 2015 for the Western Interconnect as the basis for its analysis.
Correlations between the Northwest and other Western Interconnect load areas
represent how electricity demand changes together across the system. This method
avoids oversimplifying Western Interconnect loads. Absent the use of correlations,
stochastic models may offset changes in one variable with changes in another, virtually
eliminating the possibility of broader excursions witnessed by the electricity grid. The
additional accuracy from modeling loads this way is crucial for understanding wholesale
electricity market price variation. It is vital for understanding the value of peaking
resources and their use in meeting system variation.

Tables 10.5 and 10.6 present load correlations for the 2017 IRP. Statistics are relative
to the Northwest load area (Oregon, Washington and Idaho). “NotSig” indicates no
statistically valid correlation existed in the data. “Mix” indicates the relationship was not
consistent across the 2007 to 2015 period. For regions and periods with NotSig and Mix
results, the IRP does not model correlations between the regions. Tables 10.7 and 10.8
provide the coefficient of determination values by zone.3

Table 10.5: January through June Load Area Correlations

Area Jan Feb Mar Apr May Jun
Alberta Mix Mix Mix Mix Not Sig 20%
Arizona 32% 38% Mix Not Sig Mix Not Sig
Avista 88% 86% 78% 77% 41% 79%
British Columbia 86% 88% 72% 73% 41% 61%
California Not Sig Not Sig Mix Mix 17% Not Sig
CO-UT-WY -23% Mix Mix -26% -3% -18%
Montana 55% 65% 63% 52% Mix 46%
New Mexico 6% 6% Mix Mix Mix Mix
North Nevada 58% 22% 6% Mix Mix 51%
South Idaho 79% 76% 69% Mix Mix 49%
South Nevada 52% 42% Mix Not Sig Mix 19%

3 The coefficient of determination is the standard deviation divided by the average.
SC_PR_3-2 Attachment A Page 148 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Table 10.6: July through December Load Area Correlations

Area Jul Aug Sep Oct Nov Dec
Alberta 6% Not Sig Not Sig Not Sig 12% Mix
Arizona Mix Mix Mix -21% Mix 27%
Avista 76% 78% 67% 79% 92% 92%
British Columbia 73% 56% 23% 75% 87% 83%
California Not Sig Not Sig Not Sig -12% Mix Not Sig
CO-UT-WY -2% Mix -2% -12% 26% Mix
Montana 6% 17% 6% 20% 79% 75%
New Mexico Not Sig Mix Mix Not Sig 34% 18%
North Nevada 52% 53% 27% Mix 60% 34%
South Idaho 29% 38% 32% 6% 87% 84%
South Nevada Mix 6% Mix -33% Mix 64%

Table 10.7: Area Load Coefficient of Determination (Standard Deviation/Mean)

Area Jan Feb Mar Apr May Jun
Alberta 4.9% 4.3% 4.8% 4.5% 4.9% 5.5%
Arizona 8.2% 7.2% 7.2% 10.8% 15.1% 16.2%
Avista 8.9% 8.5% 9.6% 8.7% 8.5% 10.3%
British Columbia 8.5% 7.9% 8.5% 8.0% 8.3% 8.6%
California 9.3% 9.3% 9.4% 9.9% 11.4% 12.6%
CO-UT-WY 7.8% 7.7% 7.9% 7.5% 8.7% 13.2%
Montana 7.8% 7.1% 7.7% 7.1% 7.3% 9.6%
New Mexico 8.3% 8.4% 8.0% 9.5% 13.0% 13.6%
Northern Nevada 5.6% 5.6% 5.6% 6.4% 6.0% 9.4%
Pacific Northwest 9.7% 9.2% 9.4% 8.7% 8.4% 8.9%
South Idaho 8.6% 8.2% 8.8% 9.8% 11.0% 14.9%
South Nevada 6.5% 5.8% 6.3% 11.5% 17.1% 18.3%

Table 10.8: Area Load Coefficient of Determination (Standard Deviation/Mean)

Area Jul Aug Sep Oct Nov Dec
Alberta 5.8% 5.5% 5.8% 4.6% 5.0% 4.8%
Arizona 14.0% 14.4% 15.6% 13.2% 7.5% 7.8%
Avista 12.7% 12.4% 9.8% 8.8% 11.1% 9.9%
British Columbia 9.5% 9.4% 8.8% 8.9% 10.5% 9.2%
California 13.1% 13.8% 14.6% 11.7% 9.9% 9.7%
CO-UT-WY 12.8% 12.4% 11.4% 8.3% 8.6% 8.4%
Montana 9.8% 10.1% 8.1% 7.2% 8.6% 8.1%
New Mexico 12.8% 12.5% 13.8% 10.8% 9.1% 8.8%
Northern Nevada 10.0% 9.3% 8.7% 5.9% 6.2% 6.5%
Pacific Northwest 10.6% 10.5% 9.2% 9.0% 11.7% 10.9%
South Idaho 11.4% 12.2% 12.8% 8.6% 10.6% 9.4%
South Nevada 15.7% 15.7% 17.8% 13.0% 6.8% 7.1%

SC_PR_3-2 Attachment A Page 149 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Hydroelectric Variation
Hydroelectric generation is the most commonly modeled stochastic variable in the
Northwest because historically it has a larger impact on regional electricity prices than
other variables. The IRP uses an 80-year hydroelectric record starting with the 12-
month water year beginning October 1, 1928. Every iteration starts with a randomly
drawn water year from the historical record, so each water year repeats approximately
125 times in the study (500 scenarios x 20 years / 80 water year records).

Wind Variation
Wind has the most volatile short-term generation profile of any utility-scale resource.
This makes it necessary to capture wind volatility in the power supply model to
determine the value of non-wind resources able to follow loads when wind production
varies. Accurately modeling wind resources requires hourly and intra-hour generation
shapes. For regional market modeling, the representation is similar to how AURORAXMP
models hydroelectric resources. A single wind generation shape represents all wind
resources in each load area. This shape is smoother than an individual wind plant, but
closely represents the diversity of a large number of wind farms located across a zone.

This simplified wind methodology works well for forecasting electricity prices across a
large market, but does not accurately represent the volatility of specific wind resources
Avista might select as part of its PRS. Therefore individual wind farm shapes form the
basis of wind resource options for Avista.

Fifteen potential 8,760-hour annual wind shapes represent each geographic region or
facility. Each year contains a wind shape drawn from these 15 representations. The IRP
relies on two data sources for the wind shapes. The first is BPA balancing area wind
data. The second is NREL-modeled data between 2004 and 2006.

Avista believes an accurate representation of a wind shape across the West requires
data meeting several conditions:

1. Data correlated between areas using historical data.
2. Data within load areas is auto-correlated.4
3. The average and standard deviation of each load area’s wind capacity factor is
consistent with the expected amount of energy for a particular area in the year
and month.
4. The relationship between on- and off-peak wind energy is consistent with historic
wind conditions. For example, more energy in off-peak hours than on-peak hours
where this has been experienced historically.
5. Hourly capacity factors for a diversified wind region are never greater than 90
percent due to turbine outages and wind diversity within the area.
Absent these conditions, it is unlikely any wind study provides a level of accuracy
adequate for planning efforts. Avista’s methodology, first developed for its 2013 IRP,

4 Adjoining hours or groups of hours correlate to each other.
SC_PR_3-2 Attachment A Page 150 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
attempts to meet the five conditions by first using a regression model based on historic
data for each region. The independent variables used in the analysis were month, night
or day hour type, and generation levels from the prior two hours. To reflect correlation
between regions, a capacity factor adjustment reflects historic regional correlation using
an assumed normal distribution with the historic correlation as the mean. After this
adjustment, a capacity factor adjustment accounts for hours with generation levels
exceeding a 90 percent capacity factor. Figure 10.12 shows a Northwest example of an
8,760-hour wind generation profile. This example, shown in blue, has a 31 percent
capacity factor. Figure 10.13 shows actual 2016-wind generation recorded by BPA
Transmission. The average wind fleet in BPA’s balancing authority had a 27.3 percent
capacity factor in 2016.

Figure 10.12: Wind Model Output for the Northwest Region

Forced Outages
Most deterministic market modeling represents generator-forced outages with an
average reduction to maximum capability. This over simplification represents expected
values well; however, it is better to represent the system more accurately in stochastic
modeling by randomly placing non-hydroelectric units out of service based on a mean
time to repair and on an average forced outage rate. Internal studies show this level of
modeling detail is necessary only for natural gas-fired, coal and nuclear plants with
generating capacities in excess of 100 MW. Plants under 100 MW on forced outage do
not materially affect market prices so their outages do not require stochastic modeling.
Forced outage rates and mean time to repair data for the larger units in the Western
Interconnect come from analyzing the North American Electric Reliability Corporation’s
Generating Availability Data System database, also known as GADS.

0%
20%
40%
60%
80%
100%
1 1,001 2,001 3,001 4,001 5,001 6,001 7,001 8,001
Ca
p
a
c
i
t
y

F
a
c
t
o
r
SC_PR_3-2 Attachment A Page 151 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.13: 2016 Actual Wind Output BPA Balancing Authority5

Market Price Forecast
An optimal resource portfolio cannot ignore the extrinsic value inherent in its resource
choices. To determine extrinsic value, the 2017 IRP simulation compares each
resource’s expected hourly output using forecasted Mid-Columbia hourly prices over
500 iterations of Monte Carlo-style scenario analysis.

Hourly zonal electricity prices are equal to either the operating cost of the marginal unit
in the modeled zone or the economic cost to generate and move power from another
zone to the modeled zone. A forecast of available future resources helps create an
electricity market price projection. The IRP uses regional planning margins to set
minimum capacity requirements rather than simply summing the capacity needs of
regional utilities. This reflects how some regions have resource surpluses even where
individual utilities are deficit. This imbalance can be due in part to ownership of regional
generation by independent power producers and possible differences in planning
methodologies used by utilities in the region.

AURORAXMP assigns market values to each resource alternative available to Avista, but
does not select Avista’s PRS. Several market price forecasts determine the value and
volatility of a resource portfolio. As Avista does not know what will happen in the future,
it relies on risk analysis to help determine an optimal resource strategy. Risk analysis
uses several market price forecasts with different assumptions from the Expected Case
or with changes to the underlying statistics of a study. The modeling splits alternate
cases into stochastic and deterministic studies.

5 Chart data is from the BPA at: http://transmission.bpa.gov/Business/Operations/Wind/default.aspx.
0%
20%
40%
60%
80%
100%
1 1,001 2,001 3,001 4,001 5,001 6,001 7,001 8,001
Ca
p
a
c
i
t
y

F
a
c
t
o
r
SC_PR_3-2 Attachment A Page 152 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP

A stochastic study uses Monte Carlo analysis to quantify the variability in future market
prices, and the resultant impact on individual and portfolios of resources. These
analyses include 500 iterations of varying natural gas prices, loads, hydroelectric
generation, thermal outages, and wind generation shapes. The IRP includes three
stochastic studies—the Expected Case, a case with the social cost of carbon, and a
benchmarking case excluding a cost of carbon.

Mid-Columbia Price Forecast
The Mid-Columbia market is Avista’s primary electricity trading hub. The market is
historically the lowest cost in the west due to the amount of hydroelectric generation at
the hub and its proximity to Canadian gas supplies, though other markets can be less
expensive at times when solar production is high and loads are low.

Fundamentals-based market analysis is critical to understanding the power industry
environment. The Expected Case includes two studies. The first study is a deterministic
market view using expected levels for the key assumptions discussed in the first part of
this chapter. The second is a risk or stochastic study with 500 unique scenarios based
on different underlining assumptions for natural gas prices, load, wind generation,
hydroelectric generation, forced outages, and inflation. Each study simulates the entire
Western Interconnect hourly between 2018 and 2037.

Figure 10.14 shows the Mid-Columbia stochastic market price results with horizontal
bars representing the 10th and 90th percentile range for annual prices, diamonds show
average prices, and arrows represent the 95th percentile. The 20-year nominal levelized
price is $35.85 per MWh. Table 10.9 shows the annual averages of the stochastic case
on-peak, off-peak and levelized prices. Spreads between on- and off-peak prices
average $5.09 per MWh over 20 years. The 2015 IRP annual average nominal price
was $38.48 per MWh. The market price reduction from the 2015 study results from
lower natural gas prices, lower loads, higher percentages of new lower-heat-rate natural
gas plants, and increased solar resources serving higher RPS requirements.

SC_PR_3-2 Attachment A Page 153 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.14: Mid-Columbia Electric Price Forecast Range

Table 10.9: Annual Average Mid-Columbia Electric Prices ($/MWh)

Year Flat Off-
Peak
On-
Peak
2018 23.79 19.48 27.02
2019 23.71 19.53 26.85
2020 23.99 20.16 26.85
2021 24.30 20.88 26.85
2022 25.95 22.59 28.47
2023 29.68 26.30 32.24
2024 32.03 28.90 34.38
2025 32.58 29.83 34.65
2026 34.27 31.77 36.13
2027 37.61 35.43 39.25
2028 40.18 38.28 41.60
2029 44.06 42.44 45.27
2030 46.86 45.15 48.15
2031 48.08 46.42 49.32
2032 51.10 49.17 52.55
2033 52.81 50.83 54.29
2034 55.09 53.07 56.61
2035 57.50 55.14 59.26
2036 60.52 58.24 62.22
2037 64.51 62.09 66.33
Levelized $35.85 $32.94 $38.03
$/MWh
$20/MWh
$40/MWh
$60/MWh
$80/MWh
$100/MWh
$120/MWh
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
20
1
8
-
3
7
SC_PR_3-2 Attachment A Page 154 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Negative Electric Market Prices
The price forecast includes functionality to allow prices to go negative during oversupply
events. In the past, oversupply events mostly occurred during spring periods when
hydro was at high levels and wind was at full capacity. Traditionally these events occur
at night when loads are lower. Given increasing solar penetration, negative pricing is
now occurring during the mid-afternoon. Avista models this by changing the supply
curve of the hydro resources to a negative marginal price. Whenever demand is higher
than hydro resources and must run generation, the marginal price is negative. Without
this change, prices would never go below zero. This change properly values new
resource opportunities such as storage and peaking units, but is also important to avoid
overvaluing solar and other non-dispatchable resources during oversupply events.

Greenhouse Gas Emission Levels
Greenhouse gas levels are declining regionally and nationally as lower-cost natural gas
resources displace coal-fired generation, or even forces coal plants into early
retirement. This IRP includes emissions limits and pricing as described earlier in this
chapter. Figure 10.15 shows historic and expected greenhouse gas emissions for the
Western Interconnect. Greenhouse gas emissions from electricity generation decrease
6.2 percent between 2018 and 2037, and 2018 is 15 percent lower than 2015. The
figure also includes 10th and 90th percentile statistics from the 500-iteration dataset. The
higher and lower bands show emissions depending on changes in hydroelectric
generation, load, resource availability and other factors. Lower load forecasts, lower
natural gas prices, higher RPS requirements, coal-fired generation retirements and
carbon limits drive the reductions. Once the majority of planned coal-fired plant
retirements occur by 2032, emissions rise again reflecting new load met by a mixture of
renewables and natural gas without coal retirements beyond current announcements.

Figure 10.15: Western States Greenhouse Gas Emissions

-
50
100
150
200
250
300
350
19
9
0
19
9
2
19
9
4
19
9
6
19
9
8
20
0
0
20
0
2
20
0
4
20
0
6
20
0
8
20
1
0
20
1
2
20
1
4
20
1
6
20
1
8
20
2
0
20
2
2
20
2
4
20
2
6
20
2
8
20
3
0
20
3
2
20
3
4
20
3
6
Mi
l
l
i
o
n

M
e
t
i
c

T
o
n
s

o
f

C
O
2
Historical
Expected Case
10th Percentile
90th Percentile
SC_PR_3-2 Attachment A Page 155 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.16 illustrates the Expected Case emissions intensity for the Western
Interconnect. The CPP included an option for states to meet intensity goals for covered
plants; this chart illustrates the reductions across the west to get a second look at the
effectiveness of the emission constraints modeled. Between 2018 and 2037, the
emission intensity falls 17.5 percent. Alternatively, Figure 10.17 illustrates the change in
emission intensity from 2018 to 2037 by area. All areas show declining emissions
intensity with the exception of southern Idaho. The Idaho area has few emitting
resources (the region currently imports much of its baseload power) and the added
natural gas increases its intensity. This chart shows the relationship of the emissions
intensity of facilities in the area compared to the area’s load. For example, Wyoming
exports energy as its production is greater than its local load.

Figure 10.16: Emission Intensity Metric

-
100
200
300
400
500
600
700
800
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
lb
s

p
e
r

M
W
h
SC_PR_3-2 Attachment A Page 156 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.17: Instate Emission Intensity Change from 2018 to 2037

Resource Dispatch
State-level RPS goals and greenhouse gas regulations change resource dispatch
decisions and affect future power prices. The Northwest is witnessing the market-
changing effects of more than 7,750 MW of wind. Figure 10.18 illustrates how natural
gas will increase as a percentage of Western Interconnect generation from 29 percent
in 2018 to 37 percent 2037. The increase offsets coal-fired generation, with coal
dropping from 23 percent in 2018 to 13 percent in 2037. Utility-owned solar and wind
generation increase from 11 percent in 2018 to 20 percent by 2037. New renewable
generation also reduce coal-fired generation, but natural gas-fired generation is the
primary resource meeting load growth due to economic dispatch and its addition to
serve peak load growth. Figure 10.19, illustrates the resources meeting the reduction of
coal and nuclear resources, and the increase in load. Natural gas meets 50 percent,
while renewables meet the rest.

Figure 10.18: Base Case Western Interconnect Resource Mix

729
1,238
312
1,244
64
989
563 406
2,449
3,945
601
1,002
271
1,070
203
963
556
258
1,544
3,402
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
US West AZ CA CO S. ID NM NV OWI-Mt UT WY
Ar
e
a

l
b
s

p
e
r

M
W
h
2018
2037
Nuclear
Hydro
Other
Coal
Wind
Solar
Natural Gas
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s

SC_PR_3-2 Attachment A Page 157 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.19: Western Interconnect Resource Mix Changes

Greenhouse Gas Emission Pricing
This IRP assumes the market will have emission caps; with this assumption, the
AURORAXMP model produces emission prices rather than a direct input as past IRPs.
With this new constraint, the model produces a shadow price or hurdle rate for the
plants with emission constraints. The resulting shadow prices as shown in Figure 10.20
affect the dispatch of plants in each area with reduction goals similar to models with a
carbon “price”. For Washington, the prices are near zero (depending on water year) until
the early 2020s and remain below $5 per metric ton until 2030. These prices are a result
of increasing renewables on the system and the type of regulations in place. Avista’s
facilities are not subject to these prices. The Washington projected emissions prices are
lower than the prices required in coal regions as it is affecting natural gas resources
rather than coal facilities. Natural gas prices need a lower financial disincentive to
dispatch compared to coal as natural gas is on the margin most hours, while coal
facilities are not.

In Washington, the emission policy only those plants identified by the Department of
Ecology for the Clean Air Rule have constraints; therefore, the model may find cheaper
ways to serve customers by running regulated plants only to the point of the regulation,
or importing power. The prices shown are for the average price. Prices can be
significantly higher, as shown in Figure 10.21 from the stochastic analysis. If
hydroelectric production is low and there are few alternatives to serve load, then
emissions prices could be significantly higher. Further analysis is required due to the
baseline emissions were not yet available at the time of the analysis. The AURORAXMP
model is not able to produce prices based on a three-year compliance period; these
prices assume a one-year compliance period. These prices do not represent the cost of
compliance of this rule, but rather the implied cost for the electric sector to comply with
the rule on a marginal basis. Non-electric participants subject to the rule could affect
pricing if a future allowance market creates competition for scarce compliance options
or where by building additional renewables driving down wholesale prices.
Natural Gas
53%
Hydro
2%
Solar
30%
Wind
15%
SC_PR_3-2 Attachment A Page 158 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Prices in Oregon are important to Avista, as our Coyote Springs 2 is located there. At
the time of the IRP analysis, Oregon had not identified a specific greenhouse gas policy.
This IRP uses a 30 percent reduction goal from 2015 emissions by 2025. This amount
is 10 percent lower than the Clean Power Plan new source complement mass based
goal. The resulting prices of this assumption are similar to the Washington results, as
the states have similar generation profiles after existing coal-fired facilities retire. In this
state, the average prices increase to approximately $11 per metric ton by 2037. The
resulting Montana prices are significantly higher than the coastal states, as emissions
reductions must come from low marginal cost coal. The average price starts at $6.40
per metric ton in 2024 and escalates to $27 per ton in 2037.6 Coal facilities have lower
base dispatch costs and require a high price to reduce dispatch. These results illustrate
the importance of policy making regarding emission reductions. For Avista, Colstrip is
subject to this price adder for this analysis. This analysis illustrates how placing
emission caps on individual states may drive in-state emissions lower, but will likely
cause increasing imports (or decreasing exports). The analysis also shows lowering
emissions from coal facilities requires higher pricing than areas with natural gas. For the
northwest, a carbon pricing mechanism would be more effective and less burdensome
on customers if it focused on coal rather than all resource types.

Figure 10.20: Northwest Greenhouse Gas Emission Shadow Prices

6 At the 95th percentile, the 2024 price is $17 per metric ton and $60 per metric ton in 2037.
$0
$5
$10
$15
$20
$25
$30
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
$
p
e
r

M
e
t
r
i
c

T
o
n
WA
OR
MT
SC_PR_3-2 Attachment A Page 159 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.21: Washington Clean Air Rule Pricing

Scenario Analysis
Scenario analysis evaluates the impact of changes in underlying market assumptions,
Avista’s generation portfolio, and new generation resource values. In addition to the
Expected Case, this IRP includes two stochastic analyses. The first scenario is the case
where Colstrip retires and the second scenario reduces dispatch at Colstrip to 50
percent of current levels. Both scenarios are required due to the nature of the portfolio
studies they support (as described in Chapter 11).

In past IRPs, several stochastic scenarios reviewed impacts on changes in
environmental policy. These scenarios are important to consider for resource planning,
but given uncertainty in policy, limited time for the analysis, and only minor changes
from the 2015 IRP, these additional scenarios are only indicative until greenhouse gas
policy becomes more certain. Therefore, most of the IRP scenarios focus on Avista’s
portfolio rather than the market. Per the TAC’s request, a deterministic market scenario
simulates how the energy market would change if total emissions decreased 50 percent
by 2035. This is a market scenario only, and not part of the portfolio analysis. It is
informative on the steps Avista’s portfolio would need to take to achieve this goal.

No Colstrip Scenario
The No Colstrip Scenario models the implications of retiring Colstrip Units 3 & 4 early.
The scenario values new resource options and the remaining portfolio in a marketplace
without Colstrip. In addition, this scenario provides data about the regional financial
impacts of a Colstrip closure, rather than just the impact to Avista from divestment of its
share. This scenario assumes 1,000 MW CCCT, 430 MW peakers, and 300 MW wind
replace the units. It does not attempt to represent the feasibility of this assumption, but
rather helps understand the impacts to the overall market place. To simulate all the
portfolio scenarios implications, this market scenario assumes Colstrip retires by the
end of 2023.

$0
$20
$40
$60
$80
$100
$120
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
$
p
e
r

M
e
t
r
i
c

T
o
n
Mean
95th Percentile
99th Percentile
SC_PR_3-2 Attachment A Page 160 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Without Colstrip, regional market prices increase slightly as shown in Figure 10.22.
There are small differences beginning in 2024 with a $0.93 per MWh annual average
price difference, overall prices are 2.7 percent higher without Colstrip. While these price
changes are not large, it assumes the average price over a year in average water
conditions. At times, the price impacts are much greater and without replacement
capacity, price impacts and reliability concerns increase. Beginning in 2024, the annual
production costs to all western customers’ increases by $143 million with the closure of
Colstrip, plus the capital recovery of the additional new resources to replace the
capacity estimated to be $250 million (2023 dollars). Without Colstrip, greenhouse gas
emissions decrease; in 2030 model emissions were 3.0 percent lower, or nearly 6
million metric tons per year, as shown in Figure 10.23.

Figure 10.22: Annual Mid-Columbia Flat Price Forecast Colstrip Retires Scenario
Colstrip Dispatch Reduction Scenario
One of the methods to reduce emissions in the Northwest without closing Colstrip is to
reduce its generation. This scenario shows the market implications if Colstrip dispatches
less to meet policy objectives. Because the plants are not retired, the scenario does not
require replacement of the generation capacity. Emissions at the plant decrease
beginning in 2023 and continue until the reduction reaches 50 percent of its typical
generation amount. This dispatch constraint lowers emissions and creates an emission
price for the two units. Figure 10.24 provides the resulting emission prices and emission
quantities. Colstrip emissions fall by up to 4 million metric tons annually by 2037. This is
approximately two-thirds of the emission reduction achieved by the Colstrip retires
scenarios. The emission price for this scenario starts around $7 per metric ton and
escalates to $38 per metric ton by 2037. The prices shown are the mean of the 500
simulations. The 95th percentile price in 2037 is $75 per metric ton. Prices will vary
depending on the level of hydro production among other factors such as load, wind
production and natural gas prices.

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Expected Case
Colstrip Retires
SC_PR_3-2 Attachment A Page 161 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.23: No Colstrip Scenario Annual Western U.S. Greenhouse Gas Emissions

Mid-Columbia pricing in this scenario is nearly identical to the Expected Case because
the marginal units driving prices do not change. With similar prices, total Western
Interconnect emissions fall by one percent by 2035 or 2.3 million tons as the reduction
in Colstrip operations is offset by increases in natural gas dispatch in other regions.

Figure 10.24: Colstrip Emissions & Pricing

0
50
100
150
200
250
300
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Mi
l
l
i
o
n
s

o
f

M
e
t
r
i
c

T
o
n
s
Expected Case Colstrip Retires
$0
$5
$10
$15
$20
$25
$30
$35
$40
$45
0
2
4
6
8
10
12
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
$
p
e
r

M
e
t
r
i
c

T
o
n
Mi
l
l
i
o
n

M
e
t
r
i
c

T
o
n
s
Colstrip Carbon Price
Expected Case
Colstrip Reduction
SC_PR_3-2 Attachment A Page 162 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Western Interconnect is 50 Percent Below 1990 Greenhouse Gas Levels Scenario
In each IRP, Avista studies different fundamental shifts in the electric market to
understand the impacts to the market place. Past studies included high solar
penetration, the impact of electric vehicles, and high carbon prices. This IRP uses the
new AURORAXMP constraint modeling functionality to develop a scenario that reduces
Western Interconnect emissions by 50 percent compared to 1990 emission levels. Due
to the uncertainty regarding regional conservation, load growth is the same as the
Expected Case. This is a deterministic case similar to the Expected Case’s deterministic
study. This scenario does not consider variability to hydro, natural gas prices, or other
inputs as described earlier in the chapter. Figure 10.25 illustrates the change in
greenhouse gas emissions compared to the Expected Case. Emissions in the scenario
start out lower due to changes in the new resource selection by the model because it
anticipates significant future emission limits, so the model acquires renewables earlier.

Mid-Columbia prices are significantly higher in this scenario as significant emission
prices drive emissions lower. Prices begin to deviate in 2029 when the price of carbon
escalates at a higher rate, see Figure 10.26. Electric prices levelized for 20 years are 12
percent higher than the Expected Case, but 30 to 40 percent higher in the latter half of
the study. See Figure 10.26. Carbon pricing shown below are for the entire Western
Interconnect, as if the region was a cap and trade system. The levelized price for
emission is $37.54 per metric ton between 2025 and 2037.

This aggressive reduction goal requires new renewables and more natural gas-fired
generation. Figure 10.27 illustrates the change in production in 2037 between this
scenario and the Expected Case. Natural gas generation increases 20 percent, solar 40
percent, and coal reduces 86 percent. Wind generation remains flat, as solar is a lower
cost alternative with fewer limitations. New investment in renewables drives total annual
cost to the system $15.3 billion higher than the Expected Case in the last 10 years of
the study.

SC_PR_3-2 Attachment A Page 163 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.25: Greenhouse Gas Reduction

Figure 10.26: Mid-Columbia Electric Price Comparison

0
50
100
150
200
250
300
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Mi
l
l
i
o
n

M
e
t
r
i
c

T
o
n
s
Expected Case
50 Percent CO2 Reduction
$/Ton
$10/Ton
$20/Ton
$30/Ton
$40/Ton
$50/Ton
$60/Ton
$70/Ton
$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
$80/MWh
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Expected Case
50 Percent CO2 Reduction
Carbon Pricing
SC_PR_3-2 Attachment A Page 164 of 205
Chapter 10- Market Analysis
Avista Corp 2017 Electric IRP
Figure 10.27: 2037 Generation Mix Comparison

-
10,000
20,000
30,000
40,000
50,000
60,000
Natural Gas Solar Wind Coal Other Hydro Nuclear
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Expected Case
50 Percent CO2 Reduction
SC_PR_3-2 Attachment A Page 165 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-1

11. Preferred Resource Strategy

Introduction
This chapter describes potential costs and financial risks of Avista’s new resource and
conservation strategy to meet future requirements over the next 20 years. It explains the
decision making process used to select the Preferred Resource Strategy (PRS), and the
resulting avoided costs used to set targets for future conservation acquisitions and new
resource alternatives.

The 2017 PRS describes a reasonable low-cost plan along the Efficient Frontier of
potential resource portfolios accounting for fuel supply, regulatory and price risks. Major
changes from the 2015 IRP include less energy efficiency (due to lower projected loads),
the addition of demand response and storage resources, less natural gas-fired peaking
capacity, and replacing the planned 2026 CCCT with natural gas-fired peakers.

Demand response returns to the PRS, as program options are more competitive
compared to building new resources. Storage appears for the first time in the plan as
projected costs decline and its modular sizing fits Avista’s small load growth needs. Avista
is also in the process of acquiring a 15 MW DC solar facility to sell to subscribing
commercial and industrial customers of the Solar Select program (see Chapter 4 for
further information). Due to a recent contract extension, Avista’s first resource deficit is in
the winter of 2026 after the expiration of the Lancaster PPA.

Avista will meet the Washington Energy Independence Act with current resources through
the duration of the plan and Avista anticipates reduction in greenhouse gas emissions at
its owned facilities given current policy direction at the state level.

Supply-Side Resource Acquisitions
As shown in Figure 11.1, Avista has made several generation acquisitions and upgrades
over the last 15 years, including:

 25 MW Boulder Park natural gas-fired reciprocating engines (2002);
 7 MW Kettle Falls natural gas-fired CT (2002);
Section Highlights





SC_PR_3-2 Attachment A Page 166 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-2

 35 MW power purchase agreement with the Stateline Wind Project (2004 –
2014);
 72 MW (total) hydroelectric upgrades (2007 – 2016);
 270 MW natural gas-fired Lancaster Generation Station tolling agreement
(2010 – 2026);
 105 MW Palouse Wind power purchase agreement (2012 – 2042); and
 423 kW Boulder Park Community Solar (2015)

Figure 11.1: Resource Acquisition History

Resource Deficiencies
Avista uses both single-hour and 18-hour peak events (six hours per day spread over
three consecutive days) to measure its projected resource adequacy. The single-hour
event assures the system has enough machine capacity to meet an extreme load and/or
outage event. The 18-hour methodology assures energy-limited hydroelectric resources
can meet multiday extreme weather events. For this plan, both summer and winter deficits
are slightly higher for the single-hour event than the 18-hour event.

Avista’s peak planning methodology includes operating reserves, regulation, load
following, variable generation (solar and wind) integration, and a planning margin. Avista
currently projects having adequate resources between owned and controlled generation
to meet physical energy and capacity needs until the end of 2026 when the Lancaster
power purchase agreement expires.1 See Figure 11.2 for Avista’s physical resource
positions for annual energy, summer capacity, and winter capacity. This figure accounts

1 Chapter 6 – Long-Term Position contains details about Avista’s peak planning methodology.
1,100
1,300
1,500
1,700
1,900
2,100
2,300
2,500
19
9
4
19
9
5
19
9
6
19
9
7
19
9
8
19
9
9
20
0
0
20
0
1
20
0
2
20
0
3
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
Me
g
a
w
a
t
t

C
a
p
a
c
i
t
y
Ra
t
h
d
r
u
m
Ce
n
t
r
a
l
i
a

S
a
l
e
BP

&

K
F
C
T
1/
2

C
S
2
St
a
t
e
l
i
n
e

P
P
A
Hydro Upgrades
La
n
c
a
s
t
e
r

P
P
A
1/
2

C
S
2
Pa
l
o
u
s
e

P
P
A
St
a
t
e
l
i
n
e

P
P
A

E
x
p
i
r
e
s
Co
m
m
u
n
i
t
y

S
o
l
a
r

SC_PR_3-2 Attachment A Page 167 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-3

for the effects of energy efficiency programs on the load forecast. Absent energy
efficiency, Avista would be deficient earlier.

Figure 11.2: Physical Resource Positions (Includes Energy Efficiency)

Renewable Portfolio Standards
The Washington Energy Independence Act (EIA) requires utilities with over 25,000
customers to meet 9 percent of current retail load from qualified renewable resources and
15 percent by 2020. The initiative also requires utilities to acquire all cost-effective energy
efficiency.

Avista expects to meet or exceed its EIA renewable energy requirements through the 20-
year plan with a combination of qualifying hydroelectric upgrades, the Palouse Wind
project, and the Kettle Falls Generating Station. Table 11.1 provides a list of the qualifying
generation projects and associated generation and qualifying renewable energy credits
(RECs). Figure 11.3 shows the REC position forecast. The flexibility within the EIA to use
RECs from the current year, from the previous year, or from the following year for
compliance, mitigates year-to-year variability in the output of qualifying renewable
resources.

-500
-400
-300
-200
-100
0
100
200
300
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
g
a
w
a
t
t
s
/

A
v
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Winter 1 Hour Peak (MW)
Summer 1 Hour Peak (MW)
Annual Energy (aMW)
SC_PR_3-2 Attachment A Page 168 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-4

Table 11.1: Qualifying Washington EIA Resources2

Kettle Falls GS Biomass 1983 47 374,824 355,607
Long Lake 3 Hydro 1999 4.5 14,197 14,197
Little Falls 4 Hydro 2001 4.5 4,862 4,862
Cabinet Gorge 3 Hydro 2001 17 45,808 45,808
Cabinet Gorge 2 Hydro 2004 17 29,008 29,008
Cabinet Gorge 4 Hydro 2007 9 20,517 20,517
Wanapum Hydro 2008 0 22,206 0
Noxon Rapids 1 Hydro 2009 7 21,435 21,435
Noxon Rapids 2 Hydro 2010 7 7,709 7,709
Noxon Rapids 3 Hydro 2011 7 14,529 14,529
Noxon Rapids 4 Hydro 2012 7 12,024 12,024
Palouse Wind Wind 2012 105 349,726 419,671
Nine Mile 1 & 2 Hydro 2016 4 21,950 21,950

Figure 11.3: REC Requirements versus Qualifying RECs for EIA

2 The forecasted REC’s shown are based on project capability and may differ from the EIA report due to
the EIA report may include economic dispatch. Palouse Wind receives a 20 percent bonus apprenticeship
credit increasing the number of RECs. Wanapum has no qualifying RECs until the projects uses WREGIS.
-
20
40
60
80
100
120
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Banking
Kettle Falls
Palouse Wind
Hydro Upgrades
Requirement
SC_PR_3-2 Attachment A Page 169 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-5

Resource Selection Process
Avista uses several decision support systems to develop its resource strategy, including
AURORAXMP and Avista’s PRiSM model. The AURORAXMP model, discussed in detail in
the Market Analysis chapter, calculates the operating margin (value) of every resource
option considered in each of the 500 Monte Carlo simulations of the Expected Case, as
well as Avista’s existing generation portfolio. The PRiSM model helps make resource
decisions. Its objective is to meet resource deficits while accounting for overall cost, risk,
capacity, energy, renewable energy requirements, and other constraints.

PRiSM evaluates resource values by combining operating margins with capital and fixed
operating costs. The model creates an Efficient Frontier of resources, or least-cost
portfolios, given a certain level of risk and constraints. Avista’s management selects a
resource strategy using this Efficient Frontier to meet all capacity, energy, renewable
energy, and other requirements.

PRiSM
Avista staff developed the first version of PRiSM in 2002 to support resource decision
making in the 2003 IRP. Enhancements over the years have improved the model. PRiSM
uses a mixed integer programming routine to support complex decision making with
multiple objectives. These tools provide optimal values for variables, given system
constraints.

PRiSM Model Overview
The PRiSM model requires a number of inputs:
1. Expected future deficiencies
o Greater of summer 1- or 18-hour capacity
o Greater of winter 1- or 18-hour capacity
o Annual energy
o EIA requirements
2. Costs to serve future retail loads as if served by the wholesale marketplace
3. Existing resource and conservation contributions
o Operating margins
o Fixed operating costs
4. Resource and conservation options
o Fixed operating costs
o Return on capital
o Interest expense
o Taxes
o Generation levels
o Emission levels
5. Constraints
o Must meet energy, capacity and RPS shortfalls without market reliance
o Resources quantities available to meet future deficits

PRiSM uses these inputs to develop an optimal resource mix over time at varying levels
of risk. PRiSM considers new resource costs over the next 50 years to consider long-term
SC_PR_3-2 Attachment A Page 170 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-6

resource implications, but it weights the first 25 years more than the later years to highlight
the importance of nearer-term decisions. Equation 11.1 shows a simplified view of the
PRiSM linear programming objective function.

Equation 11.1: PRiSM Objective Function

Minimize: (X1 * NPV2018-2042) + (X2 * NPV2018-2067)

Where: X1 = Weight of net costs over the first 25 years (95 percent)
X2 = Weight of net costs over the next 50 years (5 percent)
NPV is the net present value of total system cost.3

An efficient frontier captures the optimal resource mix graphically given varying levels of
cost and risk. Figure 11.4 illustrates the efficient frontier concept.

Figure 11.4: Conceptual Efficient Frontier Curve

As you attempt to lower risk, costs increase. The optimal point on the Efficient Frontier
depends on the level of acceptable risk. No best point on the curve exists, but Avista
prefers points where small incremental cost additions offer larger risk reductions.
Portfolios to the left of the curve are more desirable, but do not meet the planning
requirements or resource constraints. Examples of these constraints include
environmental costs, regulation, and the availability of commercially viable technologies.
Portfolios to the right of the curve are less efficient as they have higher costs than a

3 Total system cost is the existing resource marginal costs, all future resource fixed and variable costs, and
all future energy efficiency costs, and the net short-term market sales/purchases.

Ri
s
k
Cost
Least Cost
Least Risk
SC_PR_3-2 Attachment A Page 171 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-7

portfolio with the same level of risk. PRiSM meets all deficit projections with new
resources of the actual sizes available in the marketplace and does not use market
purchases. As discussed earlier in this chapter, reflecting real-world constraints in the
model is necessary to create a realistic representation of the future. Some constraints are
physical and others are policy driven. The major resource constraints are capacity and
energy needs, the EIA, and the greenhouse gas emissions performance standard.

Preferred Resource Strategy
The 2017 PRS consists of existing thermal resource upgrades, energy efficiency, demand
response, storage and natural gas-fired peakers (See Table 11.2 and Figure 11.5). The
15 MW (DC) solar facility for Avista’s new voluntary Solar Select Program is also included
in the resource plan4. Prior to the first capacity and energy need in 2026, the PRS shows
Avista beginning two demand response programs to reduce loads at system peak. Both
Solar Select and the DR programs will require commercial and industrial cooperation,
regulatory approvals, permitting, and starting the program early to ensure enough
participants are available when our deficit requires it.

Additional thermal based resources meet the first large deficit created by the expiration
of the Lancaster PPA. It is possible this resource could be re-acquired, or an alternative
market resource may be available at a lower cost. Without an acquisition, the first new
resource is a 192 MW of natural gas-fired peakers and upgrades at existing thermal
facilities. Given the small cost differences between the evaluated natural gas-fired peaker
technologies, the future technology decision likely will be refined in a Request for
Proposals (RFP) process. Technological changes in efficiency and flexibility may lead
Avista to revisit this resource choice closer to the actual need.

Table 11.2: 2017 Preferred Resource Strategy

Resource By the End
of Year
ISO Conditions
(MW)
Winter Peak
(MW)
Energy
(aMW)
Solar (Solar Select Program) 2018 15 0 3
Natural Gas Peaker 2026 192 204 178
Thermal Upgrades 2026-2029 34 34 31
Storage 2029 5 5 -0
Natural Gas Peaker 2030 96 102 89
Natural Gas Peaker 2034 47 47 43
Total 389 392 344
Efficiency Improvements Acquisition
Range
Winter Peak
Reduction
(MW)
Energy
(aMW)
Energy Efficiency 2018-2037 203 108
Demand Response 2025-2037 44 <0
Distribution Efficiencies <1 <1
Total 247 108

4 The size of the Solar Select facility may change from the RFP amount if program participation exceeds
the initial 15 MW program.
SC_PR_3-2 Attachment A Page 172 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-8

Figure 11.5: New Resources to Meet Winter Peak Loads

After a combination of upgrades to existing thermal facilities, new peakers, and demand
response, Avista’s customers still will require additional capacity as loads grow and
contracts expire. The next acquisition is a storage resource. The selected storage unit
has a five-megawatt capacity rating, and 15 megawatt-hours of storage. Following the
storage resource addition, a significant wholesale power contract expires at the end of
2030. To fill this gap, PRiSM selects a 96 MW natural gas fired peaker unless renewing
the contract under favorable terms benefits customers. The last selected resource of the
20-year plan is a 47 MW natural gas-fired peaker by the end of 2034.

2015 IRP Comparison
The 2017 PRS differs from the 2015 PRS shown in Table 11.3. Lower load growth and
contract changes push resource needs out to 2026 rather than by the end of 2020. New
resource needs are 191 MW lower due to lower load growth, higher expected
conservation at the time of system peak, and the addition of new demand response and
storage programs. These factors further reduce the need for new fossil fuel resources.
The 2015 PRS combined cycle plant is now too large relative to the projected need for
replacing Lancaster with a new facility. Further, market conditions are changing due the
amount of new renewable resources in the west, favoring flexible peaking resources over
historically intermediate and baseload resources. Avista preformed a scenario, discussed
in Chapter 12, showing if Avista continued assuming replacing Lancaster with a new
CCCT plant to see the cost and risk impact to the portfolio.

-
500
1,000
1,500
2,000
2,500
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
g
a
w
a
t
t
s
Demand Response Storage
Plant Upgrade Peaker
Existing Resources & Rights Load w/ Conservation + Contingency
SC_PR_3-2 Attachment A Page 173 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-9

Table 11.3: 2015 Preferred Resource Strategy

Resource By the End of
Year
ISO Conditions
(MW)
Winter Peak
(MW)
Energy
(aMW)
Natural Gas Peaker 2020 96 102 89
Thermal Upgrades 2021-2025 38 38 35
Combined Cycle CT 2026 286 306 265
Natural Gas Peaker 2027 96 102 89
Thermal Upgrades 2033 3 3 3
Natural Gas Peaker 2034 47 47 43
Total 565 597 524
Efficiency
Improvements
Acquisition
Range
Winter Peak
Reduction
(MW)
Energy
(aMW)
Energy Efficiency5 2016-2035 193 132
Distribution Efficiencies <1 <1
Total 193 132

Energy Efficiency
Energy efficiency is an integral part of the PRS. It also is a critical component of the EIA
requirement for utilities to obtain all cost-effective conservation. PRiSM considers energy
efficiency and supply side options at the same time to ensure compliance with the EIA.
PRiSM models each specific energy efficiency measure individually and does not bundle
measures. This allows the selection of different conservation amounts at each point along
the Efficient Frontier to capture changes in the risk profiles of additional conservation.
This capability improves previous IRP evaluations assuming a constant conservation
acquisition level along the entire curve.

Conservation options inclusion within PRiSM requires a load forecast without future
conservation. Due to industry-standard load forecasting methods, Avista’s load forecast
is the load expectation net of future energy efficiency. Estimating the amount of
conservation included in the forecast requires evaluating its economic potential. This
requires an iterative process with PRiSM to validate if selected conservation is similar to
the assumed conservation level in the load forecast. For example, if PRiSM selects less
conservation than originally estimated, it runs again with a lower amount of conservation
until the predetermined conservation is similar to the selected conservation on an annual
energy basis. For this IRP, selected conservation is very similar to the levels in the
forecast. The difference is three percent higher in the first 10 years, and two percent
higher over 20 years, or 1.9 aMW.

Figure 11.6 illustrates the load forecast with and without conservation. The selected 108
aMW of savings represents 53.3 percent of expected load growth between 2018 and
2037. Please refer to Chapter 5 for a detailed discussion of energy efficiency resources.

5 Total energy efficiency estimates include savings from transmission and distribution system losses.

SC_PR_3-2 Attachment A Page 174 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-10

Because portfolio analysis described in this chapter considers the impacts of transmission
and distribution losses, savings in Chapter 5 are lower than shown here.

Figure 11.6: Load Forecast with and without Energy Efficiency

Grid Modernization
Distribution feeder upgrades entered the PRS in the 2009 IRP and the grid modernization
process began with the Ninth and Central feeder in Spokane. The decision to rebuild a
feeder considers savings from reducing energy losses, operation and maintenance
savings, the age of installed equipment, reliability indices, and the number of customers
on the feeder. System reliability, instead of energy savings, generally drives feeder rebuild
decisions. Therefore, feeder upgrades are no longer included as a resource option in
PRiSM. A broader discussion of Avista’s feeder rebuild program is in Chapter 8.

Natural Gas-Fired Peakers
Avista plans to locate potential sites for new natural gas-fired generation capacity within
its service territory ahead of an anticipated need. The option of having a utility-owned site
is very low cost relative to the final acquisition cost of a natural gas-fired plant, and this
strategy ensures customers will not pay a premium over the actual cost of building a new
asset. A 2013 Action Item was to identify a utility-build natural gas resource site. Since
then, Avista procured land in North Idaho in the event a greenfield site benefits customers.
A second option for a smaller resource need is possible at the Rathdrum CT site.

Avista is not specifying a preferred peaking plant technology at this time. Tradeoffs will
occur between capital costs, size, operating efficiency, and flexibility. Relative to other
natural gas-fired peaking facilities, frame CT machines are a lower capital-cost option,
but have higher operating costs and less flexibility, while the hybrid technology and aero
-
200
400
600
800
1,000
1,200
1,400
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Net Load Forecast w/ Conservation
Expected Case Load Forecast w/o Conservation
SC_PR_3-2 Attachment A Page 175 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-11

turbines have higher capital costs, lower operating costs, and more operational flexibility.
Advances in natural gas-fired reciprocating engines are also of interest. These resources
utilize a group of smaller units to reduce the risk of a larger single plant breaking down,
have lower heat rates, and are highly flexible, but can be more capital and O&M intensive
than other technologies. Increased flexibility requirements and greenhouse gas
emissions costs could make a hybrid plant or reciprocating engines preferable. Avista
currently has enough resource flexibility to meet customer needs to drive the strategy
towards a lower cost peaker option, but potential future participation in an energy
imbalance markets may provide enough revenues for a flexible peaker to offset the higher
costs. It is also possible other resource options such as CCCT, storage, or hydro could
cost effectively compete against new peakers when procuring the new resource.

Greenhouse Gas Emissions
Chapter 10 – Market Analysis, discusses how greenhouse gas emissions decrease due
to coal plant retirements across the Western Interconnect. Avista’s projected resource
mix does not include any retirements. The only significant carbon emitting resource
leaving the portfolio is the expiration of the Lancaster PPA in 2026. Figure 11.7 presents
Avista’s expected greenhouse gas emissions (excluding Kettle Falls GS) with the addition
of 2017 PRS resources. The estimates in Figure 11.7 do not include emissions from
purchased power or adjustments to reduce emissions for off-system sales. Emissions in
2037 are 11 percent lower than the 2018/19 average emissions and 18 percent lower on
a per MWh basis. Emissions are 29 percent lower as compared to the 2015 IRP’s
estimate for 2035. The emissions reduction comes from adding natural gas-fired peaking
units instead of building a new CCCT facility, and a reduction in the dispatch at Colstrip
3 & 4 due to modeled emission regulations.

Figure 11.7: Avista Owned and Controlled Resource’s Greenhouse Gas Emissions

-
0.13
0.25
0.38
0.50
Mil
1 Mil
2 Mil
3 Mil
4 Mil
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
t
r
i
c

T
o
n
s

p
e
r

M
W
h
Me
t
r
i
c

T
o
n
s
Expected Total
Metric Tons per MWh
SC_PR_3-2 Attachment A Page 176 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-12

Capital Spending Requirements
The IRP assumes Avista will finance and own all new resources for IRP planning
purposes.6 A competitive acquisition processes may hold different result, but under this
assumption, and the resources identified in the 2017 PRS, the first capital addition to rate
base is in 2025 as capital improvements are required for the stand-by generation DR
program. In 2027, significant investment will be required for the first natural gas-fired
peaker as a replacement for the Lancaster PPA. If a new facility replaces Lancaster,
construction would begin prior to need, but the resource’s capital cost would not enter
rate base until after it is placed in service. The capital cash flows in Table 11.4 include
AFUDC, generation capital costs, and transmission investments for generation, tax
incentives, and sales taxes. Over the 20-year IRP timeframe, $538 million (nominal) in
generation and related transmission expenditure is required to support the PRS.

Table 11.4: PRS Rate Base Additions from Capital Expenditures
(Millions of Dollars)

Year Investment Year Investment
2018 0.0 2028 2.1
2019 0.0 2029 9.5
2020 0.0 2030 9.9
2021 0.0 2031 140.1
2022 0.0 2032 0.5
2023 0.0 2033 0.5
2024 0.0 2034 0.5
2025 2.3 2035 94.1
2026 2.0 2036 0.5
2027 275.7 2037 0.5
2018-27 Total 280.0 2028-37 Totals 258.2

Annual Power Supply Expenses and Volatility
PRS variance analysis tracks fuel, variable O&M, emissions, and market transaction
costs for the existing resource portfolio for each of the 500 Monte Carlo iterations of the
Expected Case risk analysis. In addition to existing portfolio costs, new resource capital,
fuel, O&M, emissions, and other costs provide a range of expected costs to serve future
loads. Figure 11.8 shows expected PRS costs through 2037 as the blue bars. In 2018,
portfolio costs average $26 per MWh. By 2037, costs approach $60 per MWh. The chart
shows a two-sigma cost range with yellow diamonds representing the lower range and
orange triangles representing the upper range. The main drivers increasing power supply
costs and volatility are natural gas prices and weather, which affect both hydroelectric
generation and load variability. Avista increases the volatility assumption of future natural
gas prices because the commodity price has unknown future risks and a history of
volatility.

6 Except for resources taking advantage of the ITC, such as solar.
SC_PR_3-2 Attachment A Page 177 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-13

Figure 11.8: Projected Power Supply Expense Range

Efficient Frontier Analysis
The Efficient Frontier analysis is the backbone of the PRS. The PRiSM model develops
the efficient frontier by simulating the costs and risks of resource portfolios using a mixed-
integer linear program. PRiSM finds an optimized least cost portfolio for a range of risk
levels. The PRS analyses examined the following portfolios.

 Least Cost: Meets all capacity, energy and RPS requirements with the least-cost
resource options. This portfolio ignores power supply expense volatility in favor of
lowest-cost resources.
 Least Risk: Meets all capacity, energy, and RPS requirements with the least-risk
mix of resources. This portfolio ignores the overall cost of the selected portfolio in
favor of minimizing year-on-year portfolio cost variability.
 Efficient Frontier: Meets all capacity, energy, and RPS requirements with sets of
intermediate portfolios between the least risk and least cost options. Given the
resource assumptions, no resource portfolio can be at a better cost and risk
combination than these portfolios.
 Preferred Resource Strategy: Meets all capacity, energy, and RPS requirements
while recognizing both the overall cost and risk inherent in the portfolio. Avista’s
management chose this portfolio as the most reasonable strategy given current
information.

Figure 11.9 presents the Efficient Frontier in the Expected Case. The x-axis is the
levelized nominal cost per year for the power supply portfolio, including capital recovery,
-
10
20
30
40
50
60
70
80
90
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Do
l
l
a
r
s

p
e
r

M
W
h

(
N
o
m
i
n
a
l
)
Expected Cost
Two Sigma Low
Two Sigma High
SC_PR_3-2 Attachment A Page 178 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-14

operating costs, and fuel expense; the y-axis displays standard deviation of power supply
costs in 2030. It is necessary to move far enough into the future so load growth provides
PRiSM the opportunity to make new resource decisions. The year 2030 is far enough into
the future to account for the risk tradeoffs of several resource decisions. Using an earlier
year to measure risk would have too few new resource decisions available to distinguish
between portfolios.

Avista chose to use the least cost portfolio for this IRP. Past IRPs selected a portfolio with
lower risk, but slightly higher cost. The main difference between this plan and prior plans
is first the choice to replace Lancaster after the expiration of the PPA with peaking plants.
Avista chose to move away from a baseload resource due to the lower capacity
requirements upon its expiration. With the lower capacity requirement, adding a CCCT
(without a partner) would increase customer’s costs until the company could grow into the
excess capacity. The second reason for the change is to take advantage of a low electric
market price forecast by selecting natural gas-fired peakers and demand response.
Avista’s resource strategy meets reliability requirements and selects new resources to
meet rapid changes in daily price volatility due to renewable resources in the region. If
Avista maintains its strategy to replace Lancaster with a new CCCT, the costs would be
0.8 percent higher (PVRR) and the risk in 2030 would increase by 10 percent. While this
scenario is similar to the portfolios on the Efficient Frontier analysis, there are other more
optimal portfolios with similar risk, but at lower cost. More information regarding this
scenario is in Chapter 12.

Figure 11.9: Expected Case Efficient Frontier

$20 Mil
$30 Mil
$40 Mil
$50 Mil
$60 Mil
$70 Mil
$80 Mil
$90 Mil
$350 Mil $400 Mil $450 Mil $500 Mil $550 Mil
20
3
0

S
t
d
e
v

o
f

P
o
w
e
r

S
u
p
p
l
y

C
o
s
t
s
Levelized Cost 2018-2042
Least Cost Preferred Resource Strategy
Least Risk
SC_PR_3-2 Attachment A Page 179 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-15

Selecting the appropriate point on the Efficient Frontier is not solvable through a
mathematical formula. Portfolio selection along the Efficient Frontier is from a
determination of management’s judgment of cost versus risk. In prior IRPs, management
selected lower heat rate facilities to protect the portfolio from wholesale market volatility
by moving down the frontier curve. In this IRP, management is pursuing a modestly higher
risk strategy by selecting peakers over CCCTs. Given the uncertainties in the marketplace
today, including carbon mitigation policies, this choice gives more flexibility. Since our
resource need is nine years away, multiple IRP’s will be able to change course if needed
when more information becomes available.

The 2015 IRP presented a method for reviewing portfolios along the Efficient Frontier as
part of a request by the Washington Commission Staff. This method is a risk adjusted
Present Value of Revenue Requirement, or PVRR, taking into account the tail risk. The
first step calculates risk adjusted PVRR for each portfolio. This calculation is the net
present value (NPV) of the future revenue requirements, plus the present value of taking
each of the future year’s tail risk, calculated by five percent of the 95th percentile’s
increase in costs. This methodology assumes the lowest NPV should yield the best
strategy. Figure 11.10 shows the results of this study on the Efficient Frontier. The first
two portfolios are the least cost adjusted for this risk calculation. The second portfolio is
0.003 percent lower cost than the PRS (Least Cost scenario), meaning the portfolios are
essentially identical. The only difference is the resources selected are after 2035.

Figure 11.10: Risk Adjusted PVRR of Efficient Frontier Portfolios

$ Bil
$1 Bil
$2 Bil
$3 Bil
$4 Bil
$5 Bil
$6 Bil
Le
a
s
t

C
o
s
t
2 3 4 5 6 7 8 9 10 11 12 13 14 15
Le
a
s
t

R
i
s
k
20

Y
e
a
r

R
i
s
k

A
d
j
u
s
t
e
d

P
V
R
R

Efficient Frontier Portfolios
SC_PR_3-2 Attachment A Page 180 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-16

To illustrate tradeoffs between the cost and risk of each portfolio along the Efficient
Frontier, a point-to-point derivative of the slope of the change in cost relative to the change
in absolute costs is useful. In this case, a greater slope indicates increasing benefits for
trading risk reduction for higher portfolio costs. Figure 11.11 illustrates the results of this
study. The PRS selected by PRiSM is the least cost portfolio, but moving down the frontier
does provide good risk versus cost tradeoffs, as the slope of the Efficient Frontier is
steeper until the sixth portfolio. As time passes, Avista may choose to move further down
the Efficient Frontier given Avista’s first resource need is not eminent.

Figure 11.11: Risk Adjusted PVRR of Efficient Frontier Portfolios

Other Efficient Frontier Portfolios
In addition to the PRS, the Efficient Frontier contains 15 additional resource portfolios.
The lower cost and higher risk portfolios contain primarily natural gas peakers and
renewable resources to reduce risk. The amount of conservation varies in these portfolios
as it lowers risk and fills deficiencies depending on the resources selected. For example,
the model must select a resource size actually available in the marketplace. Given this
“lumpiness”, it may be more efficient to meet some needs with conservation rather than
building a much larger generation asset. This discussion continues in Chapter 12 –
Portfolio Scenarios.

Table 11.5 details the selected resource totals between 2018 and 2037 for each Efficient
Frontier portfolio. Toward the middle of the Efficient Frontier, PRiSM favors wind and solar
to reduce market risk as additional conservation resources become more expensive. The
lower half of the Efficient Frontier includes portfolios with large capacity surpluses and
-
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Le
a
s
t

C
o
s
t
2 3 4 5 6 7 8 9 10 11 12 13 14 15
Le
a
s
t

R
i
s
k
Ef
f
i
c
i
e
n
t

F
r
o
n
t
i
e
r

S
l
o
p
e
Efficient Frontier Portfolios
SC_PR_3-2 Attachment A Page 181 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-17

renewable resources, meanwhile maxing out the amount of conservation included in the
model. The least risk portfolio has no financial objective and selects as many resources
as possible given the model’s constraints to lower risk. A new natural gas CCCT does not
appear anywhere on the Efficient Frontier for the first time since PRiSM was adopted in
the 2003 IRP. This is because new CCCT units are too large relative to Avista’s load
requirements.

Table 11.5: Alternative Resource Strategies (2035) along the Efficient Frontier (MW)

Po
r
t
f
o
l
i
o

Le
v
e
l
i
z
e
d

Co
s
t

20
3
0

St
d
e
v

NG

Pe
a
k
e
r

NG

C
C
C
T

Wi
n
d

So
l
a
r

De
m
a
n
d

Re
s
p
o
n
s
e
Th
e
r
m
a
l

Up
g
r
a
d
e

St
o
r
a
g
e

Hy
d
r
o

Up
g
r
a
d
e

En
e
r
g
y

Ef
f
i
c
i
e
n
c
y

Determining the Avoided Costs of Energy Efficiency
The Efficient Frontier methodology determines the avoided cost of new resource additions
included in the PRS. There are two avoided cost calculations for this IRP: one for energy
efficiency and one for new generation resources. The energy efficiency avoided cost is
higher because it includes benefits beyond generation resource value.

Avoided Cost of Energy Efficiency
Since PRiSM selects energy efficiency, the prior IRP method of calculating avoided costs
is no longer required; but estimating these values is helpful in selecting future
conservation measures for more detailed analysis between IRPs. The process used to
estimate avoided cost calculates the marginal cost of energy and capacity of the
resources selected in the PRS. The energy value uses hourly energy prices to value more
highly measures providing more contribution during system peak. The value of
conservation measures includes the energy value, the ten percent Power Act adder and
SC_PR_3-2 Attachment A Page 182 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-18

the value of loss savings.7 Reducing customer loads saves future distribution and
transmission capital and O&M costs, and is included in the conservation-avoided cost
calculation. The final component of avoided cost accounts for the savings from avoided
new capacity. This capacity value is the difference between the cost of a resource mix
and the value the mix earns from energy sales in the wholesale marketplace. Equation
11.2 describes the avoided costs to evaluate conservation measures. This equation is the
same as the 2015 IRP.

Equation 11.2: Conservation Avoided Costs

{(E + (E * L) + DC) * (1 + P)} + PCR

Where:
E = Market energy price. The price calculated by AURORAXMP is $35.85 per MWh
assuming a flat load shape.
PCR = New resource capacity savings for the PRS selection point is estimated to
be $120 per kW-year (winter savings only).
P = Power Act preference premium. This is the additional 10 percent premium
given as a preference towards energy efficiency measures.
L = Transmission and distribution losses. This component is 6.0 percent based on
Avista’s estimated system average losses.
DC = Distribution capacity savings. This levelized value is approximately $34.41
per kW-Year.

Determining the Avoided Cost of New Generation Options
The 2017 IRP’s avoided costs are in Table 11.6. However, avoided costs will change as
Avista’s loads and resources change, as well with the wholesale power marketplace
changes. The prices shown in the table represent energy & capacity values for different
periods and product types. For example, for a new project with equal deliveries over the
year in all hours has an energy value equal to the Flat Energy price shown in Table 11.6.
Traditional on-peak and off-peak pricing is also included as a comparison to the flat price.
In addition to the energy prices, this theoretical resource would also receive the capacity
value as it produces power at the time of system peak. This system peak contributing
value begins in 2027 for potential resources that can dependably meet winter peak
requirements.

Capacity values shown below are the marginal cost of the most expensive significant
resource from the PRS each year. The significant resources in this case are the natural
gas-fired peakers. These resources set the avoided capacity cost, rather than smaller
technologies, as the smaller technologies selected may not represent the marginal cost

7 The Power Act adder refers to one aspect of federal law enacted in 1980 along with the creation of the
NPCC. The NPCC includes the 10 percent adder to deferred capacity, given Avista’s new conservation
methodology using this 10 percent adder would not allow Avista’s PRiSM model to solve, as it would be
non-linear. Avista compared it’s conservation method to the older method that calculates conservation
outside PRiSM with the 10 percent adder in the 2015 IRP and both methods produced similar results.

SC_PR_3-2 Attachment A Page 183 of 205
Chapter 11 – Preferred Resource Strategy

Avista Corp 2017 Electric IRP 11-19

if changes are made to loads or resources or if the PRiSM model is able to select
resources to proper size requirement. The capacity payment applies to the capacity
contribution of the resource at the time of the winter peak hour. To obtain a full capacity
payment the resource must generate 100 percent of its capacity rating at the time of
system peak. Solar receives no payment because it does not generate at the time of
Avista’s planned system peak (winter evenings or mornings when it is still dark). Wind
resources may qualify for some contribution depending on the correlation and
diversification of the resource. For example, this IRP assumes 7.5 percent winter capacity
credit for Montana wind resources. The capacity cost methodology of this analysis is the
same as the 2015 and prior plans by using the natural gas-fired resources as the avoided
capacity unit. The only major difference from prior plans is the inclusion of specific avoided
costs for renewables.

As an alternative to showing tipping point analysis to determine when a solar or wind
resource is cost effective, the avoided energy value of these resources is part of this table.
For solar, the levelized price to be economic for customers between 2017 and 2037, is
$29.90 per MWh and for wind the economic price is $31.81 per MWh. These values do
not include costs to integrate variable energy production, reserves, or interconnection
costs, but represent the energy value of the resource’s generation. The value attributed
to these resources vary due to the time of expected delivery of the resources.

Table 11.6: 2017 IRP Avoided Costs

Year Flat
Energy
$/MWh
On-Peak
Energy
$/MWh
Off-Peak
Energy
$/MWh
Capacity
$/kW-Yr
Example
WA Solar
$/MWh
Example
WA Wind
$/MWh
2018 23.79 27.02 19.48 0 23.70 21.66
2019 23.71 26.85 19.53 0 23.28 21.71
2020 23.99 26.85 20.16 0 22.37 21.76
2021 24.30 26.85 20.88 0 21.67 21.63
2022 25.95 28.47 22.59 0 22.54 22.92
2023 29.68 32.24 26.30 0 25.36 26.35
2024 32.03 34.38 28.90 0 26.62 28.40
2025 32.58 34.65 29.83 0 26.66 28.85
2026 34.27 36.13 31.77 0 27.42 30.23
2027 37.61 39.25 35.43 171 29.51 33.25
2028 40.18 41.60 38.28 174 30.91 35.20
2029 44.06 45.27 42.44 178 33.84 38.65
2030 46.86 48.15 45.15 181 36.19 41.01
2031 48.08 49.32 46.42 185 36.88 41.98
2032 51.10 52.55 49.17 189 39.26 44.82
2033 52.81 54.29 50.83 192 40.73 46.13
2034 55.09 56.61 53.07 196 43.28 48.35
2035 57.50 59.26 55.14 200 45.96 50.51
2036 60.52 62.22 58.24 204 48.13 53.15
2037 64.51 66.33 62.09 208 51.98 57.14

SC_PR_3-2 Attachment A Page 184 of 205

SC_PR_3-2 Attachment A Page 185 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-1

12. Portfolio Scenarios

Introduction
The Preferred Resource Strategy (PRS) is Avista’s 20-year strategy to meet future loads.
Because the future is often different from the IRP forecast, the strategy needs to be
flexible enough to benefit customers under a range of plausible outcomes. This chapter
investigates the cost and risk impacts to the PRS with different futures the utility might
face. It reviews the impacts of losing a major generating unit, evaluates alternative loads,
determines the impact of unit sizing, and the selection of portfolios to the right of the
Efficient Frontier. All portfolios include the Solar Select project discussed in Chapter 11.

Load Forecast Scenarios
The PRS meets the Expected Case’s load growth of 0.45 percent and winter peak growth
of 0.39 percent over the next 20 years. Chapter 3 – Economic and Load Forecast provides
details about the alternative load forecasts and Table 12.1 summarizes the alternative
growth assumptions used to determine how the plan would change under different
economic conditions.

Table 12.1: Load Forecast Scenarios (2018-2037)

Scenario Energy
Growth (%)
Winter
Peak
Growth (%)
Summer
Peak
Growth (%)
Expected Case 0.45 0.39 0.42
High Load 0.74 0.72 0.78
Low Load 0.16 0.03 0.04

Table 12.2 shows the changes to the PRS for each load growth scenario. In each
scenario, a natural gas-fired CT is required by the end of 2026. Both the Low Load Growth
case and the PRS add a 192 MW natural gas-fired CT by the end of 2026. The High Load
Growth case requires 288 MW of additional capacity by the end of 2026. In all cases, the
thermal upgrade selection is the same, but the timing of resources change, as the
resource needs change. In both alternative scenarios, the storage facility is not cost
effective, due to the size of selected resources needed to meet capacity needs. In the
Expected Case, storage is the lowest cost resource for small incremental needs, but not
for larger requirements. The portfolios for all three cases are similar with no scenario
Chapter Highlights





SC_PR_3-2 Attachment A Page 186 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-2

requiring a different decision date for a new facility; the only major difference is the size
of the addition. Near the 2026 requirement, Avista will have a greater understanding of
its actual requirements.

Table 12.2: Resource Selection for Load Forecast Scenarios

Resource
Expected
Case's
PRS
High
Load
Growth
Low
Load
Growth
NG Peaker 335 477 192
NG Combined Cycle CT 0 0 0
Wind 0 0 0
Solar 0 0 0
Demand Response 49 49 49
Storage 5 0 0
Thermal Upgrades 34 34 34
Hydro Upgrades 0 0 0
Total 423 560 275

Colstrip Scenarios
Coal-fired power plants are facing pressure from both policy requirements and economics
to reduce their dispatch or to shut down. Avista’s TAC and state commissions asked
Avista to study the impacts of shuttering Colstrip prior to the end of its operating life. This
IRP studies two alternative shutdown scenarios including coal-fired plant dispatch is
limited due to more restrictive carbon reduction policies relative to the Expected Case’s
assumption.

In the Expected Case, Avista’s ownership interests in the plant remains cost effective for
the next 20 years, although it dispatches less due to carbon regulation projections. The
Expected Case also includes Selective Catalytic Reduction (SCR) beginning service in
2028, significant capital expenses for Coal Combustion Residual (CCR) requirements and
water management issues. Operating costs will increase when Units 1 & 2 close because
there will be additional O&M costs and possible requirements for additional mercury
controls.

Colstrip Retirement Scenario
This IRP includes two scenarios with Colstrip retiring in 2030 and 2035. Both represent
plausible early retirement dates when the plant could end service to customers. These
scenarios assume both closure dates eliminate capital spending for the SCR and shorten
capital recovery to current and future capital to five years after the retirement date. Future
capital costs are lower than the Expected Case as certain capital improvements are
cancelled. The CCR costs remain the same as in the Expected Case, but the time to
complete the projects accelerates. The scenarios do not include costs related to
employee retraining or relocation, payments to other owners, or decommissioning beyond
those already included rates.
SC_PR_3-2 Attachment A Page 187 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-3

Table 12.3 shows the resource strategy for the Colstrip retirement scenarios. For the 2030
scenario, the table includes options for natural gas peakers and a CCCT. The 2035
scenario only shows replacement with peakers, although a CCCT could replace the plant,
the cost illustration shown in 2030 represents this scenario. Figure 12.1 illustrates the
costs and power supply risks of retiring Colstrip compared to the Efficient Frontier and the
PRS. This chart shows the annual levelized costs between 2018 and 2042 on the x-axis
and the 2037 standard deviation of power supply costs on the y-axis1. A separate scenario
replacing Colstrip with energy storage and renewables appears later in this chapter.
Retiring Colstrip early increases costs compared to the PRS, while pushing the retirement
date out to 2035 is the least cost of the retirement scenarios, due to the added costs
representing a smaller portion of the financial period. To understand the cost increases
in the year of retirement, Figure 12.2 compares the annual cost of each scenario and the
PRS.

The year following the plant retirement, power supply costs increase $50 to $60 million
due to the cost of new capacity; this represents a 10 to 13 percent increase in power
supply expenses as compared to the PRS. Reduced capital spending offsets some of the
cost increases prior to the shutdown, but not enough to offset the increase. The CCCT
option costs $1.8 million more per year (0.4 percent than the peaker option, but risk is 8
percent lower.

Table 12.3: Colstrip Retires- Resource Strategy Options (ISO Conditions MW)

Resource By End of
Year
2030
Retirement
with Peaker
2030
Retirement
with CCCT
2035
Retirement
with Peaker
Natural Gas Peaker 2026 192 192 192
Thermal Upgrades 2027-2030 34 34 34
Storage 2028 5 5 5
Natural Gas Peaker 2030 288 0 96
Natural Gas CCCT 2030 0 286 0
Storage 2032 5 5 0
Natural Gas Peaker 2033 47 47 0
Natural Gas Peaker 2034 0 0 47
Natural Gas Peaker 2035 0 0 192
Total 571 569 566

Demand Response 2025-2037 44 44 48
Conservation (w/ T&D losses) 2018-2037 107 107 108

Early Colstrip retirement decreases direct greenhouse gas emissions as shown in Figure
12.3. In the natural gas-fired peaker scenario, direct emissions decrease 62 percent in

1 The risk year is shifted to 2037 rather than 2030 used in other section to reflect change risk profile changes
for portfolio choices late in the study period.
SC_PR_3-2 Attachment A Page 188 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-4

2037 compared to the PRS. If a CCCT replaces Colstrip, direct emissions fall 44 percent.
The CCCT has higher direct emissions because it dispatches more hours than the less
thermally efficient NG peaker. For the peaker scenario, Avista would rely on market
purchases except when the peaker dispatch price is less expensive than purchasing from
the market. Another method to review this scenario is the implied cost of carbon of
shutting down the units. Using the average cost change between 2031 and 2037 and
dividing by the average direct emissions reduction is an implied cost of $17.41 per metric
ton, this with the pricing included in the market price forecast totals $38.78 per metric
ton.2

Figure 12.1: Colstrip Retires Scenario Cost versus Risk

2 This does not include indirect emissions from market purchases; depending on the methodology used to
estimate these emissions the cost per ton could be higher. In the CCCT replacement scenario, the implied
cost of carbon is $48.18 per metric ton using the same methodology.
$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
$350 $400 $450 $500 $550
20
3
7

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Levelized Annual Power Supply Cost (2018-42, Millions)
Expected Case Efficient Frontier
Expected Case (PRS)
Colstrip Retires (2030- Peakers)
Colstrip Retires (2030- CCCT)
Colstrip Retires (2035- Peakers)
SC_PR_3-2 Attachment A Page 189 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-5

Figure 12.2: Annual Cost Impact with Colstrip Retirement versus PRS

Figure 12.3: Annual Greenhouse Gas Emissions with Colstrip Retirement

-$20
-$10
$0
$10
$20
$30
$40
$50
$60
$70
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
An
n
u
a
l

P
o
w
e
r

S
u
p
p
l
y

C
o
s
t
s

(
M
i
l
l
i
o
n
s
)
Colstrip Retires (2030- Peakers)
Colstrip Retires (2030- CCCT)
Colstrip Retires (2035- Peakers)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
t
r
i
c

T
o
n
s

(
M
i
l
l
i
o
n
s
)
Colstrip Retires (2030- Peakers)
Colstrip Retires (2030- CCCT)
Colstrip Retires (2035- Peakers)
Preferred Resource Strategy
SC_PR_3-2 Attachment A Page 190 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-6

High-Cost Colstrip Retention Scenario
As part of the acceptance letter from the 2015 IRP, the Washington Commission
requested a scenario with a higher than expected compliance costs to retain Colstrip and
consult with the TAC regarding carbon pricing policies in the stochastic model. This
scenario includes the following assumptions:

1) The SCR is required by the end of 2023 instead of 2028 to reflect an expansion
of EPA regional air quality programs.
2) Units 1 & 2 shut down in 2018 rather than in 2022 and shift common facility
costs earlier than in the Expected Case.
3) Adding a fabric filter (baghouse) system to enhance particulate removal by the
end of 2023.
4) State of Montana to reduce carbon emissions beginning following the Clean
Power Plan’s mass based with new sources levels, but delayed until 2024.3

The annual cost between 2018 and 2037 is 3.7 percent higher in the High-Cost Colstrip
scenario as compared to the PRS. Instead of paying these higher costs, the plant could
retire by 2023. Table 12.4 shows the resource strategy for a 2023 Colstrip retirement to
avoid the High Cost Colstrip scenario assumptions. Shutting down the plant as compared
to the High Colstrip Cost scenario would save customers 0.35 percent over running the
plant for the remainder of the IRP study period. Figure 12.4 illustrates the cost and risk
of the portfolio compared to the PRS and the Expected Case’s Efficient Frontier. Both the
high cost and retirement scenarios result in higher customer costs, but early retirement
exposes customers to more volatile power supply costs. Figure 12.5 shows the annual
costs of the two scenarios compared to the PRS. Direct emissions for the PRS and the
2023 shutdown case are in Figure 12.6. Early retirement reduces emissions to 0.9 million
metric tons if natural gas-fired peakers replace Colstrip and Lancaster and the wholesale
market serves some customer energy needs. The implied carbon cost of shutting down
the plant between 2024 and 2037 by selecting the new resource strategy is an additional
$12.21 per metric ton using the change in cost and the change in Avista’s direct emissions
from this scenario. This in total with the pricing included in the market analysis, totals
$23.88 per metric ton.

3 The average shadow price of the stochastic studies is $11.67 per metric ton between 2024 and 2037.
$6.47 in 2024 and $26.89 in 2037. The 95th percentile price in in 2024 is $16.94 per metric ton and $60.16
in 2037.
SC_PR_3-2 Attachment A Page 191 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-7

Table 12.4: Colstrip Retires in 2023 Scenario Resource Strategy

Resource By End of
Year
ISO
Conditions
(MW)
Natural Gas Peaker 2023 143
Thermal Upgrades 2023-2037 34
Natural Gas Peaker 2026 288
Natural Gas Peaker 2030 96
Storage 2035 5
Total 566

Demand Response 2025-2037 44
Conservation (w/ T&D losses) 2018-2037 107

Figure 12.2: High-Cost Colstrip Retention Scenario Efficient Frontier

$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550
20
3
0

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Levelized Annual Power Supply Cost (2018-42, Millions)
Expected Case: Efficient Frontier
Expected Case: PRS
High Colstrip Cost: PRS
High Colstrip Cost: Retire Colstrip 2023
SC_PR_3-2 Attachment A Page 192 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-8

Figure 12.3: High-Cost Colstrip Scenarios Annual Cost

Figure 12.4: Greenhouse Gas Emissions: Retire Colstrip in 2023 versus PRS

$0
$100
$200
$300
$400
$500
$600
$700
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Po
w
e
r

S
u
p
p
l
y

C
o
s
t

(
M
i
l
l
i
o
n
s
)
High Colstrip Cost: PRS
High Colstrip Cost: Retire Colstrip 2023
Expected Case: PRS
0.0
0.5
1.0
1.5
2.0
2.5
3.0
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
t
r
i
c

T
o
n
s

(
M
i
l
l
i
o
n
s
)
High Colstrip Cost: Retire Colstrip 2023
Expected Case: PRS
SC_PR_3-2 Attachment A Page 193 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-9

Colstrip Reduction Scenario
The major challenge with shutting down Colstrip prior to the end of its operational life is
the cost to replace its generation capacity. An alternative to retiring Colstrip is reducing
its dispatch. Each owner has dispatch rights and may not shut off all delivery, unless each
owner agrees. If the owners could agree, or if a program’s design could reduce dispatch
within the constraints of each owner’s control, then this scenario could be a lower cost
approach to reduce emissions than plant closure.

For this scenario, a cap on emissions is set to 50 percent of Expected Case operations,
and the plant is not able to purchase additional allowances. This methodology creates a
carbon price for the emission reduction as described in Chapter 10. Figure 12.7, illustrates
the cost and risk changes of this scenario compared to the PRS and retiring Colstrip in
2030. The cost of dispatching Colstrip at a 50 percent level is 2.2 percent higher than the
Expected Case’s PRS. Retiring the plant in 2030 and replacing it with peakers is a 1.8
percent increase and replacing the plant with a CCCT is a 2.2 percent increase. Figure
12.8 shows the change in greenhouse gas emissions. Reducing dispatch to 50 percent
levels is nearly on par from the customer cost point of view of shutting down the resource,
but if the plant needed to reduce operations less than 50 percent, then keeping the plant
available is less costly.

Figure 12.5: 50 Percent Colstrip Dispatch Reduction Scenario Cost & Risk Comparison

$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
$350 $400 $450 $500 $550
20
3
7

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Levelized Annual Power Supply Cost (2018-42, Millions)
Expected Case: Efficient FrontierExpected Case: PRSColstrip Reduction: PRS
No Colstrip Case Colstrip Retires (2030- Peakers)
No Colstrip Case Colstrip Retires (2030- CCCT)
SC_PR_3-2 Attachment A Page 194 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-10

Figure 12.6: Colstrip Dispatch Reduction Scenario Greenhouse Gas Comparison

Other Resource Scenarios
Several other resource portfolio studies using the Expected Case’s market forecast
formed the following analyses. The portfolios show the financial impact of different
choices in meeting future resource deficits. Figure 12.9 shows the levelized cost and 2030
risk compared to the Efficient Frontier.

Market Scenarios
This plan includes two wholesale market portfolio scenarios; the first uses wholesale
market purchases to meet all resource deficits with a load adjustment assuming
conservation programs end. This scenario illustrates the cost to serve the system with
market resources. The second market scenario limits new resources to conservation and
wholesale market purchases. These scenarios show the cost and risk if the utility chooses
to depend on the wholesale market for its future needs. These portfolios estimate the
value of capacity in the PRS.

If Avista ended conservation programs and used the wholesale power market for all future
deficits, the cost to serve customers would be $22 million lower per year and market risk
would be $12 million higher than the PRS. Offering conservation programs saves
customers $23 million per year along with the market risk only being $1 million higher
than the PRS. This analysis indicates that conservation is a cost effective method to
reduce risk and cost to customers. It illustrates the cost to meet capacity requirements for
a reliable system adds $23 million per year to customer costs over depending on the
wholesale market place.

0.0
0.5
1.0
1.5
2.0
2.5
3.0
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
20
2
4
20
2
5
20
2
6
20
2
7
20
2
8
20
2
9
20
3
0
20
3
1
20
3
2
20
3
3
20
3
4
20
3
5
20
3
6
20
3
7
Me
t
r
i
c

T
o
n
s

(
M
i
l
l
i
o
n
s
)
Expected Case: PRS
Colstrip Reduction: PRS
No Colstrip Case Colstrip Retires (2030- Peakers)
No Colstrip Case Colstrip Retires (2030- CCCT)
SC_PR_3-2 Attachment A Page 195 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-11

Figure 12.7: Other Resource Strategy Portfolio Cost and Risk (Millions)

No New Thermal Resources Scenario
The No New Thermal Resources scenario meets future resource deficits without adding
carbon-emitting resources. It requires a mix of new resource options adding both capacity
and energy to the system. Table 12.5 outlines the resources selected to meet Avista
capacity and energy requirements. If Avista could not construct or purchase new thermal
resources, meeting capacity deficits would require new hydro and storage technologies,
along with increased conservation and demand response. Wind and solar resources
would meet energy requirements.

This scenario is 4.1 percent higher cost than the PRS per year over the IRP study period,
but the 2030 market risk is 2.7 percent lower. Greenhouse gas emissions are 22 percent
lower than the PRS, when taking into account the added renewables to the overall
system. This scenario would require additional reliability work to determine if storage
technology and the wholesale market could together meet reliability requirements. This
scenario assumes over 10 percent of peak load is met by 215 MW of storage capacity
and 645 MWh of storage capability. Avista will need to determine if current and large
amounts of additional storage can adequately serve customer needs.

$0
$10
$20
$30
$40
$50
$60
$70
$80
$90
$100
$350 $400 $450 $500 $550
20
3
0

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Levelized Annual Power Supply Cost (2018-42, Millions)
50% Colstrip Dispatch
No New Thermal Resources
Market meets all resource deficits
Market &
conservation
meet all
resoure
deficits
PRS
14% Summer Planning Margin
PRS w/ Lower Palouse Output
No New Thermal Resources & Colstrip Retires (2035)
New CCCT Replaces Lancaster
SC_PR_3-2 Attachment A Page 196 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-12

Table 12.5: No New Thermal Resource Scenario

Resource By End of
Year
ISO
Conditions
(MW)
Storage 2026 150
Thermal Upgrades 2026-2030 44
Storage 2026-2037 65
Wind (on system) 2030 50
Hydro Upgrades 2030 68
Solar 2030-2037 250
Total 627

Demand Response 2025-2037 47
Conservation (w/ T&D losses) 2018-2037 123

Extending the no new thermal resources scenario to the Colstrip shut down in 2035
scenario requires additional storage and renewable resources. Table 12.6 outlines the
resources selected to meet deficits in this case. This scenario results in significant
increases in storage, hydro upgrades and solar resources at a capital cost exceeding
$3.1 billion through 2037 compared to the $538 million included in the PRS.

The cost, assuming Avista decisions do not affect market prices, is 9.7 percent higher
than the PRS between 2018 and 2042. In 2036, the first full year of Colstrip retirement,
costs are 45 percent higher than the PRS, and 31 percent higher than replacing Colstrip
with natural gas-fired peakers. Power Supply Cost volatility is 25 percent lower in this
scenario than the PRS and 8 percent lower than replacing Colstrip with natural gas-fired
peakers in 2037. Greenhouse gas emissions are significantly lower. The direct
greenhouse gas emissions from Avista facilities fall to 596,000 metric tons in 2037, but
renewables added to the Avista system would offset these emissions.

Even though this scenario is attractive from an environmental point-of-view, it has
significant cost implications and reliability concerns. Additional studies are required to
validate if there are any reliability concerns with meeting loads without baseload
generation as a backstop during both poor hydro years and in peak winter conditions.

SC_PR_3-2 Attachment A Page 197 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-13

Table 12.6: No New Thermal Resource and Colstrip Replacement Scenario

Resource By End of Year ISO Conditions (MW)
Storage 2026 155
Thermal Upgrades 2026-2030 44
Storage 2027-2037 225
Wind (on system) 2030-2037 250
Solar 2030-2037 550
Hydro Upgrades 2035 148
Wind (Montana) 2036 100
Total 1,472

Demand Response 2025-2037 49
Conservation (w/ T&D losses) 2018-2037 124

Low Palouse Output Scenario
Currently, Avista does not anticipate needing additional renewables to meet the
Washington EIA due to control of Palouse Wind and ownership of Kettle Falls Generation
Station. Palouse Wind has delivered power for more than four years, but only one year
has delivered the anticipated energy output. This scenario studies if Avista would require
additional renewable energy if the generation continues to be below original expectations.
The results of the scenario analysis warrant no change in resource strategy due to the
inclusion of upgrades to Kettle Falls in the PRS. This analysis also indicates less REC
sales (revenue) would be a result of lower Palouse Wind production. Given these
conclusions, Avista will continue on its current EIA compliance path, but will continue to
monitor production levels for any significant changes.

Increased Summer Planning Margin Scenario
As explained earlier, in recent IRPs Avista has not included any summer planning margin
beyond expected load expectation and reserve requirements. This IRP adds a seven
percent summer planning margin to the mandatory reserve requirements based on the
shrinking regional capacity associated with the shutdown of coal plants. The seven
percent planning margin is half of the winter planning margin. This scenario tests the
potential requirement and portfolio changes for a 14 percent summer planning margin.
Although, Avista does not currently anticipate moving to a 14 percent summer margin
until the wholesale market fails to provide adequate capacity as determined by internal or
NPCC studies. This study shows no significant change to the resource strategy until after
2035. The minor changes accelerate thermal upgrades in the PRS, although after 2035
solar resources are cost effective to provide summer peak reduction.

New CCCT Replaces Lancaster Scenario
Previous IRP’s included a scenario regarding how the previous PRS compares to the new
PRS. Since this plan’s new resource acquisition is significantly different from prior plans
in both timing and resource choice, the best way to represent this type of analysis is by
including a new CCCT rather than CT’s to replace Lancaster as this is the major change
with this plan. The levelized cost for this scenario is higher than the PRS by 0.85 percent
SC_PR_3-2 Attachment A Page 198 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-14

and 10 percent lower in 2030. In the Efficient Frontier analysis shown in Figure 12.7
above, the portfolio’s cost and risk is to the right of the Efficient Frontier. Indicating there
are more optimal portfolios to achieve similar risk savings. Table 12.7 shows the resource
strategy selection for this scenario. It is possible the CCCT is lower cost compared to
other alternatives so this portfolio option should be considered in future RFPs.

Table 12.7: New CCCT Replaces Lancaster Scenario

Resource By End of Year ISO Conditions (MW)
CCCT 2026 285
Thermal Upgrades 2026-2037 34
Natural Gas Peaker 2030 47
Storage 2036 5
Total 371

Demand Response 2032-2037 35
Conservation (w/ T&D losses) 2018-2037 103

Washington State Emission Goal Analysis
The State of Washington has a goal to reduce greenhouse gas emissions to 20 percent
below 1990 levels by 2035. No legislation or pathway to achieve this goal is set at the
time of the 2017 IRP analysis. Details regarding how to account for emissions from market
purchases have not been determined. Lastly, allocation between Washington and Idaho
will need resolution. Ignoring these issues, Figure 12.10 shows Avista’s total direct
greenhouse gas emissions since 1990 and a 20-year forecast. Historical emissions are
volatile due to hydro variability and resource changes. Avista significantly reduced its
direct emissions in 2001 by selling its share of the Centralia coal plant, but emissions later
rose due to Coyote Springs 2 and the Lancaster PPA. Hydro volatility needs addressing
by any policy to reduce emissions because poor hydro years require thermal resources
to meet load needs and they increase emissions in the regional power system.

Avista anticipates direct emissions to remain near 1990 levels and begin to decline under
average water conditions, until reaching 20 percent below 1990 levels by 2035. After
2035, emissions begin to grow as Avista’s natural gas-fired facilities increase production
to meet load growth, unless future policies require changes to Avista’s dispatch or require
the purchase of allowances to comply with state regulations. The Colstrip Reduction
scenario level meets emission reduction goals. Retiring Colstrip in 2035 could reduce
emissions by 60 percent compared to 1990 levels.

SC_PR_3-2 Attachment A Page 199 of 205
Chapter 12 – Portfolio Scenarios

Avista Corp 2017 Electric IRP 12-15

Figure 12.8: Avista Direct Greenhouse Gas Emissions

0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
19
9
0
19
9
2
19
9
4
19
9
6
19
9
8
20
0
0
20
0
2
20
0
4
20
0
6
20
0
8
20
1
0
20
1
2
20
1
4
20
1
6
20
1
8
20
2
0
20
2
2
20
2
4
20
2
6
20
2
8
20
3
0
20
3
2
20
3
4
20
3
6
Mi
l
l
i
o
n

M
e
t
r
i
c

T
o
n
s
Avista Direct Emissions 1990 Level
25% Below 1990 Levels Colstrip Retires 2035
Colstrip Reduction PRS
SC_PR_3-2 Attachment A Page 200 of 205

SC_PR_3-2 Attachment A Page 201 of 205
Chapter 13–Action Items
Avista Corp 2017 Electric IRP
13. Action Items

The IRP is an ongoing and iterative process balancing regular publication timelines with
pursuing the best 20-year resource strategies. The biennial publication date provides
opportunities to document ongoing improvements to the modeling and forecasting
procedures and tools, as well as enhance the process with new research as the
planning environment changes. This section provides an overview of the progress made
on the 2015 IRP Action Plan and provides the 2015 Action Plan.

Summary of the 2015 IRP Action Plan
The 2015 Action Plan included three categories: generation resource related analysis,
energy efficiency, and transmission planning.

2015 Action Plan and Progress Report

Generation Resource Related Analysis
 Analysis of continued feasibility of the Northeast Combustion Turbine due to its age.
o Northeast is a 39 year old peaking unit permitted to run 100 run hours per
year per unit. This action item is to determine if the unit should be
available for the full 20-years of the IRP and if it should be considered for
a capacity upgrade described in Chapter 9. Avista determined Northeast is
a viable plant for the 20-year planning horizon. The plant has few
operating run hours and it is not expected to reach its next maintenance
cycle for hot gas path inspection due to run hour limitations. The unit is
designed and used to meet extreme peak load conditions and to provide
non-spinning reserves, it meets these needs at little cost to customers.

 Continue to review existing facilities for opportunities to upgrade capacity and
efficiency.
o Avista included several options to upgrade both hydro and thermal
generating facilities in this IRP, these options are identified in Chapter 4.
Further, Avista completed an upgrade to the Coyote Springs 2 facility in
2016, increasing winter peak capacity by 16 MW and increasing its
efficiency by 0.8 percent by utilizing a hot gas path upgrade during its
latest maintenance outage period.

 Increase the number of manufacturers and sizes of natural gas-fired turbines
modeled for the PRS analysis.
o Avista reviewed the thermal generation sizes and manufacturers when
selecting resources to model for this IRP. Given Avista’s new generation
capacity need is not until 2026, additional resources beyond those
identified in Chapter 4 are unnecessary at this time. Avista studied many
alternative natural gas-fired resources and selected the lowest cost and
sizeable resource to meet Avista’s deficits.

SC_PR_3-2 Attachment A Page 202 of 205
Chapter 13–Action Items
Avista Corp 2017 Electric IRP
 Evaluate the need for, and perform if needed, updated wind and solar integration
studies.
o Avista determined it is not necessary to update or develop variable
integration study at this time. This is due to the fact the generation and
pricing scenarios used from the previous study are still relevant. Further,
Avista prefers to conduct these updated studies using intra hour modeling
technology. This is currently being developed and may be available for the
2019 IRP.

 Participate and evaluate the potential to join a Northwest EIM.
o Avista is conducting a cost/benefit analysis associated with joining the
CAISO EIM. This analysis will be complete in the fall of 2017. Avista is
also evaluating other factors influencing the decision to join the CAISO
EIM. These include the reduction of near term market liquidity as other
utilities join the EIM and the additional integration of renewable resources
in our service territory. Avista anticipates making a decision on joining the
CAISO EIM and the associated timing by the end of 2017.

 Monitor regional winter and summer resource adequacy.
o Avista continues to monitor resource adequacy for both the Northwest and
Avista. Avista is concerned the region may not have adequate resources
given announcements of large baseload plants, further, new analysis
shown by the Northwest Power and Conservation Council show summer
peaking is starting to be a concern. Given this change, Avista
implemented a 7 percent planning margin in the summer (in addition to
operating reserves). Avista will continue to follow regional analysis by
participating in the Resource Adequacy Advisory Committee.

 Participate in state level implementation of the CPP.
o Since the 2015 IRP, the Clean Power Plan is on hold by the US Supreme
Court. Further, the new Federal Administration has appeared to pause the
Clean Power Plan. This IRP does assume many of the goals of the CPP
will ultimately be implemented at a later date.

Energy Efficiency and Demand Response
 Continue to study and quantify transmission and distribution efficiency projects as
they apply to EIA goals.
o This IRP includes new assumptions for T&D benefits based on new
analysis, as discussed in Chapter 5.

 Complete energy efficiency potential assessment on Avista’s generation facilities.
o Since the 2015 IRP, Avista has completed additional analysis on owned
generation facilities, further, the costs have come down as some projects
are lighting related. An updated analysis is provided in Chapter 5.

SC_PR_3-2 Attachment A Page 203 of 205
Chapter 13–Action Items
Avista Corp 2017 Electric IRP
Transmission and Distribution Planning
 Work to maintain Avista’s existing transmission rights, under applicable FERC
policies, for transmission service to bundled retail native load.
o Avista has maintained its existing transmission rights to meet native
customer load.

 Continue to participate in BPA transmission processes and rate proceedings to
minimize the costs of integrating existing resources outside of Avista’s service area.
o Avista is actively participating in the BPA transmission rate proceedings.

 Continue to participate in regional and sub-regional efforts to establish new regional
transmission structures to facilitate long-term expansion of the regional transmission
system.
o Avista staff participates in and leads many regional transmission efforts
including the Columbia Grid and the Northern Tier Transmission Group
Forums.

2017 IRP Two Year Action Plan
Avista’s 2017 PRS provides direction and guidance for the type, timing, and size of
future resource acquisitions. The 2017 IRP Action Plan highlights the activities planned
for possible inclusion in the 2019 IRP. Progress and results for the 2017 Action Plan
items are reported to the TAC and the results will be included in Avista’s 2019 IRP. The
2017 Action Plan includes input from Commission Staff, Avista’s management team,
and the TAC.

Generation Resource Related Analysis
 Continue to review existing facilities for opportunities to upgrade capacity and
efficiency.
 Model specific commercially available storage technologies within the IRP; including
efficiency rates, capital cost, O&M, life cycle, and ability to provide non-power supply
benefits.
 Update the TAC regarding the EIM study and Avista plan of action.
 Monitor regional winter and summer resource adequacy, provide TAC with additional
Avista LOLP study analysis.
 Update the TAC regarding progress regarding Post Falls Hydroelectric Project
redevelopment.
 Perform a study to determine ancillary services valuation for storage and peaking
technologies using intra hour modeling capabilities. Further, use this technology to
estimate costs to integrate variable resources.
 Monitor state and federal environmental policies effecting Avista’s generation fleet.

SC_PR_3-2 Attachment A Page 204 of 205
Chapter 13–Action Items
Avista Corp 2017 Electric IRP
Energy Efficiency and Demand Response
 Determine whether or not to move the T&D benefits estimate to a forward looking
value versus a historical value.
 Determine if a study is necessary to estimate the potential and costs for a winter and
summer residential demand response program and along with an update to the
existing commercial and industrial analysis.
 Use the utility cost test methodology to select conservation potential for Idaho
program options.
 Share proposed energy efficiency measure list with Advisory Groups prior to CPA
completion.

Transmission and Distribution Planning
 Work to maintain Avista’s existing transmission rights, under applicable FERC
policies, for transmission service to bundled retail native load.
 Continue to participate in BPA transmission processes and rate proceedings to
minimize costs of integrating existing resources outside of Avista’s service area.
 Continue to participate in regional and sub-regional efforts to facilitate long-term
economic expansion of the regional transmission system.
 IRP & T&D planning will coordinate on evaluating opportunities for alternative
technologies to solve T&D constraints.

SC_PR_3-2 Attachment A Page 205 of 205