20150831Avista 2015 IRP.pdf

Avista Corp.l4ll East Mission P.O.Box3727
Spokane. Washingon 99220-0500
Telephone [phone]
Toll Free 800-727-9170
August 28,2015
Jean D. Jewell, Secretary
Idaho Public Utilities Commission
Statehouse Mail
W . 472 Washington Street
Boise,Idaho 83720
RE: Avista Utilities 2015 Electric Integrated Resource Plan
Dear Ms. Jewell:
Per the Commission's Integrated Resource Plan Requirements outlined in Case No.U-1500-
165, Order No. 22299, Case No.GNR-E-93-1, Order No. 24729 and Case No.GNR-E-93-3, Order
No. 25260 , Avista Corporation dlblal Avista Utilities, hereby submits for filing an original
(unbound, unstapled copy), andT copies of its 2015 Electric Integrated Resource Plan (IRP). A copy
of the IRP and Appendices are being provided electronically with each hard copy of the IRP (inside
the front cover). Paper use and printing costs have been reduced by putting supporting documents on
the Company' s web site at www. avistautilities. com/resources/plans/electric. asp.
The Company submits the IRP to Public Utility Commissions in Idaho and Washington
every two years as required by state regulation. Avista regards the IRP as a methodology for
identifring and evaluating various resource options and as a process by which to establish a plan of
action for resource decisions.
The 2013 Plan is notable for the following:
o Population and employment growth are starting to recover from the Great
Recession.
) -_.)t-
;-.f
fb
I,1.-l ,:
-:t
o Natural gas-fired plants represent the largest portion of generation potential.
o The IRP anticipates the first new generation will be a natural gas-fired peaker, to
come online by the end of 2020 to replace expiring contracts and to serve load
growth.o Coal-fired generation at Colstrip remains a cost effective and reliable source of
power to meet the future energy needs of customerso Energy efficiency offsets more than half of projected load growth through the20-
year IRP timeframe.
Please direct any questions regarding this report to Clint Kalich at [phone].
Sincerely,
/t{Lh^da" Ge,nta,Ly
Linda Gervais
Manager, Regulatory Policy
Avista Utilities
509-495-497s
linda. servais@ avistacorp. com
2lP age

2015
Electric
Integrated Resource Plan
August 31, 2015
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 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.
Table of Contents

Avista Corp 2015 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 .................................................................................... 1-6
Greenhouse Gas Emissions .................................................................................................... 1-6
Action Items .............................................................................................................................. 1-8
2. Introduction and Stakeholder Involvement ................................................................ 2-1
IRP Process ............................................................................................................................. 2-1
2015 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 ................................................................................................ 3-14
Monthly Peak Load Forecast Methodology ............................................................................ 3-21
Simulated Extreme Weather Conditions with Historical Weather Data ................................. 3-22
Testing for Changes in Extreme Temperature Behavior ........................................................ 3-26
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 .................................................................................................................. 4-4
Power Purchase and Sale Contracts ....................................................................................... 4-6
Customer-Owned Generation ................................................................................................ 4-10
Solar ....................................................................................................................................... 4-11
5. Energy Efficiency & Demand Response ..................................................................... 5-1
Introduction ............................................................................................................................... 5-1
The Conservation Potential Assessment ................................................................................. 5-2
Overview of Energy Efficiency Potential .................................................................................. 5-4
Conservation Targets ............................................................................................................... 5-7
Energy Efficiency-Related Financial Impacts ........................................................................... 5-8
Integrating Results into Business Planning and Operations .................................................... 5-8
Demand Response ................................................................................................................. 5-11
Generation Efficiency Audits of Avista Facilities .................................................................... 5-15
6. Long-Term Position ....................................................................................................... 6-1
Introduction & Highlights .......................................................................................................... 6-1
Reserve Margins ...................................................................................................................... 6-1
Energy Imbalance Market ........................................................................................................ 6-8
Balancing Loads and Resources ............................................................................................. 6-9
Washington State Renewable Portfolio Standard .................................................................. 6-12
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
FERC Transmission Planning Requirements and Processes.................................................. 8-1
BPA Transmission System ....................................................................................................... 8-4
Avista’s Transmission System ................................................................................................. 8-4
Transmission System Information ............................................................................................ 8-5
Table of Contents

Avista Corp 2015 Electric IRP

ii
Distribution System Efficiencies ............................................................................................... 8-8
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 ......................................................................... 9-12
Thermal Resource Upgrade Options ..................................................................................... 9-15
Ancillary Services Valuation ................................................................................................... 9-16
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 ................................................................................................................. 10-25
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 ..................................................................................................... 11-15
Determining the Avoided Costs of Energy Efficiency ........................................................... 11-19
Determining the Avoided Cost of New Generation Options ................................................. 11-20
12. Portfolio Scenarios ...................................................................................................... 12-1
Introduction ............................................................................................................................. 12-1
Other Resource Scenarios ................................................................................................... 12-11
Resource Tipping Point Analyses ........................................................................................ 12-13
13. Action Items ................................................................................................................. 13-1
Summary of the 2013 IRP Action Plan................................................................................... 13-1
2013 Action Plan and Progress Report – Supplemental ........................................................ 13-3
2015 IRP Two Year Action Plan ............................................................................................. 13-5
Production Credits .................................................................................................................. 13-6

Table of Contents

Avista Corp 2015 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 ............................................. 1-4
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, 1971-2014 ....................................... 3-2
Figure 3.2: MSA Population Growth, 2007-2014 .......................................................................... 3-3
Figure 3.3: MSA Non-Farm Employment Breakdown by Major Sector, 2014 .............................. 3-4
Figure 3.4: MSA Non-Farm Employment Growth, 2007-2014 ..................................................... 3-4
Figure 3.5: MSA Personal Income Breakdown by Major Source, 2013 ....................................... 3-5
Figure 3.6: MSA Real Personal Income Growth, 1970-2013 ....................................................... 3-6
Figure 3.7: Forecasting IP Growth.............................................................................................. 3-10
Figure 3.8: Industrial Load and Industrial (IP) Index .................................................................. 3-10
Figure 3.9: Population Growth vs. Customer Growth, 2000-2014 ............................................. 3-11
Figure 3.10: Forecasting Population Growth .............................................................................. 3-12
Figure 3.11: Long-Run Annual Residential Customer Growth ................................................... 3-16
Figure 3.12: Load Scenarios with PV Shocks ............................................................................ 3-17
Figure 3.13: Load Growth Scenarios with PV Shocks................................................................ 3-17
Figure 3.14: Average Megawatts, High/Low Economic Growth Scenarios ................................ 3-19
Figure 3.15: UPC Growth Forecast Comparison to EIA ............................................................. 3-20
Figure 3.16: Load Growth Comparison to EIA ........................................................................... 3-20
Figure 3.17: Peak Load Forecast 2015-2035 ............................................................................. 3-24
Figure 3.18: Peak Load Forecast with 1 in 20 High/Low Bounds, 2015-2035 ........................... 3-25
Figure 4.1: 2016 Avista Capability & Energy Fuel Mix ................................................................. 4-1
Figure 4.2: Avista’s Net Metering Customers ............................................................................. 4-10
Figure 5.1: Historical and Forecast Conservation Acquisition (system) ....................................... 5-2
Figure 5.2: Analysis Approach Overview ..................................................................................... 5-3
Figure 5.3: Cumulative Conservation Potentials CPA versus PRiSM .......................................... 5-7
Figure 5.4: Existing & Future Energy Efficiency Costs and Energy Savings ............................... 5-8
Figure 6.1: 2020 Market Reliance & Capacity Cost Tradeoffs ..................................................... 6-4
Figure 6.2: Planning Margin Survey Results ................................................................................ 6-5
Figure 6.3: Single Largest Contingency Survey Results (2014 Peak Load) ................................ 6-6
Figure 6.4: 95th Percentile Capacity Requirements ..................................................................... 6-7
Figure 6.5: 99th Percentile Capacity Requirements ..................................................................... 6-8
Figure 6.6: Winter 1 Hour Capacity Load and Resources .......................................................... 6-10
Figure 6.7: Summer 18-Hour Capacity Load and Resources .................................................... 6-11
Figure 6.8: Annual Average Energy Load and Resources ......................................................... 6-12
Figure 7.1: Draft Clean Power Plan 2030 Emission Intensity Goals ............................................ 7-7
Figure 8.1: NERC Interconnection Map ....................................................................................... 8-2
Figure 9.1: Northwest Wind Project Levelized Costs per MWh ................................................... 9-6
Figure 9.2: Solar Nominal Levelized Cost ($/MWh) ..................................................................... 9-8
Figure 9.3: Historical and Planned Hydro Upgrades .................................................................. 9-13
Figure 9.4: Storage’s Value Stream ........................................................................................... 9-17
Figure 9.5: Avista’s Monthly Up/Down Regulation Surplus ........................................................ 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 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: 2030 Adjusted State Carbon Intensity CPP Goals ............................................... 10-11
Table of Contents

Avista Corp 2015 Electric IRP

iv
Figure 10.9: Historical Stanfield Natural Gas Prices (2004-2015) ........................................... 10-12
Figure 10.10: Stanfield Annual Average Natural Gas Price Distribution .................................. 10-13
Figure 10.11: Stanfield Natural Gas Distributions .................................................................... 10-14
Figure 10.12: Wind Model Output for the Northwest Region ................................................... 10-18
Figure 10.13: 2014 Actual Wind Output BPA Balancing Authority ........................................... 10-18
Figure 10.14: Mid-Columbia Electric Price Forecast Range .................................................... 10-21
Figure 10.15: Western States Greenhouse Gas Emissions ..................................................... 10-22
Figure 10.16: EPA’s CPP Annual Emissions Intensity for the West ........................................ 10-23
Figure 10.17: EPA’s CPP 2030 State Goal vs. Modeling Result ............................................. 10-23
Figure 10.18: Base Case Western Interconnect Resource Mix ............................................... 10-24
Figure 10.19: Annual Mid-Columbia Flat Price Forecast Benchmark Scenario ....................... 10-25
Figure 10.20: Benchmark Scenario Annual Western U.S. Greenhouse Gas Emissions ......... 10-26
Figure 10.21: Annual Mid-Columbia Flat Price Forecast Colstrip Retires Scenario ................ 10-27
Figure 10.22: No Colstrip Scenario Annual Western U.S. Greenhouse Gas Emissions ......... 10-27
Figure 10.23: Social Cost of Carbon Scenario Emission Prices .............................................. 10-28
Figure 10.24: Annual Mid-Columbia Flat Price Forecast Social Cost of Carbon Scenario ...... 10-29
Figure 10.25: Social Cost of Carbon Scenario Western US Greenhouse Gas Emissions ...... 10-29
Figure 10.26: Draft CPP as Proposed Scenario Flat Mid-Columbia Electric Prices ................ 10-30
Figure 10.27: Draft CPP as Proposed Scenario Western Greenhouse Gas Emissions .......... 10-31
Figure 10.28: Draft CPP as Proposed Scenario 1941 Water Year Annual Costs .................... 10-32
Figure 10.29: CPP as Proposed 1941 Water Year Scenario Mid-Columbia Electric Prices .... 10-33
Figure 11.1: Resource Acquisition History ................................................................................. 11-2
Figure 11.2: Physical Resource Positions (Includes Energy Efficiency) .................................... 11-3
Figure 11.3: REC Requirements vs. Qualifying RECs for Washington State EIA ..................... 11-4
Figure 11.4: Conceptual Efficient Frontier Curve ....................................................................... 11-6
Figure 11.5: New Resources Meets Winter Peak Loads............................................................ 11-8
Figure 11.6: Energy Efficiency Annual Expected Acquisition Comparison .............................. 11-10
Figure 11.7: Load Forecast with and without Energy Efficiency .............................................. 11-10
Figure 11.8: Avista Owned and Controlled Resource’s Greenhouse Gas Emissions ............. 11-12
Figure 11.9: Power Supply Expense Range ............................................................................ 11-14
Figure 11.10: Expected Case Efficient Frontier ........................................................................ 11-16
Figure 11.11: Risk Adjusted PVRR of Efficient Frontier Portfolios ........................................... 11-17
Figure 11.12: Risk Adjusted PVRR of Efficient Frontier Portfolios ........................................... 11-18
Figure 12.1: Linear versus Integer Efficient Frontier Difference ................................................. 12-2
Figure 12.2: Colstrip Retires Scenario Efficient Frontier Analysis .............................................. 12-5
Figure 12.3: Colstrip Retires in 2026 Scenario Power Supply Cost Impact ............................... 12-6
Figure 12.4: Colstrip Retires in 2027 Emissions ........................................................................ 12-6
Figure 12.5: High-Cost Colstrip Retention Scenario Efficient Frontier ....................................... 12-8
Figure 12.6: High-Cost Colstrip Scenarios Annual Cost ............................................................ 12-8
Figure 12.7: Social Cost of Carbon Impact to Efficient Frontier ................................................. 12-9
Figure 12.8: Colstrip Retires in 2027 Portfolio Efficient Frontier .............................................. 12-10
Figure 12.9: Colstrip Retires in 2027 Portfolio Emissions ........................................................ 12-10
Figure 12.10: Other Resource Strategy Portfolio Cost and Risk (Millions) .............................. 12-11
Figure 12.11: Risk Adjusted PVRR (2016- 2035) ..................................................................... 12-13
Figure 12.11: Utility Scale Solar Tipping Point Analysis (2014 $) ............................................ 12-14
Figure 12.13: Utility Scale Storage Tipping Point Analysis (2014 $) ........................................ 12-15

Table of Contents

Avista Corp 2015 Electric IRP

v
Table of Tables

Table 1.1: The 2015 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 ..................................................... 3-9
Table 3.2: Customer Growth Correlations, January 2005-December 2013 ............................... 3-11
Table 3.3: Average Annual PV Scenario Load Growth for Selected Periods ............................ 3-18
Table 3.4: High/Low Economic Growth Scenarios (2015-2035) ................................................ 3-18
Table 3.5: Load Growth for High/Low Economic Growth Scenarios (2015-2035) ..................... 3-19
Table 3.6: Forecasted Winter and Summer Peak Growth, 2015-2035 ...................................... 3-24
Table 3.7: Energy and Peak Forecasts ...................................................................................... 3-25
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 .................................................................................................... 4-9
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) ............................. 5-7
Table 5.3: Commercial and Industrial Demand Response Achievable Potential (MW) ............. 5-13
Table 6.1: Washington State EIA Compliance Position Prior to REC Banking .......................... 6-13
Table 8.1: 2015 IRP Requested Transmission Upgrade Studies ................................................. 8-7
Table 8.2: Third-Party Large Generation Interconnection Requests ............................................ 8-7
Table 8.3: Completed and 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 ................................................................................... 9-7
Table 9.5: Storage Power Supply Value .................................................................................... 9-17
Table 9.6: Natural Gas-Fired Facilities Ancillary Service Value ................................................. 9-18
Table 10.1: AURORAXMP Zones ................................................................................................. 10-2
Table 10.2: Added Northwest 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 ................................................... 10-15
Table 10.7: Area Load Coefficient of Determination (Standard Deviation/Mean) .................... 10-15
Table 10.8: Area Load Coefficient of Determination (Standard Deviation/Mean) .................... 10-16
Table 10.9: Annual Average Mid-Columbia Electric Prices ($/MWh) ....................................... 10-21
Table 11.1: Qualifying Washington EIA Resources ................................................................... 11-4
Table 11.2: 2015 Preferred Resource Strategy .......................................................................... 11-8
Table 11.3: 2013 Preferred Resource Strategy .......................................................................... 11-9
Table 11.4: PRS Rate Base Additions from Capital Expenditures ........................................... 11-13
Table 11.5: Avista Medium-Term Winter Peak Hour Capacity Tabulation ............................... 11-15
Table 11.6: Avista Medium-Term Summer 18-Hour Sustained Peak Capacity Tabulation ..... 11-15
Table 11.7: Alternative Resource Strategies along the Efficient Frontier (MW) ....................... 11-19
Table 11.8: Updated Annual Avoided Costs ($/MWh).............................................................. 11-21
Table 12.1: Efficient Frontier with Linear Programming ............................................................. 12-2
Table 12.2: Load Forecast Scenarios (2016-2035) .................................................................... 12-3
Table 12.3: Resource Selection for Load Forecast Scenarios ................................................... 12-3
Table 12.4: Colstrip Retires in 2026 Scenario Resource Strategy ............................................. 12-5
Table 12.5: Colstrip Retires in 2022 Scenario Resource Strategy ............................................. 12-7

Avista Corp 2015 Electric IRP
2015 Electric IRP Introduction

Avista has a 125-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 2015 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 2015 IRP shows Avista has adequate resources between owned and contractually
controlled generation, combined with conservation and market purchases, to meet
customer needs through 2020. In the longer term, plant upgrades, energy efficiency
measures, and additional natural gas-fired generation are integral parts of Avista’s 2015
Preferred Resource Strategy.

Changes
Major changes from the 2013 IRP include:
 Average annual load growth reduced to 0.6 percent from just over 1 percent in
2013. This combined with a short term purchase power agreement delays the
need for a new natural gas-fired resource by one year.
 Less contribution from natural gas-fired peakers due to lower projected loads.
 The elimination of demand response (temporarily reducing the demand for
energy) due to higher estimated costs.

Highlights
Some highlights of the 2015 IRP include:
 Population and employment growth is starting to recover from the Great
Recession.
 Natural gas-fired plants represent the largest portion of generation potential.
 The first anticipated resource acquisition is a natural gas-fired peaker by the end
of 2020 to replace expiring contracts and to serve load growth.
 Colstrip remains a cost effective and reliable source of power to meet future
customer needs.
 Energy efficiency offsets more than half of projected load growth through the 20-
year IRP timeframe.

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 mandates. Avista’s professional energy analysts
use sophisticated modeling tools and input from over 75 invited participants to develop
each plan. The participants in the public process include customers, academics,
Avista Corp 2015 Electric IRP
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 2017 IRP in early 2016. 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.
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-1

1. Executive Summary

Avista’s 2015 Electric Integrated Resource Plan (IRP) guides its resource strategy over
the next two years and resource 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 2015 Preferred Resource Strategy (PRS)
includes energy efficiency, generation upgrades, and new natural gas-fired generation.

PRS development depends on modeling techniques to balance cost, reliability, rate
volatility, and renewable resource requirements. Avista’s management and the
Technical Advisory Committee (TAC) guide its development and the IRP document by
providing 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 experiences its highest peak loads in the
winter. Its peak planning methodology includes operating reserves, regulation, load
following, wind integration, and a 14 percent planning margin over winter-peak load
levels. The company has adequate resources, combined with conservation and market
purchases, to meet peak load requirements through 2020. Figure 1.1 shows Avista’s
resource position through 2035.

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

-
500
1,000
1,500
2,000
2,500
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
Me
g
a
w
a
t
t
s
Existing Resources Load w/ Conservation + Cont.
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-2

A short-term capacity need exists in the winter of 2015-2016, but is short-lived due to a
150 MW capacity sale contract ending in 2016. Avista addressed this deficit with market
purchases; so, the first long-term capacity deficit begins in 2021. Resources acquired to
meet projected winter deficiencies will provide capacity in excess of summer needs.
Chapter 6 – Long Term Position details Avista’s resource needs.

Modeling and Results
Avista uses a multiple-step approach to develop its PRS. It begins by identifying and
quantifying potential new generation resources to serve projected electricity demand
across the West. This Western Interconnect-wide study determines the impact of extra-
regional markets on the Northwest electricity marketplace of which Avista is a part. It
then maps existing Avista resources to the transmission grid in a model simulating
hourly operations for the Western Interconnect from 2016 to 2035, the IRP study
timeframe. The model adds new resources and transmission to the Western
Interconnect as regional loads grow and older resources are retired. 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 2035.

Electricity and Natural Gas Market Forecasts
Figure 1.2 shows the 2015 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 $38.48 per MWh in nominal dollars over the 2016-2035 timeframe.

Figure 1.2: Average Mid-Columbia Electricity Price Forecast

$0
$20
$40
$60
$80
$100
$120
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
1
6
-
3
5
$
p
e
r

M
W
h
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-3

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 $4.97 per dekatherm over the next 20
years. See Chapter 10 – Market Analysis for details on the natural gas and electricity
price forecasts.

Figure 1.3: Stanfield Natural Gas Price Forecast

Energy Efficiency Acquisition
Avista commissioned a 20-year Conservation Potential Assessment in 2015. The study
analyzed over 3,000 equipment and 2,300 measure options for residential, commercial,
and industrial energy efficiency applications. Data from this study formed the basis of
the IRP conservation potential evaluation. Figure 1.4 shows how historical efforts in
energy efficiency have decreased Avista’s load requirements by 127 aMW, or
approximately eleven percent of its total load in 2014. The cumulative line shows the
summation of all efficiency acquisitions and the online 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.

$0
$2
$4
$6
$8
$10
$12
$14
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
1
6
-
2
0
3
5
Do
l
l
a
r
s

p
e
r

D
e
k
a
t
h
e
r
m
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-4

Figure 1.4: Annual and Cumulative Energy Efficiency Acquisitions

Preferred Resource Strategy
The PRS results from careful consideration by Avista’s management and the TAC of
information gathered and analyzed in the IRP process. It meets future load growth with
upgrades at existing generation facilities, energy efficiency, and natural gas-fired
technologies, as shown in Table 1.1.
Table 1.1: The 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 Efficiency 2016-2035 193 132
Distribution Efficiencies <1 <1
Total 193 132

0
40
80
120
160
200
0
4
8
12
16
20
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
Cu
m
u
l
a
t
i
v
e

S
a
v
i
n
g
s

(
a
M
W
)
An
n
u
a
l

S
a
v
i
n
g
s

(
a
M
W
)
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-5

The 2015 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 2013 IRP include a reduced contribution from natural gas-fired peakers and the
elimination of demand response because of lower projected load growth, more thermal
plant upgrades and higher demand response costs.

Each new resource and energy efficiency option is valued against the Expected Case
Mid-Columbia electricity market to identify its future value, as well as its inherent risk
measured by year-to-year portfolio 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 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. An 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 expected 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, 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

$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
2
7

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 2016-2040
Least Cost
Preferred Resource Strategy
Least Risk
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-6

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 voters approved the Energy Independence Act (EIA) through Initiative 937
in the November 2006 general election. The EIA requires utilities with over 25,000
customers to meet three percent of retail load from qualified renewable resources by
2012, nine percent 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, Kettle Falls Generating
Station output and renewable energy certificate (REC) purchases. Figure 1.6 shows
Avista’s EIA-qualified generation; Chapter 6 – Long-Term Position includes a more in-
depth discussion of this topic.

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

Greenhouse Gas Emissions
The regulation of greenhouse gases, or carbon emissions, is in various stages of
development and implementation throughout the country. Some states have active cap
and trade programs, emissions performance standards, renewable portfolio standards,
or a combination of active and proposed regulations affecting emissions from electric
generation resources. The Environmental Protection Agency’s (EPA) June 2014 Clean
Power Plan (CPP) draft proposal aimed to reduce greenhouse gas emissions from
-
20
40
60
80
100
120
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
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Banking Kettle Falls
Palouse Wind Hydro Upgrades
Requirement
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-7

existing fossil-fueled electric generating units by establishing state-by-state emission
rate targets calculated based on four building blocks. The EPA issued the final CPP rule
on August 3, 2015, which was after modeling for this IRP was completed. The analysis
of the final CPP rule, and subsequent state implementation plans, will occur in the 2017
IRP. The 2015 IRP reduces emissions consistent with the EPA draft rule. All active
regulations affecting generation in the Western Interconnect are included in the IRP,
including a $12 per metric ton carbon cost that escalates over time. Figure 1.7 shows
Avista’s projected greenhouse gas emissions for its existing and new generation assets.

Figure 1.7 shows that Avista emissions will increase modestly over the IRP timeframe.
Figure 1.8 shows that, unlike Avista, western-region emissions likely will fall from
historic levels. This discrepancy occurs because Avista does not own any of the less-
cost-effective coal and natural gas-fired plants projected to retire over the IRP
timeframe. More details on state and federal greenhouse gas policies are in Chapter 7.
Results of greenhouse-gas policy scenarios are in Chapter 12.

Figure 1.7: 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
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
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
Chapter 1- Executive Summary
Avista Corp 2015 Electric IRP 1-8

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

Action Items
The 2015 Action Items chapter updates progress made on Action Items in the 2013 IRP
and outlines activities Avista intends to perform between the publication of this report
and publication of the 2017 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
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
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
111(d) plants
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 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 2015 IRP is Avista’s fourteenth 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 2015 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 more
than 75 invited 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 2015 IRP. The first meeting was on May 29,
2014; the last occurred on June 24, 2015. 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 http://www.avistautilities.com/inside/resources/irp/electric. 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, Case No. GNR-E-93-1, Order No.
24729, and Case No. GNR-E-93-3, Order No. 25260.
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 Electric IRP
Table 2.1: TAC Meeting Dates and Agenda Items

Meeting Date Agenda Items
TAC 1 – May 29, 2014  TAC Meeting Expectations
 2013 IRP Commission Acknowledgements
 2013 Action Plan Update
 Energy Independence Act Compliance
 Pullman Energy Storage Project
 Demand Response Study Discussion
 Draft 2015 Electric IRP Work Plan
TAC 2 – September 23, 2014  Introduction & TAC 1 Recap
 Conservation Selection Methodology
 Load and Economic Forecast
 Shared Value Report
 Generation Options
 Clean Power Plan Proposal Discussion
TAC 3 – November 21, 2014  Introduction & TAC 2 Recap
 Planning Margin
 Colstrip Discussion
 Cost of Carbon
 IRP Modeling Overview
 Conservation Potential Assessment
TAC 4 – February 24, 2015  Introduction & TAC 3 Recap
 Demand Response Study
 Natural Gas Price Forecast
 Electric Price Forecast
 Resource Requirements
 Interconnection Studies
 Market Scenarios and Portfolio Analysis
TAC 5 – May 19, 2015  Introduction & TAC 4 Recap
 Review of Market Futures
 Ancillary Services Valuation
 Conservation Potential Assessment
 Draft 2015 PRS & Portfolio Analysis
TAC 6 – June 24, 2015  Introduction & TAC 5 Recap
 Avista Community Solar
 2015 Action Plan
 Final 2015 PRS
 2015 IRP Document Introduction

Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 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 2015 IRP TAC process.

Table 2.2: External Technical Advisory Committee Participating Organizations

Organization
AEG
As You Sow
Birch Energy Economics
City of Spokane
Clearwater Paper
Earth Justice
Eastern Washington University
Eugene Water & Electric Board
GE Energy
Gonzaga University
Grant PUD
Idaho Department of Environmental Quality
Idaho Public Utilities Commission
Inland Empire Paper
Montana Environmental Information Center
NW Energy Coalition
PacifiCorp
Pend Oreille PUD
Puget Sound Energy
Pullman City Council
Renewable Northwest
Residential and Small Commercial Customers
Resource Development Associates
Sierra Club
Spokane Neighborhood Action Partners
The Energy Authority
Washington State Office of the Attorney General
Washington Department of Enterprise Services
Washington State Department of Commerce
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.
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 Electric IRP
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
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.

2015 IRP Outline
The 2015 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 2015 IRP.

Chapter 2: Introduction and Stakeholder Involvement
This chapter introduces the IRP and details public participation and involvement in the
IRP planning process.
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 Electric IRP
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.

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 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.

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 2015 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 2013 IRP. It
details the action items Avista will focus on between publication of this plan and the next
one.
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 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
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

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
–

–

–

–

–

–

Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 Electric IRP
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 11- Preferred Resource Strategy
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
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
Chapter 2- Introduction and Stakeholder Involvement
Avista Corp 2015 Electric IRP
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

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) – Plan includes a two-
year action plan that implements 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. (Description not
required)
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) – The integration of the
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

Chapter 3: Economic & Load Forecast

Avista Corp 2015 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 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 explains that when regional employment growth is
stronger than U.S. growth over the business cycle, its cause is increased in-migration.
The reverse holds true. Figure 3.1 shows annual population growth since 1971. 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
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 2015 Expected Case energy forecast grows 0.6 percent per year,
replacing the 1.0 percent annual growth rate in the 2013 IRP.
 Peak load growth is higher than energy growth, at 0.74 percent in the winter
and 0.85 percent in the summer.
 Retail sales and residential use per customer forecasts continue to decline
from 2013 IRP projections.
 Testing performed for this IRP shows that historical extreme weather events
are valid for peak load modeling.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-2
to 2013. Accelerating service area employment growth in 2013 helped push population
growth above one percent in 2014.

Figure 3.1: MSA Population Growth and U.S. Recessions, 1971-2014

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 population growth
in 2014 relative to the U.S. The association of employment growth to population growth
has a one year lag. That is, 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.

3 Data Source: Bureau of Economic Analysis and U.S. Census.

-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
An
n
u
a
l

G
r
o
w
t
h
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-3
Figure 3.2: MSA Population Growth, 2007-2014

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 MSAs.4 Approximately 70 percent of employment in the three
MSAs is in private services, followed by government (18 percent) and private goods-
producing sectors (13 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.

Between 1990 and 2007 non-farm employment growth averaged 2.7 percent per year.
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. By the fourth quarter 2014, service area employment growth began
exceeding U.S. growth rates.
4 Data Source: Bureau of Labor and Statistics 5 Data Source: Bureau of Labor and Statistics.

1.8%
1.4%
1.1%
0.7%
0.6%0.6%
0.8%
1.2%
1.0%1.0%
0.9%0.8%
0.7%0.7%0.7%0.7%
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
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
U.S.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-4
Figure 3.3: MSA Non-Farm Employment Breakdown by Major Sector, 2014

Figure 3.4: MSA Non-Farm Employment Growth, 2007-2014

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
6 Data Source: Bureau of Economic Analysis.
Private Goods
Producing, 13%
Private Service
Producing, 69%
Federal Government,
2%
State Government,
4%
Local Government,
12%
2.2%
0.5%
-4.6%
-1.6%
0.4%0.7%
2.1%
1.7%
1.1%
-0.6%
-4.3%
-0.7%
1.2%
1.7%1.7%1.9%
-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
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
U.S.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-5
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.

Figure 3.5: MSA Personal Income Breakdown by Major Source, 2013

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 22 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 roughly 60 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
for Avista’s service area and the U.S. Note that in the 1980-90 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. As a result, 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,
56%
Dividends,
Interest, and
Rent, 22%
Transfer
Receipts, 22%
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-6
Figure 3.6: 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 a set of 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.

8 For schedules representing a single customer, were there is no customer count and for street lighting,
total load is forecast directly without first forecasting UPC.
2.3%
1.4%
2.5%
0.7%
1.4%
2.1%
2.3%2.4%
0.7%
1.8%
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 2013
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.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-7
Equation 3.1: Generating Schedule Total Load

Where:

 = the forecast for month t, year j = 1,…,5 beyond the
current year, yc ,for schedule s.
 = the UPC forecast.
 = the customer forecast.

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

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); average household size (AHS); 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.

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.

9 Faruqui, Ahmad (2000). Making Forecasts and Weather Normalization Work Together, Electric Power
Research Institute, Publication No. 1000546, Tech Review, March 2000.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-8
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.

The choice of a 20-year moving average for defining normal weather reflects several
factors. First, recent climate research from the National Aeronautic 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 that summer
temperatures in the Northern Hemisphere have increased about 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 1951-1981 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 10
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, AHS, and IP appear in Equation 3.2, then they must also be
forecasted for five years to generate the UPC forecast. The assumption in the five-year
forecast for this IRP is that RAP will increase two percent annually. This rate reflects the
average annual real growth rate for the 2005-2013 period. AHS is constant at the 2012
level.11 This reflects the relative stability of AHS over the 2006-2013 period. Table 3.1
shows the schedules using these three drivers.

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 11 AHS only appears in the forecast equation for Washington Schedule 1 UPCAHS is not a statistically
significant predictor of UPC and the sign on the estimated regression coefficient is not stable for Idaho
Schedule 1.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-9
Table 3.1: UPC Models Using Non-Weather Driver Variables

Schedule Variables Comment
Washington:
Residential Schedule 1 RAP, AHS
Commercial Schedule 31 RAP Commercial pumping schedule
Industrial Schedule 31 RAP
Industrial Schedules 11, 21, and 25 IP
Idaho:
Residential Schedule 1 RAP AHS not a statistically significant or
stable driver
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.

Equation 3.3: IP Regression Equation

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.12, 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.

12 Data Source: U.S. Federal Reserve and Avista records. 13 Figure 3.8 excludes one large industrial customer with significant load volatility.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-10
Figure 3.7: Forecasting IP Growth

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 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
70
80
90
100
110
120
70 GWh
80 GWh
90 GWh
100 GWh
110 GWh
120 GWh
Ja
n
-
9
7
Au
g
-
9
7
Ma
r
-
9
8
Oc
t
-
9
8
Ma
y
-
9
9
De
c
-
9
9
Ju
l
-
0
0
Fe
b
-
0
1
Se
p
-
0
1
Ap
r
-
0
2
No
v
-
0
2
Ju
n
-
0
3
Ja
n
-
0
4
Au
g
-
0
4
Ma
r
-
0
5
Oc
t
-
0
5
Ma
y
-
0
6
De
c
-
0
6
Ju
l
-
0
7
Fe
b
-
0
8
Se
p
-
0
8
Ap
r
-
0
9
No
v
-
0
9
Ju
n
-
1
0
Ja
n
-
1
1
Au
g
-
1
1
Ma
r
-
1
2
Oc
t
-
1
2
Ma
y
-
1
3
De
c
-
1
3
Ju
l
-
1
4
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
Average GDP
Growth Forecasts:
 IMF, FOMC,
Bloomberg, etc.
 Average
forecasts out 5-
yrs.
U.S Industrial Production
Index (IP) Growth Model:
 Model links year y GDP
growth year y IP
growth.
 Federal Reserve
industrial production
index is measure of IP
growth.
 Forecast out 5-yrs.
Generate Average, High,
and Low IP Forecast:
 Forecast annual IP
growth using the GDP
forecast average.
 Convert annual
growth scenario to a
monthly basis to
project out the
monthly level of the IP
index.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-11
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.

Table 3.2: Customer Growth Correlations, January 2005-December 2013

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

Figure 3.9 shows the relationship between annual population growth and year-over-year
customer growth.14 For the last 15 years electricity customer growth has closely
followed population growth in the combined Spokane-Kootenai MSAs. Both population
and customer growth have averaged 1.2 percent annually over the 2000-14 period.

Figure 3.9: Population Growth vs. Customer Growth, 2000-2014

Figure 3.9 demonstrates that 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. That is, for schedule 1
14 Data Source: Bureau of Economic Analysis, U.S. Census, and Avista records.
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
An
n
u
a
l

G
r
o
w
t
h
Avista WA-ID MSAs
System Customers
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-12
in Washington and Idaho, an Expected Case forecast is made using an ARIMA times-
series model. 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 3.10: 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

Equation 3.5: Kootenai Employment Forecast

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.
Average GDP
Growth Forecasts:
 IMF, FOMC,
Bloomberg,
etc.
 Average
forecasts out
5-yrs.
Non-farm Employment
Growth Model:
 Model links year y, y-
1, and y-2 GDP
growth to year y
regional employment
growth.
 Forecast out 5-yrs.
 Averaged with GI
forecasts.

Regional Population Growth
Models:
 Model links regional, U.S.
growth to Spokane and
Kootenai population growth.
 Forecast out 5-yrs for
Spokane, WA; and Kootenai,
ID
 Averaged with IHS
forecasts.
 Growth rates used to adjust
Expected Case ARIMA
customer forecasts for WA-
ID Residential Schedule 1
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-13
 DKC and DHB = structural change (SC) dummy variables for the closing
of Kaiser Aluminum in Spokane.
 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.

The same average GDP growth forecasts used for the IP growth forecasts are inputs to
generate five-year employment growth forecasts. Employment forecasts are averaged
with IHS Connect’s (formerly Global Insight) forecasts for the same counties. Averaging
reduces 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

Equation 3.7: Kootenai Population Forecast

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 HIS forecasts) become inputs to Equations 3.6 and 3.7 to
generate population growth forecasts. These forecasts, averaged with IHS’s forecasts
for the same MSAs, produce a final population forecast. This population growth forecast
is used to adjust the Expected Case ARIMA generated forecasts for Residential
Schedule 1 customers. This adjustment reconciles forecasted growth with forecasted
population growth.

Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-14
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 a growing electric vehicle (EV) and residential rooftop photovoltaic solar
(PV) fleets. The long-run modeling approach starts with Equation 3.8.

Equation 3.8: Residential Long-Run Forecast Relationship

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.15 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 pays attention to varying cy for
economic reasons and uy due to increased PV penetration.

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 2020 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 2019 is approximately 1 percent, assuming a gradual
decline starting in 2020. This value was generated with the Employment and
Population forecast Equations 3.4, 3.5, 3.6, and 3.7 in conjunction with IHS
Connect’s employment and population forecasts for the 2020-2024 period. The
Expected Case assumes long-run U.S. employment growth of approximately 1.4
percent and service area employment growth of approximately 1.5 percent. These
numbers result from assumed U.S. long-run GDP growth of approximately 2.4
percent. 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 assumption of high correlation is consistent with the high
historical correlation between residential and commercial load growth. The 0.5
15 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.

Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-15
percent spread is in 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. For 2020-2035, annual
industrial load growth is set at 0.5 percent and streetlight load growth at 0.1 percent.
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 2026. Up to
2026, this is the same as the nominal price increasing 4 percent a year assuming a
non-energy inflation rate of 2 percent. The real price increase assumption is zero
starting in 2026. 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 -0.20. 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. The AHS-elasticity of UPC is set at 2.3. This assumes AHS is constant up to 2025,
then starts to slowly decline through 2040. AHS-elasticity estimates are from the
five-year UPC forecast equations for Residential Schedule 1 in Washington and
Idaho, using the customer-weighted average between Washington and Idaho.

7. From 2020 to 2023, 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 2014 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.

8. Electric vehicles grow at a rate consistent with present adoption rates. Using Electric
Power Research Institute data, Avista estimates that as of 2015 there are around
400 EVs registered in its service area. The forecasted rate of adoption over the
2020-2035 period is a function of forecasted residential customer growth over the
same period. The EV adoption rate assumption uses historical data for the 2010-
2013 period to establish the relationship between residential customers and EVs.
This analysis shows that for every 100 residential customers added, approximately
three new registered EVs are added to the Avista service area. However, since
Avista does not serve 100 percent of all loads in the counties it serves, so this
adoption rate is reduced by 50 percent. Each EV uses 2,500 kWh per year in the
forecast.

9. 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
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-16
average PV system is forecast at the current median of 3.0 kilowatts and a 13
percent capacity factor. As of 2014, residential PV penetration was about 0.06
percent. The growth assumption is approximately 0.01 percent per year to 2040,
resulting in a 2035 penetration rate of 0.29 percent. This slow rate of PV penetration
growth is consistent with recent history.

Figure 3.11: Long-Run Annual Residential Customer Growth

Load Scenarios with PV
In addition to the Expected Case forecast, three alternatives illustrate the impacts of
varying PV penetration by 2025: 1 percent (low shock scenario); 5 percent (medium
shock scenario); and 10 percent (high shock scenario). In each scenario, the
penetration rate is constant after 2025. Each shock case assumes that the PV system
size grows each year so that by 2035 the typical system size equals 5 kilowatts. All
remaining assumptions in the PV penetration cases remain unchanged from the
Expected Case. Figure 3.12 presents results of the Expected Case and shock
scenarios. Figure 3.13 shows the annual growth rate in the load shown in Figure 3.12.
In all PV scenarios, load growth returns to the Expected Case by 2026 when the
penetration rate stabilizes. Table 3.3 shows the average annual PV scenario growth
rates in native load for the five-year forecast and long-run forecast.

0.7%
0.8%
0.9%
1.0%
1.1%
1.2%
1.3%
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
An
n
u
a
l

G
r
o
w
t
h
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-17
Figure 3.12: Load Scenarios with PV Shocks

Figure 3.13: Load Growth Scenarios with PV Shocks

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

M
e
g
a
w
a
t
t
s
Expected Case
Exponential Low Shock
Exponential Medium Shock
Exponential High Shock
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
0.5%
1.0%
1.5%
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
An
n
u
a
l

G
r
o
w
t
h
Expected Case
Exponential Low Shock
Exponential Medium Shock
Exponential High Shock
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-18
Table 3.3: Average Annual PV Scenario Load Growth for Selected Periods

PV Scenario 2015-2019
(Percent)
2020-2035
(Percent)
2015-2035
(Percent)
Expected Case (0.1%) 0.73 0.47 0.53
Low Shock (1%) 0.73 0.46 0.52
Medium Shock (5%) 0.73 0.38 0.46
High Shock (10%) 0.73 0.28 0.39

The model suggests that with PV penetration between 0.3 percent and 1 percent, load
growth after 2020 averages around 0.5 percent, a slight decrease from the 0.6 percent
assumption in the Expected Case. Penetration rates 5.0 percent and higher result in
noticeable load growth declines.

Native Load Scenarios with Low/High Economic Growth
Native load changes in the PV scenarios because of varying PV growth assumptions.
For load growth scenarios, Expected Case PV assumptions remain constant while
regional economic growth levels vary. The high and low scenarios use population
growth Equations 3.6 and 3.7, holding U.S. employment growth constant at 1.4 percent,
but varying MSA employment growth at higher and lower levels gauges the impacts on
population growth and utility loads. See Table 3.4. 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 3.14.

Table 3.4: High/Low Economic Growth Scenarios (2015-2035)

Economic
Growth
Annual U.S.
Employment Growth
(percent)
Annual Service Area
Employment Growth
(percent)
Annual Population
Growth
(percent)
Expected Case 1.4 1.5 1.0
High Growth 1.4 2.3 1.6
Low Growth 1.4 0.7 0.8

Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-19
Figure 3.14: Average Megawatts, High/Low Economic Growth Scenarios

Table 3.5 is the average annual load growth rate over the 2015-2035 period. The low
growth scenario predicts a slight load decline over 2020-2022 due to the impact of the
phased-in efficiency standards discussed in Item 7 of the Expected Case Assumptions
listed above.

Table 3.5: Load Growth for High/Low Economic Growth Scenarios (2015-2035)

Economic Growth Average Annual Native Load
Growth
(percent)
Expected Case 0.53
High Growth 0.83
Low Growth 0.23

Long-Run Forecast Residential Retail Sales
Focusing on residential kWh sales, Figure 3.15 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 2014 Annual Energy Outlook. Avista’s
forecast never shows positive UPC growth; in contrast, the EIA forecasts positive UPC
growth returning in 2033. The EIA forecast reflects a population shift to warmer-climate
states where air conditioning is typically required most of the year.

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

M
e
g
a
w
a
t
t
s
Expected Case
High Economic Growth
Low Economic Growth
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-20
Figure 3.15: UPC Growth Forecast Comparison to EIA
Figure 3.16 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.

Figure 3.16: Load Growth Comparison to EIA

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

G
r
o
w
t
h
EIA Refrence Case Use Per Household Growth
Expected Case's UPC Growth
-1.0%
-0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
An
n
u
a
l

G
r
o
w
t
h
EIA Purchased Residential Load Growth
Expected Case's Load Growth
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-21
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. This
said, 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

Where:


= 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.
 and = heating and cooling degree days the day before the
peak.
 = squared value of HDDd,t,y. and = heating
and cooling degree days the day before the peak.
 = maximum peak day temperature minus 65 degrees. This term
provides a better model fit than the square of CDD.
 = 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 and February 2012.
 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 out N years by using forecasted levels of GDP as shown in Equation
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-22
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:

For summer, this is defined as the forecasted series 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.)

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

Where:

 = 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.

Similarly, the model for cooling degree days is:
16 Forecasted GDP is generated by applying the averaged GDP growth forecasts used for the employment and
industrial production forecasts discussed previously.
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-23
Equation 3.11: Peak Load Simulation Equation for Summer Months

Where:

 = 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.

Given 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

Equation 3.13: Current Year Peak Load for Summer Months

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:

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.

Table 3.6 shows estimated peak load growth rates with and without the two large
industrial customers. Figure 3.17 shows the forecasted time path of peak load out to
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-24
2040, and Figure 3.18 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 3.6: Forecasted Winter and Summer Peak Growth, 2015-2035

Category Winter
(Percent)
Summer
(Percent)
Excluding Large Industrial Customers 0.74 0.85
Including Large Industrial Customers 0.68 0.79

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 3.17 will converge
with the blue winter line in approximately year 2100. Figure 3.18 shows that the winter
high/low bound considerably larger than summer, and reflects a greater range of
temperature anomalies in the winter months. Table 3.7 shows the energy and peak
forecasts.

Figure 3.17: Peak Load Forecast 2015-2035

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
Me
g
a
w
a
t
t
s
Winter Peak
Summer Peak
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-25
Figure 3.18: Peak Load Forecast with 1 in 20 High/Low Bounds, 2015-2035

Table 3.7: Energy and Peak Forecasts

Year
Energy
(aMW)
Winter Peak
(MW)
Summer Peak
(MW)
2016 1,074 1,718 1,582
2017 1,084 1,731 1,596
2018 1,091 1,744 1,610
2019 1,097 1,756 1,623
2020 1,099 1,768 1,635
2021 1,102 1,780 1,648
2022 1,105 1,792 1,661
2023 1,110 1,804 1,674
2024 1,115 1,816 1,686
2025 1,120 1,828 1,699
2026 1,125 1,840 1,713
2027 1,131 1,853 1,726
2028 1,137 1,865 1,739
2029 1,143 1,878 1,753
2030 1,150 1,891 1,766
2031 1,156 1,903 1,780
2032 1,163 1,916 1,794
2033 1,169 1,929 1,808
2034 1,176 1,942 1,822
2035 1,183 1,955 1,836

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
Me
g
a
w
a
t
t
s
Winter Peak Summer Peak
Winter- High Winter- Low
Summer- High Summer- Low
Chapter 3: Economic & Load Forecast

Avista Corp 2015 Electric IRP 3-26
Testing for Changes in Extreme Temperature Behavior
The impacts of global warming and the relevance of historical temperature data when
forecasting future peak loads, drives much of the recent load forecasting debates. To
validate the use of historical temperatures in the peak load forecast, an analysis was
conducted using the same GISS methodology and reference period referenced in the
UPC forecast methodology section. In particular, using 1951-1981 as the reference
period, Avista examined daily temperature anomalies using daily temperature data from
the Spokane International Airport going back to 1947. The analysis focused on the core
summer months (June, July, and August) and winter months (December, January, and
February). The GISS study only considered 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
1951 to 1981 reference period. In contrast, while Avista analysis shows increased
average temperatures compared to the reference period, 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.

Chapter 4: Existing Supply Resources
Avista Corp 2015 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 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 42 percent, than hydroelectric at 37 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: 2016 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
Hydro
10%
Natural Gas
42%
Coal13%
Biomass & Wind
7%
Annual Energy
Section Highlights
Hydroelectric represents about half of Avista’s winter generating capability.

Seven percent of Avista’s generating


Chapter 4: Existing Supply Resources
Avista Corp 2015 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 stripped of 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 a separate license authorized by 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. 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
transmission grid.

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. Avista
redeveloped the Huntington Park area around this facility in 2014 in honor of the
company’s 125th anniversary.

Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-3
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 is undergoing substantial upgrades scheduled for completion in 2016. Two 8-
MW units will replace its existing 3-MW units. Once operational, the new units will add
1.4 aMW of energy beyond the plant’s original configuration and bring total operating
capability to 32 MW. The nameplate rating of the facility will rise to 36 MW. In addition to
capacity upgrades, the facility will receive new hydraulic governors, static excitation
systems, switchgear, station service, control and protection packages, ventilation,
rehabilitation of intake gates and sediment bypass system, and other investments.

Long Lake
The Long Lake development is located northwest of Spokane and maintains the Lake
Spokane reservoir, also known as Long Lake. The plant received new runners in the
1990s, bringing the project’s four units to 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. A new runner upgrade in 2001
added 0.6 aMW of energy generation to the project. 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.
Current projects include replacing station service equipment, updating the powerhouse
crane, and developing new control schemes and panels. After the preliminary work is
Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-4
completed, replacing generators, turbines, and unit protection and control systems on
the four units will start.

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.

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

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.

Total Hydroelectric Generation
Avista’s hydroelectric plants have 1,065.4 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 the 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 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
Cabinet Gorge Clark Fork Clark Fork, ID 265.2 270.5 123.6
Noxon Rapids Clark Fork Noxon, MT 518.0 610.0 196.5

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.
Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-5
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 2 Boardman, OR Gas 2003 312.0 277.0 287.3
Kettle Falls Kettle Falls, WA Wood 1983 47.0 47.0 50.7
Kettle Falls CT1 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. The Avista 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 near Rathdrum, Idaho connects to the Avista transmission system. It entered
service in 1995 and has a maximum capacity of 178.0 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
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 is limited to run no more than
approximately 550 hours 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.
1 The Kettle Falls CT numbers include output of the natural gas-fired turbine plus the benefit of its steam
to the main unit’s boiler.
Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-6
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 with a maximum
capacity of 285 MW in the winter and 250 MW in the summer, with duct burners
providing additional capacity of up to 27 MW. The plant nameplate rating of the plant is
287.3 MW.

Recent upgrades to Coyote Springs 2 include cooling optimization and cold day
controls. The cold day controls remove firing temperature suppression that occurs when
ambient temperatures are below 60 degrees. The upgrade improves the heat rate by
0.5 percent and output by approximately 2.0 MW during cold temperature operations.
The cooling optimization package improves compressor and natural gas turbine
efficiency, resulting in an overall increase in plant output of 2.0 MW. In addition to these
upgrades, Coyote Springs 2 now has a Mark VIe control upgrade, a new digital front
end on the EX2100 gas turbine exciter, and model-based control with enhanced
transient capability. Each of these upgrades allows Avista to maintain high reliability,
reduce future O&M costs, maintain compliance with WECC reliability standards, and
help prevent damage to the machine during electrical system disturbances.

Kettle Falls Generation Station and Kettle Falls Combustion Turbine
The Kettle Falls Generating Station, a biomass facility, entered service in 1983 near
Kettle Falls, Washington. It is among the largest biomass 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 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. 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 2015 IRP. Tables 4.3 through 4.5 summarize
Avista’s contracts.
Chapter 4: Existing Supply Resources
Avista Corp 2015 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 when
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 2015, the contracts provide approximately 160 MW of capacity
and 96 aMW of energy. The Douglas PUD (2018) and Chelan PUD (2020) contracts
expire over the next five years. Avista may extend these contracts or even gain
additional capacity in auctions; however, there are no guarantees to extend contract
rights. Due to this uncertainty around future availability and cost, the IRP does not
include these contracts in the resource mix beyond their expiration dates.

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 in earnest to decide whether the current
treaty will continue, should be ended, or if a new agreement will be reached. 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, as
part of 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.

Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-8
Table 4.3: Mid-Columbia Capacity and Energy Contracts

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 16.9
Grant PUD Wanapum 3.7 Dec-2001 Dec-2052 34.5 27.2
Chelan PUD Rocky Reach 5.0 Jan-2016 Dec-2020 58.1 18.4
Chelan PUD Rock Island 5.0 Jan- 2016 Dec-2020 20.1 25.7
Douglas PUD Wells 3.3 Feb-1965 Aug-2018 27.9 16.5
Canadian Entitlement -10.1 -5.7
2016 Total Net Contracted Capacity and Energy 155.3 99.0

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.

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 further discussion
of this obligation, and how Avista plans to account for it, is contained 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.
Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-9
Table 4.4: PURPA Agreements

Meyers Falls Hydro
Technology
Systems Inc.
Hydro Kettle Falls,
WA
12/2013 1.30 1.05
Spokane
Waste to
Energy
City of Spokane Municipal
Waste
Spokane, WA 12/2017 18.00 16.00
Spokane
County
Digester
Spokane County Municipal
Waste
Spokane, WA 8/2016 0.26 0.14
Plummer Saw
Mill
Stimson Lumber Wood
Waste
Plummer, ID 11/2016 5.80 4.00
Deep Creek Deep Creek
Energy
Hydro Northpoint, WA 12/2016 0.41 0.23
Clark Fork
Hydro
Clark Fork LLC. Hydro Clark Fork, ID 12/2017 0.22 0.12
Upriver Dam2 City of Spokane Hydro Spokane, WA 12/2019 17.60 6.17
Sheep Creek
Hydro
Sheep Creek
Hydro Inc.
Hydro Northpoint, WA 6/2021 1.40 0.79
Ford Hydro LP Ford Hydro Ltd
Partnership
Hydro Weippe, ID 6/2022 1.41 0.39
John Day
Hydro
David Cereghino Hydro Lucille, ID 9/2022 0.90 0.25
Phillips Ranch Glenn Phillips Hydro Northpoint, WA n/a 0.02 0.01

Table 4.5: Other Contractual Rights and Obligations

PGE Capacity Exch. Exchange System 12/2016 -150 -150 0
Douglas Settlement Purchase Hydro 9/2018 2 2 3
Energy America Sale CEC RECs3 12/2019 50 50 50
WNP-3 Purchase System 6/2019 82 0 42
Lancaster Purchase Natural Gas 10/2026 279 228 215
Palouse Wind Purchase Wind 12/2042 0 0 40
Nichols Pumping Sale System n/a -1 -1 -1
2 Energy estimate is net of the city’s pumping load. 3 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.
Chapter 4: Existing Supply Resources
Avista Corp 2015 Electric IRP 4-10
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 increased to 38 per year. The increase likely
was in response to 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 2020. The Washington utility taxes credit
finances these incentives that rise to as much as $1.08 per kWh.

Avista had 208 customer-installed net-metered generation projects on its system at the
end of 2014 representing a total installed capacity of 1.8 MW. Eighty-four percent of
2014 installations are in Washington, with most located in Spokane County. In that year,
Avista credited customers $245,884 for the energy created via the Washington state tax
incentive–an average of $281 per MWh. Figure 4.2 shows annual net metering
customer additions. 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 current tax incentives
are nearing optimal payback, the number of new net-metered systems rose in 2014. If
tax subsidies end without a significant reduction in technology cost, the interest in net
metering likely will return to pre-tax incentive levels. 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

0.0
0.3
0.7
1.0
1.3
1.7
2.0
0
10
20
30
40
50
60
19
9
9
20
0
1
20
0
2
20
0
3
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
Cu
m
u
l
a
t
i
v
e

C
a
p
a
c
i
t
y

(
M
W
)
An
n
u
a
l

N
e
w

C
u
s
t
o
m
e
r
s
ID
WA
Cumulative Capacity MW
Chapter 4: Existing Supply Resources
Avista Corp 2015 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. 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 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 will utilize approximately
$600,000 each year in state tax incentives.

Boulder Park Community Solar Project

Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
5. Energy Efficiency & Demand Response

Introduction
Avista began offering energy efficiency programs to its customers in 1978. Recent
programs include the distribution in the summer of 2011 of 2.3 million compact
fluorescent lights (CFLs) to residential and commercial customers for an estimated
energy savings of 39,005 MWh. The 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 197 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 assumed measure life accounts for the difference between the
cumulative and online trajectories in Figure 5.1. Currently 127 aMW of conservation
serves customers, representing nearly 11 percent of loads.

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 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.

Section Highlights
 Current Avista-sponsored conservation reduces retail loads by nearly 11
percent, or 127 aMW.
 This IRP evaluated over 3,000 equipment options and over 2,300 measure
options covering all major end use equipment, as well as devices and actions
to reduce energy consumption for this IRP.
 This 2015 IRP is the first to co-optimize conservation and demand response
options with generation resource options using our PRiSM model.

 Bas
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Figure 5.1: Historical and Forecast 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) to develop an independent 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 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 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.

aMW
40 aMW
80 aMW
120 aMW
160 aMW
200 aMW
aMW
4 aMW
8 aMW
12 aMW
16 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
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
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Figure 5.2: Analysis Approach Overview

1. Market Assessment: 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.
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, economic, and achievable
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.

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.
Avista data

Avista data Avista data/secondary data

Energy market profiles by end
use, fue/secondary data Develop prototypes and
perform energy analysis
Forecast assumptions:
Customer growth
Price forecast
Purchase shares
Codes and standards
Energy efficiency measure list
measure costs and savings
analysis
Base-year energy consumption
by state, fuel, and sector

Avista data

Avi
Energy market profiles by end
use, fuel, segment, and vintage
Baseline forecast by end use
Energy efficiency potential
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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 2013 when the study began. Codes and standards have direct bearing
on the amount of energy efficiency potential that exists 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 Sixth Power Plan, the Regional Technical
Forum, and other measures applicable to Avista. The approximately 6,000 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 CPA includes
measure costs, energy and capacity savings, estimated useful life, and other
performance factors identified for the list of measures and economic screening
performed 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 Sixth
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, three types of potential are in this study, as
discussed below. Table 5.1 shows the CPA results for technical, economic, and
achievable potential.

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.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Table 5.1: Cumulative Potential Savings (Across All Sectors for Selected Years)

2016 2017 2020 2025 2035
Cumulative (GWh)
Achievable Potential 34 74 236 575 1,090
Economic Potential 68 138 360 733 1,292
Technical Potential 173 344 837 1,581 2,506

Cumulative (aMW)
Achievable Potential 3.9 8.5 26.9 65.6 124.5
Economic Potential 7.7 15.9 41.1 83.7 147.5
Technical Potential 19.8 39.3 95.5 180.5 286.1

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 that
would result 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 the only considered portion of
current equipment stock is that reaching the end of its useful life 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.

Economic Potential
Economic potential includes the purchase of the most efficient cost-effective option
available for each given equipment or non-equipment measure.3 Cost effectiveness
is determined by applying the Total Resource Cost (TRC) test using all quantifiable
costs and benefits, regardless of who accrues them, and inclusive of non-energy
benefits as identified by the NPCC.4 Measures passing the economic screen
represent aggregate economic potential. As with technical potential, economic
potential calculations use a phased-in approach. Economic potential is a hypothetical
upper-boundary of savings potential representing only economic measures; it does
not consider customer acceptance and other factors.

Achievable Potential
Achievable potential refines economic potential, accounting for expected program
participation, customer preferences, and budget constraints. It estimates achievable
savings attainable through Avista energy efficiency programs when considering market
3 The Industry definition of economic potential and the definition of economic potential referred to in this
document are consistent with the definition of “realizable potential for all realistically achievable units”. 4 There are other tests to represent economic potential from the perspective of stakeholders (e.g.,
Participant or Utility Cost), but the TRC is generally accepted as the most appropriate representation of
economic potential because it tends to represent the net benefits of energy efficiency to society. The
economic screen uses the TRC as a proxy for moving forward and representing achievable energy
efficiency savings potential for measures that are most cost-effective.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
maturity and barriers, customer willingness to adopt new technologies, incentive levels,
as well as whether the program is mature or represents the addition of a new program.

During this stage, AEG applied market acceptance rates based upon NPCC-defined
ramp rates from the Sixth Power Plan, taking into account market barriers and measure
lives. However, AEG adjusted the ramp rates for the measures and equipment to reflect
Avista’s market-specific conditions and program history. In some cases Avista ramp
rates exceed the NPCC’s, illustrating a mature energy efficiency program reaching a
greater percentage of the market than estimated by the now five-year-old Sixth Power
Plan. In other cases, where a program does not currently exist, a ramp rate could be
less than the NPCC’s ramp rate, acknowledging the additional design and
implementation time necessary to launch a new program. Other examples of ramp rate
changes include measures or equipment where the regional market shows lower
adoption rates than historically estimated by the NPCC, such as heat pump water
heaters. AEG’s CPA forecasts incremental annual achievable potential for all sectors at
3.9 aMW (34,106 MWh) in 2016, increasing to cumulative savings of 124.5 aMW
through 2035.

PRiSM Co-Optimization
For the first time, this IRP used a second methodology to identify achievable
conservation potential. This method selects conservation measures concurrently with
supply side resources in Avista’s PRiSM model. This methodology was the result of a
2013 IRP Action Item to streamline the process of selecting conservation in conjunction
with the efficient frontier modeling process. See Chapter 11 for more details about the
PRiSM model. The method inputs all measures with TRCs less than 130 percent of the
avoided cost rate, adjusted for ramp rates used for achievable potential. The 130-
percent threshold ensures that conservation options are available in the lower-risk
region of the efficient frontier, just as PRiSM includes higher-cost supply-side options
that help mitigate risk. The conservation resources compete with supply- and demand
response options to meet Avista resource deficits. Each conservation program’s winter
and summer peak contribution, plus the value of its energy savings are considered.

Given the change to evaluating conservation directly in PRiSM, results were also
compared to the historical method. Figure 5.3 shows both CPA and PRiSM
conservation estimates. The results were very similar, with PRiSM selecting 0.4 aMW
more conservation than the CPA over the 20-year horizon. The similar result is
evidence that the avoided cost method used for previous IRPs was accurate. However,
using PRiSM for program selection allows conservation selections to change with
differing resource strategies across the efficient frontier.5 Previously a change in
resource selection required a feedback loop with AEG to re-run the CPA with new
avoided costs. With the new approach, no feedback loop is required. Given the results
of this methodology, Avista will likely use this method in future IRPs for conservation
selection.

5 For example, pursuing a least-cost strategy might have less conservation resource than pursuing a
least-cost strategy where more costly supply-side resources are being avoided through conservation.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Figure 5.3: Cumulative Conservation Potentials CPA versus PRiSM

Conservation Targets
The IRP process provides conservation targets for the 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. Table 5.2 contains achievable conservation
potential for 2016-2017 using both the AEG and PRiSM methodologies. Also included is
the energy savings expected from the 2016 and 2017 feeder upgrade projects. See
Chapter 8 – Transmission and Distribution Planning for more information.

Table 5.2: Annual Achievable Potential Energy Efficiency (Megawatt Hours)

Year Methodology Washington Idaho
2016 AEG CPA 22,863 11,243
2016 PRiSM Selection 22,747 11,213

2017 AEG CPA 26,930 13,217
2017 PRiSM Selection 26,799 13,186

2016 WA Feeder Upgrades 485 1,118
2017 WA Feeder Upgrades 0 0

2016 Facility Efficiencies 0 300
2017 Facility Efficiencies 151 0
4
27
66
98
124
4
27
65
98
125
0
20
40
60
80
100
120
140
2016 2020 2025 2030 2035
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
AEG's CPA
PRiSM Selection
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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.6 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 Sixth 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 2016 conservation acquisition cost to electric customers is $11.6 million. This
amount grows by 224% to $26 million by 2026, a total of $186 million over this 10-year
period. Costs continue increasing after 2026 to more than $31 million in 2035. Figure
5.4 shows the annual cost in millions of nominal dollars for the utility to acquire the
projected electric achievable potential.

Figure 5.4: Existing & Future Energy Efficiency Costs and Energy Savings

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
6 The EIA defines cost effective as 10 percent higher than the cost a utility would otherwise spend on
energy acquisition.
0
10
20
30
40
50
60
70
80
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
Energy Savings (aMW)
Spending (millions $)
Levelized Cost ($/MWh)
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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. 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
Avista offers most residential energy efficiency programs through prescriptive or
standard offer programs targeting a range of end uses. Programs offered through this
prescriptive approach during 2014 included space and water heating conversions,
ENERGY STAR® homes, space and water equipment upgrades, and home
weatherization. The appliance programs offered by ENERGY STAR® phased out in
2013 due to results of a Cadmus net-to-gross study indicating market transformation to
a point that incentives are no longer required. Other non-appliance ENERGY STAR®
programs continue.

Avista offers its remaining residential energy efficiency programs through other
channels. For example, JACO, a third party administer, operates a refrigerator/freezer
recycling program. UCONS administers a manufactured home duct-sealing program.
CFL buy-downs at the manufacturer level provide customers access to lower-priced
lamps. Home energy audits, subsidized by a grant from the American Recovery and
Reinvestment Act (ARRA), ended in 2012. This program offered home inspections
including numerous diagnostic tests and provided a leave-behind kit containing CFLs
and weatherization materials. ARRA funds also helped support another program aimed
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
at helping to remove the financial roadblocks to implementing energy efficiency for
customers. This program used ARRA funds to buy down the interest rate on loans
geared directly towards installing energy efficiency measures in the home. This loan
program ended December 31, 2014, after helping fund 269 projects.

Avista processed 5,300 residential energy efficiency rebates in 2014, benefiting
approximately 4,000 households. Rebates of over $2.3 million offset customer
conservation-implementation costs. Third-party contractors implemented a second
appliance-recycling program and a manufactured home duct-sealing program. Avista
participated in a regional upstream buy-down program called Simple Steps Smart
Savings to provide customers reduced cost lighting and showerheads through
participating retailers. Finally, Avista distributed over 7,700 CFLs, and provided expert
advice, at various community events throughout the service territory. Residential
programs contributed 25,397 MWh and 355,443 therms of energy savings in 2014.

Avista successfully launched a three-year cost-effective behavioral program in June
2013 using the Opower Home Energy Report platform, where participating customers
receive a peer-comparison report in the mail every two months. Since launch of the
program, Avista has seen a higher than expected ramp rate of energy savings for
participating customers as measured in the statistically valid Randomized Control Trial
method. Uptake in other energy efficiency programs increased as well. The Opower
Home Energy Report contributed 8,131 MWh of savings in 2014.

Low-Income Sector Overview
During 2014, six community action agencies administered Avista low-income programs,
targeting a range of end-uses including space and water heating conversions, ENERGY
STAR® refrigerators, and weatherization improvements. Beyond direct energy
efficiency measures, Avista funding goes towards health and safety improvements
considered necessary to ensure the habitability of low-income homes and protect the
efficiency measures. The funding also allows the agencies to receive an administration
fee for program delivery.

Avista processed approximately 1,400 low-income sector rebates in 2014, benefitting
360 households.7 During 2014, Avista reimbursed the six agencies over $2.6 million for
energy efficiency upgrades where some measures were fully subsidized and others
capped based on avoided costs. The agencies spent nearly $394,000 on health and
human safety, or 13 percent of their total expenditures–within their 15 percent
allowance for this spending category. The low-income energy efficiency programs
contributed 400 MWh of electricity savings and 14,944 therms of natural gas savings in
2014.

Non-Residential Sector Overview
Marketing and the new energy efficiency program development starts with measures
highlighted in the CPA. All electric-efficiency measures with simple paybacks exceeding
one year, but less than eight years for lighting measures or 13 years for other
7 Washington agencies had up to $2.0 million available for energy efficiency improvements. Idaho had
$700,000 available for energy efficiency improvements and $50,000 for conservation education.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
measures, automatically qualify for the non-residential portfolio. The IRP provides
account executives, program managers/coordinators, and energy efficiency engineers
to support program implementation. However, characteristics of a non-residential facility
override any high-level program prioritization.

For the non-residential sectors, including multi-family, Avista offers energy efficiency
programs on a site-specific or custom basis. Avista offers prescriptive approaches when
treatments result in similar savings and the technical potential is high. As an example,
the prescriptive lighting program is not purely prescriptive in the traditional sense, such
as with residential applications where homogenous programs are provided for all
residential customers. It is a more prescriptive approach applied for these similar
applications.

Non-residential prescriptive programs offered by Avista include, but are not limited to,
space and water heating conversions and equipment upgrades, appliance and cooking
equipment upgrades, personal computer network controls, commercial clothes washers,
lighting, motors, refrigerated warehouses, traffic signals, and vending controls. Also
included are residential program offerings, including site-specific multi-family measures
and multi-family market transformation.

Avista processed 1,100 energy efficiency projects resulting in the payment of over $4.6
million in rebates paid directly to non-residential customers to offset the cost of their
energy efficiency projects in 2014. These projects contributed 24,400 MWh of electricity
and 262,000 therms of natural gas savings.

PECI’s Energy Smart Grocer is a regional turnkey program administrated for several
years in the Avista service territory. It will approach saturation levels during the early
part of the IRP 20-year planning horizon. The Energy Smart Grocer program contributed
3,275 MWh of the 24,400 MWh of non-residential program savings in 2014.

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. Avista intends to study converting
decorative lighting and larger-wattage (200 watt and 400 watt) streetlights in the future.

Demand Response
Over the past decade, demand response (DR) gained growing attention as an
alternative for meeting peak load growth. Demand response reduces load to specific
customers during peak demand periods. Enrolling customers allows the utility to modify
their usage pattern in exchange for bill discounts. National attention focuses on
residential programs to control water heaters, space heating, and air conditioners. A
2013 Action Item suggested further study of the DR potential based on its selection as a
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
PRS resource from 2022 to 2027 in that plan. Avista retained AEG to study the potential
of future commercial and industrial programs.

Past Programs
Avista’s experience with DR dates back to 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 rely on 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
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 that uncovered a
control board in a customer’s heat pump that caused 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 STP 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.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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, which might limit
customer interest.

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 2016–2035. 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
that may hinder acquisition.

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. Table 5.3 details achievable
demand response potential for the programs studied by AEG.

Table 5.3: Commercial and Industrial Demand Response Achievable Potential (MW)

Program 2016 2020 2025 2030 2035
Direct Load Control 0.6 6.5 6.7 6.9 7.2
Firm Curtailment 5.8 17.5 17.4 17.4 17.5
Opt-in Critical Peak Pricing 0.1 1.4 4.3 4.3 4.4
Opt-out Critical Peak Pricing 6.3 4.4 12.9 13.0 13.1

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,
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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 that 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. 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. An opt-in rate that 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.

Studies show that dynamic pricing programs such as critical peak pricing 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 demand response implemented through energy
management and control systems.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
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 hours. The AEG DR study included standby
generation in its firm curtailment section. 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 their use during peak hours. 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 20 MW of standby generation resources are available
for utility use over a five-year acquisition period. To test the concept, a pilot using Avista
backup generation facilities is likely. The pilot would provide a cost estimate and
illustrate the engineering necessary to bring a standby generation program to fruition.
The IRP assumes a standby generation program would cost $50 to $85 per kW in
upfront investments, plus $10 to $15 per kW-year in O&M costs.

In May 2015, the federal courts overturned rules limiting the availability of standby
generation resources. This ruling creates uncertainty around using standby generation
to serve utility requirements. The ruling requires new rules to be developed to determine
the amount of hours and environmental conditions these units could be used.

Generation Efficiency Audits of Avista Facilities
A 2013 IRP Action Item was the study of potential for energy efficiency opportunities at
company generation facilities. During 2015, Avista performed preliminary energy
efficiency audits at all of its hydroelectric dams and most thermal generation facilities
Avista owns or is a partial owner in, excluding Colstrip. The preliminary scoping audits
focused on lighting, shell, heating ventilation and air conditioning (HVAC), and motor
controls on processes. Table 5.4 summarizes these potential projects, Table 5.5
summarizes the planned projects for 2016 – 2017, and Appendix D contains a complete
description of the study findings. A discussion of some of the major identified categories
follows. Studies will continue into 2016 and the findings reported in the 2017 IRP.

Lighting Projects
Avista’s generation 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
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Rapids is the only facility that has completed a lighting retrofit. Little Falls, Nine Mile,
Cabinet Gorge and Long Lake lighting upgrades are planned in 2016 and 2017.

Shell Projects
Shell projects include measures keeping conditioned air within buildings. A generation
facilities review found no capital shell measures with significant savings potential.
However, small maintenance weatherization efforts could improve occupant comfort.

Table 5.4: 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
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

8 Also saves 23,911 therms of natural gas per year.
Chapter 5–Energy Efficiency & Demand Response
Avista Corp 2015 Electric IRP
Table 5.5: Planned Generation Facility Efficiency Upgrades 2016 – 2017

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

HVAC Projects
Noxon Rapids is the only hydroelectric project with heating and cooling equipment. Its
water-source heat pump system includes air handlers and hydronic unit heaters. In
addition to efficiency gains, replacing this system would reduce annual maintenance.

Cabinet Gorge does not have active heating or cooling systems. Ducted hydronic coils
flush air outside during spring and summer nights. A water-source heat pump would
increase overall heating and cooling efficiency.

In most cases waste heat from the hydroelectric generating equipment supplies heat to
facilities in winter months. When idle, facilities typically motor a unit during the winter
months to keep the facility above freezing. Unit heaters could provide a more efficient
heat source, and the control room could be thermally isolated from the rest of the plant
to ensure only required areas are heated.

Given the relative efficiency of existing thermal facilities heating systems, HVAC
equipment improvements make sense only when each unit reaches the end of its useful
life.

Controls on Process Motors
Most motor loads at the hydroelectric facilities operate limited hours, often less than 30
hours per year. They do not consume enough electricity to justify the cost of installing
new variable speed drives. Coyote Springs 2 has potential for variable-speed motors in
its compressed-air systems. The Little Falls exhaust fan could benefit from the
installation of variable speed drives.

Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-1
6. Long-Term Position

Introduction & Highlights
This chapter describes the analytical framework used to develop Avista’s net 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.

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 regional
load. Annual peaks can occur either in the winter or in the summer; but on a planning
basis using extreme weather conditions, Avista is winter peaking. The winter peak
generally occurs in December or January, but may happen in November or February
where weather events occur in these months. As described in Chapter 4 – Existing
Resources, Avista’s resource mix contains roughly equal splits between 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 contingencies. Reserve margins, on average, increase customer rates when
compared to resource portfolios without reserves because of the additional 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 revenues.

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:

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)
Section Highlights
 Avista’s first long-term capacity deficit net of energy efficiency is in 2021; the
first energy deficit is in 2026.
 Including operating reserves, Avista plans to a 22.6 percent planning margin.
 The 2015 IRP meets all EIA mandates over the next 20 years with a
combination of qualifying hydroelectric upgrades, purchased RECs, Palouse
Wind, and Kettle Falls.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-2
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’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 therefore includes operating reserves in its definition of planning
margin. Per Western Electric Coordinating Council (WECC) requirements, Avista must
maintain 1.5 percent of current load and 1.5 percent of on-line generation as spinning
reserves and 1.5 percent of current load and 1.5 percent of on-line generation as
standby reserves.2 Avista must also hold load regulation reserves to meet load following
and regulation requirements of within-hour load and generation variability.

Avista participates in regional Energy Imbalance Market (EIM) studies and committees.
An EIM, where adopted, would create a trading market for regulation services, among
other products. While the new market may not reduce the amount of required capacity,
it may lower customer rates by providing Avista another market to buy and sell short-
term capacity products and services.

Planning Margin
Utility capacity planning begins with identifying the broader regional capacity position,
as regional surpluses can offset 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 6,000 MW of natural
gas-fired generation, about 3,500 MW was constructed immediately after the crisis.
During this same time, Oregon and Washington added 7,850 MW of wind generation.
With recent wind additions in the mix, due to wind’s lack of on-peak capacity
contribution, the region is approaching load-resource capacity balance, while retaining
an energy surplus.
1 http://www.nerc.com/pa/RAPA/ri/Pages/PlanningReserveMargin.aspx 2 Spinning reserves sync to the system while stand-by reserves must be available within 10 minutes.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-3
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 5 percent LOLP threshold. Five percent LOLP means utilities meet all customer
demand in all hours of the year in 19 of 20 years; one loss-of-load event is allowed in a
20-year period. Regional LOLP analysis is beyond the scope of an IRP. Fortunately, the
NPCC conducts regional LOLP studies. Based on their work, the Northwest begins to
fail the five-percent LOLP measure in the winter of 2020-21 when three major coal
generators retire.3 The NPCC identifies a need of 1,150 MW of natural gas-fired
capacity to eliminate potential 2021 resource shortfalls. The projected shortages occur
primarily in the winter, with a small chance of shortage in the summer. At the time of
writing, the NPCC had not translated its LOLP study results into a regional planning
margin statistic. Absent NPCC translation to planning margin level, Avista made its own
estimate using NPCC data and historical methodology to perform the translation.
Including operating reserves, the Northwest planning margin is between 23 and 24
percent.

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 with 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 target a lower planning margin. The
harder question is how much reliance it should place on the wholesale market. Previous
IRPs have shown charts like Figure 6.1, the tradeoff between added resources, i.e.,
planning margin, and higher system costs and wholesale market reliance. For example,
were Avista an electrical island like AELP, a 5 percent LOLP would require a 31 percent
planning margin, adding nearly $40 million annually to rates. On the opposite end of the
spectrum, if the marketplace had 275 megawatts available, a 12 percent planning
margin would meet the 5 percent LOLP for no added cost. Figure 6.1 also explains that
in 2020, absent any resource additions or market reliance, Avista projects a 12 percent
reserve margin.

3John Fazio, NPCC, http://www.nwcouncil.org/media/7149183/may-1-2015-raac-steer-2020-21-
adequacy-assessment.pdf. The 8.3 percent LOLP result primarily is due to the retirements of the
Boardman and Centralia coal-fired plants, and to a lesser extent regional load growth.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-4
Figure 6.1: 2020 Market Reliance & Capacity Cost Tradeoffs
Avista reviewed planning margins used by transmission organizations and utilities
across the country. The results varied depending on the depth and breadth of their
interconnections and the types and quantities of resources within their systems. One
challenge in comparing planning margins across utilities is determining whether they
include ancillary service, or operating reserve, obligations in their planning margins.
Figure 6.2 illustrates the findings of our review of utility 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 metric to consider 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 met 16 percent of weather-adjusted peak load in 2014, a
high statistic relative to our Western Interconnect peers. Figure 6.3 illustrates the single
largest contingencies for selected utilities in the West. Excluding Avista, the average
percentage of peak load is 11 percent; the high is 33 percent for Sierra Pacific (553 MW
Tracy CCCT), and the low is 5 percent for BC Hydro.

Some resource planners argue planning margins should be no smaller than a utility’s
single largest contingency on the basis that where your 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
0
5
10
15
20
25
30
35
40
45
-
50
100
150
200
250
300
12%13%15%16%18%19%21%22%24%25%27%28%30%31%
In
c
r
e
m
e
n
t
a
l

C
o
s
t

(
$
M
i
l
l
/
Y
r
)
Ma
r
k
e
t

C
o
n
t
r
i
b
u
t
i
o
n

(
M
W
)
Planning Margin
MW
Annual Cost
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-5
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
load reductions.

Figure 6.2: Planning Margin Survey Results

0%
8%
10%
10%
11%
12%
12%
12%
12%
12%
12%
12%
13%
13%
13%
14%
14%
14%
14%
14%
15%
15%
15%
15%
15%
15%
15%
15%
15%
15%
15%
15%
15%
15%
16%
16%
16%
16%
17%
17%
18%
18%
20%
20%
20%
20%
20%
22%
23%
25%
Northwestern
Hydro QuebecFortis
Idaho Power
Minnesota PowerEntergy-New Orleans
Sunflower CoopKansas City B of PU
Oklahoma Gas & Electric
Salt River ProjectNevada Power
PGEIndianoplis Light & Power
Public Service Co of NM
PacifiCorpSPP
XCEL-New Mexico
ERCOTDuke-Indiana
PSE (2018-19)Duke-Carolina's
MISO
TVAISO New England
California PUCBasin Electric
San Diego Gas & Electric
Roseville ElectricPlatte River Power Authority
APSUNS Electric
El Paso Electric
Sierra PacificTri-State G&T
Dominion
PJMPSE (2020+)
XCEL-ColoradoEWEB
NYISO
Colorado Springs UtilitiesClark PUD
Nova Scotia PowerHydro One
FPL
Progress EnergyBC Hydro
New Brunswick PowerCowlitz PUD
LADWP
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-6
Figure 6.3: Single Largest Contingency Survey Results (2014 Peak Load)
Flexibility Requirements
Renewable portfolio standards, large federal and state subsidies, high feed-in tariff and
PUPRA prices, and falling equipment and installation costs have led to more intermittent
wind and solar generation installations in the Northwest. Unlike traditional generation
resources, intermittent generation variability consumes system capacity. This is similar
to holding generation capacity for unknown changes in load, but differs because
changes in renewable generation output are much larger and more volatile than load
changes on a per-MW of capacity basis. Avista and many of its peer utilities have
conducted studies to ensure they have enough flexible capacity to support intermittent
resources. However, analytical methods contained in these studies are not fully mature
because it is a relatively new concept for the industry.

Avista has identified an initial analytical process to study flexibility requirements for this
IRP. The first step looks at system variation on different time horizons. The analysis
looks at the five-, 10-, 15- and 60- minute periods in calendar year 2013. The study
estimated the amounts of capacity reserves required in the 95th and 99th percentile, or
8,322 and 8,672 hours of the 8,760 hours of a year. While Avista will need to meet all
needs during the calendar year, some reliance on the wholesale marketplace is
appropriate. Figures 6.4 and 6.5 outline the amount of capacity required to meet load
and wind variation, and operating reserve requirements, at the 95th and 99th percentiles.
Over the five-minute time range, Avista needs 100 MW to 107 MW of flexible resources.
Extending the time horizon to 10 minutes, 110 MW to 122 MW are required. Between
120 MW and 137 MW are required for 15-minute interval variation. Over an hour, total
33%
20%
16%
16%
15%
13%
12%
10%
10%
9%
9%
8%
7%
7%
6%
6%
5%
Sierra Pacific (Tracy CCCT-553)
Avista-Winter (Coyote Springs 2-312)
Portland General Electric (Boardman-517)
Avista-Summer (Coyote Springs 2-277)
PacifiCorp-West (Chehalis-477)
Public Service of NM (San Juan-248)
El Paso Electric (Palo Verde-207)
Nevada Power (Lenzie-551)
Idaho Power (Langley Gultch-318)
Public Service of CO (Comanche-525)
PacifiCorp-East (Lake Side 2-628)
LADWP (Scattergood-450)
Arizona Public Service (Redhawk-500)
Bonneville Power Admin (Coulee-805)
Salt River Project (Springerville-415)
Puget Sound Energy (Mint Farm-297)
BC Hydro (Various-500)
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-7
needs are 196 MW to 260 MW. Regulation-capable resources are required to meet
much of the variation under 15 minutes, though the 44 MW of non-spinning reserve can
be met with stand-by ready resources. For the hour, incremental capacity requirements
over the five- to 15-minute intervals increases, but standby resources meet the
requirement.

Figures 6.4 and 6.5 identify the requirements for flexible resources on the system, but
they do not identify the resources available to meet them. Avista outlines in Chapter 4
resources currently meeting its flexibility requirement. We typically use a combination of
Mid-Columbia contracts and Clark Fork generators to provide regulation and load
following services, but natural gas-fired peaking resources sometimes meet non-spin or
supplemental operating requirements. Recently added controls at Coyote Springs 2
allow it to provide regulation services, taking advantage of its flexibility when online.
Figure 9.5 in Chapter 9 shows the excess reserves by month available to meet flexibility
requirements.

Figure 6.4: 95th Percentile Capacity Requirements

12 22 32
108
-10 -19 -27
-98
44 44 44
44
44 44 44
44
100 110 120
196
-200
-100
0
100
200
300
5 Min 10 Min 15 Min 1 Hour
Me
g
a
w
a
t
t
s
Non-Spin Reserves Spin Reserves
Down Up
Total Up Reg
Regulation Load Following
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-8
Figure 6.5: 99th Percentile Capacity Requirements

Avista’s Planning Margin and Flexibility Reserve Levels
The NPCC Draft Seventh Power Plan finds the region is surplus capacity through 2020.
Avista will not acquire additional capacity until its expected peak loads, plus reserve
margins, exceed resources beyond 2020 either on a single-hour or on a sustained 3-
day basis. To meet customer loads in a reliable and cost-effective manner, Avista
retains resources capable of a minimum of 114 percent of its one-in-two winter peak
load forecast.4 Further, it plans to meet spin- and non-spin requirements, as set by the
WECC. Lastly, Avista retains an additional 16 MW of regulation to serve load and wind
generation variation within the peak hour. The winter total requirement equates to a
22.6 percent planning margin. This level is in line with NPCC estimates for an adequate
supply, as described earlier in this chapter.

The NPCC study shows the region has a minimal chance of a load loss event in
summer months. Given this low probability, Avista’s summer planning margin is
comprised only of balancing area reserve requirements and 16 MW of regulation. Avista
will monitor the summer market depth and will revise its planning margin assumption if
regional capacity surpluses fall due to load growth or exports.

Energy Imbalance Market
Avista is participating 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
4 One-in-two load is the peak load day during an average coldest winter day.
19 34 49
172
-16 -27 -38
-132
44 44 44
44
44 44 44
44
107 122 137
260
-200
-100
0
100
200
300
5 Min 10 Min 15 Min 1 Hour
Me
g
a
w
a
t
t
s
Non-Spin Reserves Spin Reserves
Down Up
Total Up Reg
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-9
(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.

The MC Initiative’s core goal is to lower overall load serving costs by voluntarily re-
dispatching resources. Balancing Authorities (BA) can collectively reduce within-hour
balancing resources and maintain their systems if the EIM captures regional load and
resource diversity and BAs agree on protocols for allocating reserves and ramping
capability obligations among participants. The EIM does this by executing a security-
constrained economic dispatch process every five minutes instead of the current one-
hour term. The process accounts for the capabilities and prices of the volunteered and
committed generating resources for re-dispatch, and the real-time capability of the
transmission system to accommodate flows resulting from a central market-instructed
re-dispatch.

The name “energy imbalance market” implies the core function is managing intra-hour
imbalances – such as load forecast error, generator station error – particularly from
variable energy resources – or both. While covering these imbalances is an integral part
of the EIM, it is not the main objective of the overall economic optimization process. The
market allows BAs to use lower-cost third-party generation when sufficient real-time
transmission exists available to replace their higher-cost generation resources.

The MC Initiative formed an Analytical Team to evaluate the potential production cost
savings within the Northwest Power Pool area. An Executive Committee instructed the
Analytical Team to identify a minimum high-confidence range of potential savings, using
a production cost model with updated grid assumptions provided by members. The
base case results range from approximately $40 million to $90 million per year in
regional gross annual savings. Additional sensitivities resulted in savings of $70 to $80
million dollars to the region. This analysis indicates Avista would conservatively observe
approximately 5 percent of the total regional benefits, or $2 to $5 million. The Executive
Committee currently is evaluating implementation costs to determine if they are lower
than expected savings.

Savings estimates do not reflect significant additional benefits of reducing reserve
requirements in the region. These benefits may add $100 million or more to expected
annual benefit.

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 months than is 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 6.6 illustrates the winter balance of loads
and resources; the first year Avista identifies a significant winter capacity deficit is
January 2021. The load resource comparison removes conservation from the load
forecast to show the total resource need. Conservation will lower this need, but the plan
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-10
requires new generating resources to meet remaining shortfalls. At the time of the IRP
analysis, Avista had small short-term deficits in 2015 and 2016, but those positions
have been filled with market purchases. Chapter 11 – Preferred Resource Strategy
provides more details about the short-term position.

Figure 6.6: Winter 1 Hour Capacity Load and Resources

Avista plans to meet its summer peak load with a smaller planning margin than in the
winter. During summer months, only operating reserve and regulation obligations are
included in the planning margin. Market purchases in the deep regional market will
satisfy any weather-induced load variation or generation forced outage that otherwise
would be included in the planning margin. Resource additions serving winter peaks
meet smaller summer deficits as well.

Figure 6.7 shows Avista’s summer resource balance. This chart differs from the winter
load and resource balance by using an 18-hour sustained peak rather than the single-
hour peak. Longer-term sustained peaks are more constraining in summer months due
to reservoir restrictions and lower river flows, reducing the amount of continuous
hydroelectric generation available to meet loads.

0
500
1,000
1,500
2,000
2,500
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
Me
g
a
w
a
t
t
s
Existing Resources
Load w/o Conservation + Cont.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-11
Figure 6.7: Summer 18-Hour Capacity Load and Resources

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 planning on the amount of energy available during the “critical water” period of
1936/37.5 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 2015
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 instead
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.
5 The critical water year represents the lowest historical generation level in the streamflow record.
0
500
1,000
1,500
2,000
2,500
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
Me
g
a
w
a
t
t
s
Existing Resources
Load w/o Conservation + Cont.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-12
Beyond load and hydroelectric variability, Avista’s 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 194 aMW in 2016. See Figure 6.8 for the
summary of the annual average energy load and resource net position.

Figure 6.8: Annual Average Energy Load and Resources

Washington State Renewable Portfolio Standard
In the November 2006 general election, Washington voters approved the EIA. The EIA
requires utilities with more than 25,000 customers to source 3 percent of their energy
from qualified renewables by 2012, 9 percent by 2016, and 15 percent by 2020. Utilities
also must acquire all cost effective conservation and energy efficiency measures. In
2011, Avista acquired the output from the Palouse Wind project through a 30-year
power purchase agreement to help meet the EIA goal. In 2012, an amendment to the
EIA allowed some biomass facilities built prior to 1999 to qualify under the law
beginning in 2016. This amendment allows Avista’s 50-MW Kettle Falls project to qualify
and help meet EIA goals.

Table 6.1 shows the forecast amount of RECs Avista needs to meet Washington state
law 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
0
500
1,000
1,500
2,000
2,500
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
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Existing Resources
Load w/o Conservation + Cont.
Chapter 6: Long-Term Position
Avista Corp 2015 Electric IRP 6-13
renewables in 2030. With this ability, Avista does not need additional EIA resources
over the planning horizon of this IRP. It may have surplus renewables depending upon
the qualifying output of Kettle Falls. Kettle Falls qualifying output may vary depending
upon the quantity of fuel meeting the EIA old growth provision, the availability of fuel,
and 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.

Table 6.1: Washington State EIA Compliance Position Prior to REC Banking

2016 2020 2025 2030 2035
Percent of Washington Sales 9% 15% 15% 15% 15%
2-Year Rolling Average Washington Retail Sales
Estimate 645 662 671 682 696

Renewable Goal -58 -99 -101 -102 -104
Incremental Hydroelectric 23 23 23 23 23
Net Renewable Goal -35 -77 -78 -79 -82

Other Available REC's
Palouse Wind with Apprentice Credits 48 48 48 48 48
Kettle Falls (67% Capacity Factor) 31 31 31 31 31
Net Renewable Position (before rollover RECs) 44 3 1 0 -2

Net Renewable Position with Kettle Falls at
90% Capacity Factor 55 14 12 11 8

Chapter 7–Policy Considerations
Avista Corp 2015 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 continue to evolve
since publication of the last IRP; most recently, EPA released the Clean Power Plan in
August 2015. Current and proposed regulations by the EPA, among other agencies,
coupled with political and legal efforts, have particular implications for coal generation,
as they involve regional haze, coal ash disposal, mercury emissions, water quality, and
greenhouse gas emissions. This chapter discusses pertinent public policy issues
relevant to the IRP.

Environmental Issues
The evolving nature of environmental regulation creates unique resource planning
challenges. If avoiding certain air emissions were the only issue facing electric utilities,
resource planning would only require a determination of the amounts and types of
renewable generating technology and energy efficiency to acquire. However, the need
to maintain system reliability, acquire resources at least cost, mitigate price volatility,
meet renewable generation requirements, manage financial risks, and meet changing
environmental requirements sometimes creates conflict. Each generating resource has
distinctive operating characteristics, cost structures, and environmental regulatory
challenges that can change significantly based on timing and location.

Traditional thermal generation technologies, like coal and natural gas-fired plants,
provide reliable capacity and energy. Mine-mouth coal-fired units, like Avista’s shares in
Colstrip Units 3 and 4, have high capital costs and long permitting and construction lead
times, and relatively low and stable fuel costs. New coal plants are difficult, if not
impossible, to site today due to state and federal laws and regulations, local opposition,
their relatively high costs when compared to natural gas-fired plants, and additional
environmental concerns. Remote locations increase costs from either the transportation
of coal to the plant or the transportation of the generated electricity by the plant to load
centers.

Compared to coal, natural gas-fired plants have low capital costs, can typically be
located closer to load centers, can be constructed in relatively short time frames, emit
less than half the greenhouse gases of conventional coal generation, have fewer other
emissions and waste product issues, and are often the only utility-scale baseload
resource available. Higher fuel price volatility has historically affected the economics of
Chapter Highlights


Avista’s Climate Policy Council monitors greenhouse gas legislation and
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
natural gas-fired plants, their performance decreases in hot weather conditions, it is
increasingly difficult to secure sufficient water rights for their efficient operation, and they
emit significant greenhouse gases relative to renewable resources.

Renewable energy technologies, including wind, biomass, and solar generation, have
different challenges. Renewable resources are attractive because they have low or no
fuel costs and few, if any, direct emissions. However, solar and wind-based renewable
generation resources have limited or no capacity value for the operation of Avista’s
system, and their variable output presents integration challenges requiring additional
non-variable capacity investments. Even with significant decreases in equipment and
installation costs, renewables are high-cost and suffer from integration challenges.

Renewable projects also draw the attention of environmental groups interested in
protecting visual aspects of landscapes and wildlife populations. Similar to coal plants,
renewable resource projects are often located to maximize their capability rather than to
be near load centers. The need to site renewable resources in remote locations often
requires significant investments in transmission interconnection and capacity expansion,
as well as mitigating possible wildlife and aesthetic issues. Some of these issues may
be alleviated with distributed resources, but the price differentials of distributed
resources make them more difficult to develop at utility scale. Unlike coal or natural gas-
fired plants, the fuel for non-biomass renewable resources may not be transportable
from one location to another to utilize existing transmission facilities 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 is uncertain for several reasons.
Federal investment and production tax credits begin expiring for projects starting
construction after 2013. The continuation of credits and grants cannot be relied upon in
light of the impact such subsidies have on the finances of the federal government, and
the relative maturity of wind and solar technologies. Many relatively unpredictable
factors affect the costs of renewable technologies, such as renewable portfolio standard
goals, construction and component prices, international trade issues, and currency
exchange rates. Capital costs for wind and solar have decreased over the last several
IRPs, but future costs remain uncertain.

Uncertainty still exists about final design and scope of greenhouse gas regulation.
Pockets of strong regional and national support to address climate change exist, but
little political will at the national level to implement significant new laws exists beyond
the regulations proposed by the EPA and is unpredictable going forward. However,
since the 2013 IRP publication, changes in the approach to greenhouse gas emissions
regulation have occurred, including:

 The EPA proposed actions to regulate greenhouse gas emissions under the
CAA through the proposed CPP; and
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
 California’s cap and trade regulation continues scheduled expansion throughout
the economy and includes new linkages with Quebec, and an October 2013
compact to link future programs with British Columbia, Oregon, and Washington.

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 of Avista participate on certain
topics as needed. The monthly meetings for this group include work divided into
immediate and long-term concerns. The 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 data requests
about climate change issues. Longer-term issues involve emissions tracking and
certification, considering the merits of different greenhouse gas policies, actively
participating in the development of 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.

Greenhouse Gas Emissions Concerns for Resource Planning
Resource planning in the context of greenhouse gas emissions regulation raises the
relationships between Avista’s obligations for environmental stewardship and cost
implications for customers. Resource planning considers the cost effectiveness of
resource decisions, as well as the need to mitigate the financial impact of potential
future emissions risks. Although some parties advocate for the immediate reduction or
elimination of certain resource technologies, such as coal or even natural gas-fired
plants, there are economic and reliability limitations among concerns related to pursuing
this type of policy. Technologically, it is possible to replace fossil-fueled generation with
renewables, but this approach results in increased cost to customers and results in
reliability challenges.

State and Federal Environmental Policy Considerations
The CPP is the focus of federal greenhouse gas emissions policies in the 2015 IRP. In
the 2013 plan, Avista did not include a specific dollar amount for cap and trade or a
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
carbon tax on the modeling of the Western Interconnect. Modeling for jurisdictions with
existing costs, such as California and British Columbia, included the appropriate costs.
The 2013 IRP had an implied cost from the replacement of retired coal capacity. The
Expected Case in this IRP includes the probability of a cost of carbon. Details about the
cost of carbon and the modeling results are in Chapter 10 – Market Analysis. The
Expected Case also includes proposed regulatory mechanisms through sections 111(b)
for new sources and 111(d) for existing sources of the Clean Air Act (CAA) as described
below.

The President’s Climate Action Plan, released on June 25, 2013, outlined the Obama
administration’s three pillars of executive action regarding climate change. The pillars
include:
 Reducing U.S. carbon emissions through the regulation of emissions from power
plants, increased use of renewables and other clean energy technologies, and
stronger energy efficiency standards (reflected in the CPP);
 Making infrastructure preparations to mitigate the impacts of climate change; and
 Working on efforts to reduce international greenhouse gas emissions and
prepare for the impacts of climate change.

A presidential memo with several climate related policies went to the EPA Administrator
on the same day as the Climate Action Plan. It directed the EPA to:
 Issue new proposed greenhouse gas emissions standards for new electric
generation resources by September 30, 2013.
 Issue new proposed standards for existing and modified sources by June 1,
2014, final standards by June 1, 2015, and require state implementation plans by
June 30, 2016.

The EPA answered the administration by issuing a new proposal to limit carbon dioxide
emissions from new and modified coal and natural gas-fired electric generating units in
late 2013, and from existing sources in June 2014. Details of these proposals are later
in this chapter.

The federal Production Tax Credit (PTC), Investment Tax Credit (ITC), and Treasury
grant programs are key federal policy considerations for incenting the development of
renewable generation. The current PTC and ITC programs are available for non-solar
projects that began construction before the end of 2013 and for solar projects before the
end of 2016. Avista did not model an extension of these tax incentives because of the
uncertainty of their continuation. This situation may change and would affect modeling
assumptions for the 2017 IRP. Extension of the PTC may accelerate the development
of some regional renewable energy projects. This may affect the development of
renewable projects in the Western Interconnect, but not necessarily for Avista, because
the current resource mix and low projected load growth do not necessitate the
development of new renewables in this IRP.

Chapter 7–Policy Considerations
Avista Corp 2015 Electric 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 that 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.

Promulgated PSD permit rules may affect Avista’s thermal generation facilities in the
future. These rules can affect the amount of time it takes to obtain permits for new
generation and 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. The Title V operating permit for Colstrip Units 3 and 4 expires in 2017.
The Coyote Springs 2 permit expires in 2018. A new Title V operating permit for the
Kettle Falls generating station is expected in 2016, and the Rathdrum CT expires in
2016. 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.

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)
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
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
will guide modeling assumptions for the 2017 IRP.

Clean Power Plan Proposal
The EPA issued the draft CPP on June 2, 2014. The modeling for this IRP was based
on the CPP proposal. This plan aims to reduce national greenhouse gas emissions from
covered fossil-fueled electric generating units by 30 percent by 2030 from a 2005
baseline, with an interim goal in 2020. The draft rule calculated emission rate targets for
each state using a combination of four building blocks:

1. Heat rate improvements at coal plants up to 6 percent;
2. Displacement of coal-fired and oil-fired steam generation by increasing
utilization of natural gas-fired combined cycle plants up to a 70 percent capacity
factor;
3. Use of more low- or zero-carbon emitting generation resources (including 6
percent of nuclear capacity); and
4. Increase demand side efficiency by 1.5 percent per year between 2020 and
2029.

The EPA used 2012 data for the baseline for each state. The building blocks could
constitute the best system of emission reduction a state could propose in its compliance
plan. However, states might also propose to comply through other measures, including
a cap and trade form of regulation. The state of Washington, through the provisions of
the EIA (Chapter 19.285 RCW), currently applies renewable energy and energy
efficiency standards to Avista’s electric operations. The state also imposes an
emissions performance standard under Chapter 80.80 RCW to long-term financial
commitments made by electric utilities when acquiring new baseload generation or
upgrading existing fossil-fueled baseload generation.

Several aspects of the proposed CPP are problematic. The TAC discussed these issues
in several of its meetings. Issues include the impact of the 2012 baseline year on
hydroelectric generation, the affect on combined cycle resources in Idaho, the
immediate impact of the first two building blocks on the 2020 interim goal, and the short
time to develop regional solutions in light of the interim goal and legislation that may be
required from some of the states. Some adjustments to modeling for the 2015 IRP
attempt to alleviate some of these issues to make them into a workable plan. Updates to
2017 IRP modeling assumptions will account for changes made in the final CPP and
subsequent state implementation plans. The EPA issued the final CPP on August 3,
2015. The final rule differs from the proposed rule in many ways including the removal
of the fourth building block (energy efficiency), movement of the start date from 2020 to
2022, and adjusted goals for many states. The 2017 IRP will account for these changes,
since modeling for the 2015 concluded in early 2015.

Figure 7.1 includes the IRP’s adjusted 2030 goal in comparison to the 2012 baseline.
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
The orange portion of the bar shows the proposed reduction. Washington State has the
highest percentage reduction, followed by Arizona. Idaho has the lowest reduction after
an assumed adjustment for 2012 partial year of operations at Langley Gulch.

Figure 7.1: Draft Clean Power Plan 2030 Emission Intensity Goals

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 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 and 4’s existing emission
control systems should be sufficient to meet mercury limits. For the remaining portion of
the rule specifically addressing air toxics (including metals and acid gases), the joint
owners of Colstrip are currently evaluating what type of new emission control systems
0
500
1,000
1,500
2,000
2,500
AZ CA CO ID MT NM NV OR UT WA WY
EP
A

l
b
s

p
e
r

M
W
h
2030 Goal EPA Reduction
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
will be required to meet MATS compliance in 2016. Avista is unable to determine to
what extent, or if there will be any, material impact to Colstrip Units 3 and 4 at this time.

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, 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.

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. The state of
Washington requires mandatory greenhouse gas emissions reporting similar to the EPA
requirements. Oregon has similar reporting requirements.

Coal Ash Management and Disposal
On December 19, 2014, the EPA issued a final rule regarding coal combustion residuals
(CCR). This will affect 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 final rule has not yet been published in the Federal Register. 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 at this 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 an emissions performance standard 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 jurisdictions where
Avista operates 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 such as Avista. There are 29 states, plus the District of Columbia,
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
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 a renewable portfolio standard
(RPS). There is no indication that Idaho is moving toward the active regulation of
greenhouse gas emissions beyond the CPP. The Idaho Department of Environmental
Quality will administer greenhouse gas standards under its CAA delegation from the
EPA.

Montana Policy Considerations
Montana has a non-statutory goal to reduce greenhouse gas emissions to 1990 levels
by 2020. Montana’s RPS law, enacted through Senate Bill 415 in 2005, requires utilities
to meet 10 percent of their load with qualified renewables from 2010 through 2014, and
15 percent beginning in 2015. Avista is exempt from the Montana RPS and its reporting
requirements beginning on January 2, 2013, with the passage of SB 164 and its
signature by the Governor.

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 meeting Montana’s mercury emissions
goal.

Oregon Policy Considerations
The State of Oregon has a history of considering 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
not adopted any comprehensive requirements. 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 paying into the Climate Trust of Oregon. Senate
Bill 838 created a renewable portfolio standard requiring large electric utilities to
generate 25 percent of annual electricity sales with renewable resources by 2025.
Intermediate term goals include 5 percent by 2011, 15 percent by 2015, and 20 percent
by 2020. Oregon ceased being an active member in the Western Climate Initiative in
November 2011. The Boardman coal plant is the only active coal-fired generation facility
in Oregon; by the end of 2020, it will cease burning coal. The decision by Portland
General Electric to make near-term investments to control emissions from the facility
and to discontinue the use of coal, serves as an example of how regulatory,
environmental, political, and economic pressures can culminate in an agreement that
results in the early closure of a coal-fired power plant.
1 http://www.dsireusa.org/rpsdata/index.cfm
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
Washington State Policy Considerations
Similar circumstances leading to the closure of the Boardman facility in Oregon
encouraged TransAlta, the owner of the Centralia Coal Plant, to agree to shut down one
unit at the facility by December 31, 2020, and the other unit by December 31, 2025. The
confluence of regulatory, environmental, political, and economic pressures brought
about its scheduled closure. The state of Washington 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. Washington’s EIA, passed in the November 2006 general
election, established a requirement for utilities with more than 25,000 retail customers to
use qualified renewable energy or renewable energy credits to serve 3 percent of retail
load by 2012, 9 percent by 2016, and 15 percent by 2020. Failure to meet these RPS
requirements results in at least a $50 per MWh fine. The initiative also requires utilities
to acquire all cost-effective conservation 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.
A utility can also comply with the renewable energy standard by investing in at least 4
percent of its total annual retail revenue requirement on the incremental costs of
renewable energy resources and/or renewable energy credits. In 2012, Senate Bill 5575
amended the EIA to define Kettle Falls Generating Station and other legacy biomass
facilities that commenced operation before March 31, 1999, as EIA qualified resources
beginning in 2016. A 2013 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 beginning in
2016. The 2015 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
Chapter 7–Policy Considerations
Avista Corp 2015 Electric IRP
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 (Commerce) has
commenced a process expected to adopt a lower emissions performance standard in
2013; a new standard would not be applicable until at least 2017. Commerce filed a final
rule with 970 pounds per MWh for greenhouse gas emissions on March 6, 2013, with
rules becoming effective on April 6, 2013.2

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 did 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 regulatory program will not itself include the establishment of
an emissions trading market, but other entities could develop such a system to facilitate
trading.

2 http://www.commerce.wa.gov/Programs/Energy/Office/Utilities/Pages/EmissionPerfStandards.aspx

Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-1

8. Transmission & Distribution Planning

Introduction
Avista delivers electricity from generators to customer meters through a network of
conductors and ancillary equipment. Avista categorizes its energy delivery systems
between transmission and distribution voltages. Avista’s transmission system operates
at 115 and 230 kV nominal voltages; the distribution system operates between 4.16 and
34.5 kV, but typically at 13.2 kV in urban service centers. In addition to voltages, the
transmission system operates distinctly from the distribution system. For example, the
transmission system is a network linking multiple sources with multiple loads, while the
distribution system configuration uses radial feeders to link a single source to multiple
loads.

Coordinating transmission system operations and planning activities with regional
transmission providers maintains reliable and economic transmission service for our
customers. Transmission providers and interested stakeholders coordinate regional
planning, construction, and operations under Federal Energy Regulatory Commission
(FERC) rules and guidance from state and local agencies. This chapter complies with
Avista’s FERC Standards of Conduct compliance program governing communications
between Avista merchant and transmission functions.

This chapter describes Avista’s completed and planned distribution feeder upgrade
program, the transmission system, completed and planned upgrades, and estimated
costs and issues of new generation resource integration.

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

Local Transmission Planning Report
Avista’s local planning report is the product of both a local transmission planning
process and an annual planning assessment. Attachment K to Avista’s Open Access
Transmission Tariff (OATT) FERC Electric Volume No. 8 outlines the local transmission
Chapter Highlights





Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-2

planning process. This process identifies single-system projects needed to mitigate
future reliability and load-service requirements for the Avista transmission system.

The annual planning assessment is outlined by North American Electric Reliability
Corporation (NERC) Reliability Standard TPL-001-4. The planning assessment
determines where the system may have the inability to meet performance requirements
as defined in the NERC Reliability Standards and identifies corrective action plans
addressing how to meet the performance requirements. The planning assessment
includes performing steady state contingency analysis, voltage collapse, and transient
technical studies.

The local planning report supports compliance with the local transmission planning
process and applicable NERC reliability standards. The local planning report, with its
associated collection of single-system projects and corrective Action Plans, provides a
10-year transmission system expansion plan by including all transmission system facility
improvements.

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 also 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
8.1.

Figure 8.1: NERC Interconnection Map

1 https://www.wecc.biz/Pages/About.aspx
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-3

Peak Reliability
The Peak Reliability (Peak) organization took over the role of reliability coordinator from
WECC on February 12, 2014. Peak is wholly independent of WECC, performing the
reliability coordinator and interchange authority functions for the Western
Interconnection.2

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. Smaller, principally non-generating utilities participate in an indirect manner
through their member systems, such as the BPA.

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). The
TPC exists as a forum addressing northwest electric planning issues and concerns,
including a structured interface with external stakeholders.

ColumbiaGrid
ColumbiaGrid began 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 members, Avista participates in the NTTG planning process to foster
collaborative relationships with our interconnected utilities.
2 https://www.peakrc.com/aboutus/Pages/History.aspx

Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-4

BPA Transmission System
BPA owns and operates over 15,000 miles of transmission-level facilities and owns over
three-quarters of the region’s high voltage (230 kV or higher) transmission grid. Avista
uses BPA transmission to transfer output from its remote generation sources to Avista’s
transmission system, including its share in Colstrip Units 3 and 4, Coyote Springs 2, and
its WNP-3 settlement contract. Avista also contracts for BPA transmission to transfer
power to several delivery points on the BPA system serving portions of our retail load
and for selling surplus power to other parties in the region.

Avista participates in BPA transmission rate case processes and in BPA’s Business
Practices Technical Forum to ensure charges remain reasonable and support system
reliability and access. Avista works with BPA and other regional utilities to coordinate
major transmission facility outages.

Future electric grid expansion likely will require transmission expansion by federal and
other entities. BPA is developing several transmission projects in the Interstate-5
corridor and in southern Washington to maintain reliable system operation and integrate
regional wind generation resources. Each project has the potential to increase BPA
transmission rates and thereby affect Avista’s costs.

Avista’s Transmission System

Reliability and Operations
Avista plans and operates its transmission system pursuant to applicable criteria
established by the NERC, WECC, and NWPP. Through involvement in WECC and
NWPP standing committees and sub-committees, Avista participates in developing new
and revised criteria while coordinating transmission system planning and operation with
neighboring systems. Mandatory reliability standards promulgated through FERC and
NERC subject Avista to periodic performance audits through these regional
organizations.

Avista’s transmission system provides reliable and efficient transmission service from
the company’s generation resources to its retail and wholesale customers.
Transmission capacity surplus to retail load service needs is available to other parties
pursuant to FERC regulations and the terms and conditions of Avista’s OATT. Avista
markets its unsold surplus transmission capacity on a long-term (greater than one year)
basis and short-term basis to other parties as part of Avista’s overall resource
optimization efforts.

System Topology
Avista owns and operates over 2,200 miles of electric transmission facilities. This
includes approximately 685 miles of 230 kV line and 1,527 miles of 115 kV line. Figure
8.2 illustrates Avista’s transmission system.

Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-5

Figure 8.2: Avista Transmission Map

Avista owns an 11 percent interest in 495 miles of double circuit 500 kV lines between
Colstrip and Townsend, Montana. The transmission system includes switching stations
and high-voltage substations with transformers, monitoring and metering devices, and
other system operation-related equipment. The system transfers power from Avista’s
generation resources to its retail load centers. Avista has network interconnections with
the following utilities:

 BPA
 Chelan County PUD
 Grant County PUD
 Idaho Power Company
 NorthWestern Energy
 PacifiCorp
 Pend Oreille County PUD

Transmission System Information
Since the 2013 IRP, Avista completed several transmission projects to support new
generation, increase reliability, and provide system voltage support.

Transmission Line Upgrades
 Chelan – Stratford 115 kV: line reconductor
 Garden Springs to Hallet & White section of South Fairchild 115 kV Tap: line
reconductor
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-6

 Irvin – Opportunity 115 kV line: new line section
 Burke to Montana border section of Burke – Thompson Falls A&B 115 kV lines
 Southern half of Bronx – Cabinet Gorge 115 kV line: line reconductor

Stations
 Stratford 115 kV – station rebuild
 Odessa 115 kV – capacitor bank installed
 Lancaster 230 kV station interconnection
 Lind 115 kV – capacitor bank installed
 Moscow 230/115 kV – station rebuild
 Blue Creek 115 kV – station rebuild
 Beck Road 115 kV – new station
 Clearwater 115 kV – station upgrade
 Lewiston Mill Road 115 kV – new station
 North Lewiston 115 kV Distribution Substation

Planned Projects
Avista plans to complete several re-conductor projects throughout its transmission
system over the next decade. These projects focus on replacing decades-old small
conductor with new conductor capable of greater load-carrying capability and fewer
electrical losses. The following list gives an example of planned transmission projects:

Transmission Lines
 Addy – Devil’s Gap 115 kV
 Bronx – Cabinet Gorge 115 kV (2011-2017)
 Burke – Pine Creek 115 kV (2012-2015)
 Benton – Othello 115 kV (2014-2016)
 Devils Gap – Lind 115 kV (2014-2016)
 Devil’s Gap – Stratford 115 kV (2019)
 Coeur d’Alene – Pine Creek 115 kV (2014-2018)
 Spokane Valley Reinforcement Project (2011-2016)
Stations
 Irvin 115 kV Switching Station [Spokane Valley Reinforcement] (2016)
 Millwood 115 kV Distribution Substation [Spokane Valley Reinforcement] (2013)
 Harrington 115 kV Distribution Substation (2014)
 Noxon 230 kV Switching Station (2013-2018)
 9th & Central 115 kV Distribution Substation (2015)
 Greenacres 115 kV Distribution Substation (2014)
 Beacon 230/115 kV Station Partial Rebuild (2017+)
 Saddle Mountain 115 kV Station (new, 2018)
 Westside 230/115 kV transformer (2016)

IRP Generation Interconnection Options
Table 8.1 shows the projects and cost information for each of the IRP-related locational
studies where Avista evaluated new generation options. The study details for each
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-7

project, including cost and integration options, are in Appendix E. These studies provide
a high-level view of the generation interconnect requests, and are similar to third-party
feasibility studies performed under Avista’s generator interconnection process. Because
the FERC does not allow complete charging of integration costs benefiting the overall
transmission system to the new generator, it is unlikely that the entirety of these figures
will actually be charged to a new interconnected generator. There are cost ranges for
each proposed generation project because there are alternate solutions to reinforce the
transmission system to support the proposed interconnected generation levels.

Table 8.1: 2015 IRP Requested Transmission Upgrade Studies

Project Size (MW) Cost Estimate (Millions)3

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. The studies
typically take at least one year to complete. 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. Each of the proposed projects becomes public to
some degree, but customer names remain anonymous. Table 8.2 lists major projects
currently in Avista’s interconnection queue.

Table 8.2: Third-Party Large Generation Interconnection Requests

Project Size (MW) Type Interconnection
#43 150 Wind Lind 115 kV Substation
#44 600 Pumped Hydro Colstrip 500 kV System

3 Cost estimates are in 2014 dollars and use engineering judgment with a 50 percent margin for error.
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-8

Distribution System Efficiencies
Avista’s distribution system consists of approximately 330 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.

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 the summer of 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 both
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.

Grid Modernization
Building on a 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 customers with refreshed system feeders with new automation capabilities
across the company’s distribution system. Table 8.3 contains a list of completed and
planned feeder upgrades.

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 work
of 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
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-9

distribution system that is more efficient, easier to maintain, and more reliable for our
customers.

Figure 8.3 illustrates the reliability advantages and reasons for the Grid Modernization
Program. Prior to the 2009 feeder rebuild pilot program, 39 outages per year were
expected. After the project, outages declined significantly to an average of 20 unique
outages. In the past two years, only one outage occurred. The program is in its second
year of regular funding and is realizing its intended purpose of capturing energy savings
through reduced losses, increased reliability, and decreased O&M costs. Table 8.3
shows the feeders addressed through this program to date and projects currently in
progress. The total energy savings from both re-conductor and transformer efficiencies
for all completed feeders is approximately 7,479 MWh annually.

Figure 8.3: Spokane’s 9th and Central Feeder (9CE12F4) Outage History

0
10
20
30
40
50
60
70
2006 2007 2008 2009 2010 2011 2012 2013 2014
Nu
m
b
e
r

o
f

O
u
t
a
g
e
s
Chapter 8 – Transmission & Distribution Planning

Avista Corp 2015 Electric IRP 8-10

Table 8.3: Completed and Planned Feeder Rebuilds

Feeder Area Year
Complete
Annual Energy
Savings (MWh)
9CE12F4 Spokane, WA (9th & Central) 2009 601
BEA12F1 Spokane, WA (Beacon) 2012 972
F&C12F2 Spokane, WA (Francis & Cedar) 2012 570
BEA12F5 Spokane, WA (Beacon) 2013 885
WIL12F2 Wilbur, WA 2013 1,403
CDA121 Coeur d’Alene, ID 2013 438
OTH502 Othello, WA 2014 21
RAT231 Rathdrum, ID 2014 0
M23621 Moscow, ID 2015 413
WIL12F2 Wilbur, WA 2015 1,403
WAK12F2 Spokane, WA (Waikiki) 2016 175
RAT233 Rathdrum, ID 2019 471
SPI12F1 Northport, WA (Spirit) 2019 127
Total 7,479

Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
9. Generation Resource Options

Introduction
Several generating resource options are available to meet future load growth. 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 2015 IRP to meet future needs. The resources described in this
chapter are mostly generic, as actual resources identified through a competitive process
may differ in size, cost, and operating characteristics due to siting or engineering
requirements.

Assumptions
Avista only considers 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 wind, energy storage, photovoltaic solar,
hydroelectric upgrades, and thermal unit upgrades. Several other resource options
described later in the chapter were not included in the PRS analysis, but discussed as
potential resource options that may respond to a future RFP. The IRP excludes
potential contractual arrangements with other energy companies as an option in the
plan, but such arrangements may be an option when Avista seeks new resources
through a competitive acquisition process.

The resource costs of each resource option include transmission expenses, as
described in Chapter 8 – Transmission & Distribution Planning. Levelized costs result
from discounting nominal cash flows by a 6.58 percent-weighted average cost of capital
approved by the states of Idaho and Washington in recent rate case filings. All costs in
this section are in 2015 nominal dollars unless otherwise noted.

Many renewable resources are eligible for federal and state tax incentives. Federal
solar tax benefits fall by two-thirds after 2016; federal production tax credits (PTCs) are
no longer available unless meeting certain provisions. Incentives, to the extent they are
Section Highlights


Upgrades to Avista’s


Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
available, are included in IRP modeling. The IRP amortizes investment tax credits over
the life of the asset per regulatory accounting rules.

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
characteristic and cost information from Thermoflow.

Levelized resource costs illustrate the cost 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 the facilities. 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 planned to operate only a few hours each year. Levelized energy
costs fairly compare renewable resources to the energy component of natural gas-fired
resources because renewable technologies are not dispatchable.

The following cost items are in the levelized cost calculations 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, a cash
grant, or an investment tax credit (ITC), available to qualified generation options.
 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 that are not based on
generation levels.
 Variable O&M: Costs per MWh related to incremental generation.
 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.
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
 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.1
Table 9.1 compares the levelized costs of different resource types.

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

Advanced Large Frame CT $58 $130 220
Modern Large Frame CT $57 $124 186
Advanced Small Frame CT $64 $151 102
Frame/Aero Hybrid CT $46 $164 106
Small Reciprocating Engine Facility $41 $159 93
Modern Small Frame CT $59 $188 49
Aero CT $54 $202 45
1 x 1 Advanced CCCT $37 $211 362
1 x 1 Modern CCCT $37 $210 306

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 600
MW. Avista would need to share the plant with one or more utilities 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 some water
is available for plant cooling, but only enough for a hybrid system utilizing the benefits of
combined evaporative and convective technologies.

1 Peak credit is the amount of capacity a resource contributes at the time of system peak load.
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
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 being commonly used technologies in the
Northwest. Where 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.2 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 machine’s heat rate is 6,720 Btu/kWh in 2016.3 The larger machine is 6,631
Btu/kWh. The plants include duct firing for 7 percent of rated capacity at a heat rate of
7,912 and 7,843 Btu/kWh, respectively.

The IRP includes a 3 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 anticipated capital costs for the two CCCTs, located in Idaho on Avista’s
transmission system with AFUDC on a green field site, are $1,177 per kW for the
smaller machine and $1,120 per kW (2016$) 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
allow the plants to start and ramp quickly, providing regulation services and reserves for
load following and to integrate variable resources such as wind and solar.

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 and operational estimates based on internal engineering estimates. All
2 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. 3 Heat rates shown are the higher heating value.
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
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
$638 $2.08 9,931 $3.65 1 203 203 $129
Modern Large
Frame CT
$667 $2.08 10,007 $2.60 1 170 170 $114
Advanced Small
Frame CT
$853 $3.13 11,265 $2.60 1 96 96 $82
Frame/Aero
Hybrid CT
$1,016 $3.13 8,916 $3.13 1 101 101 $103
Small
Reciprocating
Engine Facility
$546 $8.33 7,700 $3.13 10 9.3 93 $51
Modern Small
Frame CT
$1,265 $4.17 10,252 $2.60 1 45 45 $57
Aero CT $1,316 $6.25 9,359 $2.60 1 42 42 $56
1 x 1 Modern
CCCT
$1,120 $18.7
5
6,771 $3.91 1 341 341 $382
1 x 1 Advanced
CCCT
$1,177 $15.6
3
6,845 $3.13 1 286 286 $336

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 its 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, three options exist: contracting for firm natural gas
transportation rights, on-site oil, or 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.
The 2013 “Fiscal Cliff” deal in the U.S. Congress extended the PTC for wind through
December 31, 2013, with provisions allowing projects to qualify after 2013 if
construction began in 2013. This IRP does not assume the PTC extends beyond this
term, but does assume the preferential five-year tax depreciation remains.

Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
Wind resources benefit from having no emissions or fuel costs, but they are not
dispatchable, and have high capital and labor costs on a per-MWh basis when
compared to most other resource options. Wind capital costs in 2016, including AFUDC,
are $2,234 per kW, with annual fixed O&M costs of $46 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 $4.30 per kW-year in 2016. 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 35 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 impacts 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
2016, the estimated “all-in” cost is $102 per MWh; but, direct cost to customers would
be $70 per MWh with the ITC. This plan assumes wind resources selected in the PRS
include the 20 percent REC apprenticeship adder for the EIA. Qualification for the adder
requires 15 percent of construction labor by state-certified apprentices.

Figure 9.1: Northwest Wind Project Levelized Costs per MWh

102
109
116
124
131
70 71 75 80 86
60 61 65 69 74
$0
$20
$40
$60
$80
$100
$120
$140
2016 2020 2025 2030 2035
No
m
i
n
a
l

L
e
v
e
l
i
z
e
d

$
/
M
W
h
Expected Case 30% ITC 30% ITC + 40% CF
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
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 that 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.

Where a substantial amount of PV solar is added to a summer peaking utility system,
such as one located in the Desert Southwest, the peak hour recorded prior to the
installation will be reduced, but the peak hour will shift 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 that solar does
not reduce Avista’s winter peak, reduces the summer peak, and is less effective at
reducing peak as more solar is installed.

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%

Solar-thermal technologies can produce capacity factors as much as 30 percent higher
than PV solar projects and can store energy for several hours 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 the
IRP.

Utility-scale PV solar capital costs in the IRP, including AFUDC, are $1,500 per kW for
fixed panel and $1,600 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
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
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 to account for panel degradation.

Figure 9.2 shows the levelized costs of solar resources, including applicable federal and
state incentives, on-line dates, and capacity factors. The costs are specific to Avista
acquisition and ownership. The State of Washington offers a number of incentives for
solar installations. First, 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.

Figure 9.2: Solar Nominal Levelized Cost ($/MWh)

Energy Storage
Increasing solar and wind generation on the electric grid makes energy storage
technologies attractive from an operational perspective. Storage could be an ideal way
to smooth out renewable generation variability, 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 over supply 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,
flow batteries, flywheels, and compressed air.

Storage may become an important part of the nation’s electricity grid if the technology
overcomes a number of large 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 high, at near $3,455 per kW, or nearly
150
137 141 148 155
138
129 133 139 145
94 88 91 95 99
$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
$200
2016 2020 2025 2030 2035
No
m
i
n
a
l

L
e
v
e
l
i
z
e
d

$
/
M
W
h
Expected Case 30% ITC 30% ITC + 20% CF
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
three times the initial cost of a natural gas-fired peaking plant that can provide many of
the same capabilities without the electricity consumption characteristics of storage.
Storage costs will decline over time, and Avista continues to 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. Finally, early technology adoption can be risky,
with industry examples of battery fires and financial issues.

To learn more about storage technology and its potential, Avista recently installed a
vanadium flow battery in Pullman, Washington. This installation, known as the Turner
Energy Storage Project, will provide insight about the technology’s reliability, its
potential benefit to the transmission and/or distribution systems, and potential power
supply benefits including oversupply events. The battery has one megawatt of power
capability and three megawatt-hours of energy storage. A Washington state grant for
research and development partially funded this storage project.

Turner Energy Storage Project, Pullman, WA

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 in meeting the need for additional operational flexibility if upfront capital
costs and operational losses fall.

In addition to capital costs, storage projects O&M costs are $20 per kW-year, and
recharge costs use off-peak Mid-Columbia energy prices. Levelized storage project
costs are highly 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.
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
The nominal levelized capacity cost for storage is approximately $580 per kW-year and
energy costs $35/MWh.

Other Generation Resource Options
A thorough IRP analyzes generation resources not readily available in large quantities,
not commercially available, not economically ready for utility-scale development, or
prohibited by state policy. Several emerging technologies, like energy storage, are
attractive from an operational or environmental perspective, but are significantly higher-
cost than other technologies providing similar capabilities at lower cost. The resources
include biomass, geothermal, co-generation, nuclear, landfill gas, and anaerobic
digesters. This plan does not model these resource options explicitly, but continues to
monitor their viability.

Exclusion from the PRS is not the last opportunity for non-modeled technologies to be
part of Avista’s future portfolio. The resources compete with those included in the PRS
through competitive acquisition processes. Competitive acquisition processes identify
technologies that might displace resources otherwise included in the IRP strategy.
Another possibility is acquisition through federal PURPA mandates. PURPA provides
non-utility developers the ability to sell qualifying power to Avista at set prices and
terms.4

Woody Biomass Generation
Woody biomass generation projects use waste wood from lumber mills or forest
restoration processes. 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 project, Avista will
consider it for a future acquisition.

Geothermal Generation
Northwest utilities have shown increased interest in geothermal energy over the past
several years. It 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
projects are likely to be developed.

Geothermal energy struggles to compete due to high development costs stemming from
having to drill several holes thousands of feet below the earth’s crust; each hole can
4 Rates, terms, and conditions are available at www.avistautilities.com under Schedule 62.
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
cost over $3 million. Ongoing geothermal costs are low, but the capital required locating
and proving a viable site is significant. Costs shown in this section do not account for
the dry-hole risk associated with sites that do not prove to be viable after drilling has
taken place.

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 it 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 has 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.

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 fewer than five megawatts.
Most facilities are located at large dairies and feedlots. A survey of Avista’s service
territory found no large-scale livestock operations capable of implementing this
technology.

Wastewater treatment facilities can also 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 waste water system under a
PURPA contract with a Spokane County facility.

Small Cogeneration
Avista has few industrial customers capable of developing cost-effective cogeneration
projects. 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.

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.

Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
A big challenge in developing any new cogeneration project is aligning the needs of the
cogenerator and the utility’s 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.

Nuclear
Avista does not include nuclear plants as a resource option in the IRP given the
uncertainty of their economics, the apparent lack of regional political support for 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 a reality.

The viability of nuclear power could change as national policy priorities focus attention
on de-carbonizing the nation’s energy supply. The lack 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. New smaller, and more modular,
nuclear design could increase the potential for nuclear by shortening the permitting and
construction phase, and make these traditionally large projects better fit the needs of
smaller utilities.

Coal
The coal generation industry is at a crossroads. In many states, like Washington, new
coal-fired plants are unlikely due to emission performance standards and the shortage
of utility scale carbon capture and storage projects. Federal guidelines under section
111(b) of the CAA and the CPP likely prevent or restrict the construction of new coal
generation. The final rule was not available at the time this section’s drafting. The risks
associated with future carbon legislation and projected low natural gas costs make
investments in this technology challenging.

Hydroelectric Project Upgrades and Options
Avista continues to upgrade its hydroelectric facilities. The latest hydroelectric upgrade
added nine megawatts to the Noxon Rapids Development in April 2012. Figure 9.3
shows the history of upgrades to Avista’s hydroelectric system. Avista added 40.1 aMW
of incremental hydroelectric energy between 1992 and 2012. Upgrades completed after
1999 can qualify for the EIA, thereby reducing the need for additional renewable energy
options.

Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
Figure 9.3: Historical and Planned Hydro Upgrades

Avista is currently upgrading the Nine Mile powerhouse, replacing two of its four turbine
generator units. Avista removed the last two original 1908 units in 2013 and began a
project to replace the 107-year old technology with new turbine generators, generator
step-up transformer, switchgear, exciters, governors and controls in 2014. Avista
expects to complete the project in 2016.

The Spokane River hydroelectric construction occurred in the late 1800s and early
1900s, when the priority was to meet then-current loads. The developments currently do
not capture a majority of the river flow as their original designs only met then-current
loads and not river capacity. In 2012, Avista reassessed its Spokane River
developments to evaluate opportunities to take advantage of 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 at each. Each upgrade option should qualify for the EIA,
meeting the Washington state 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. A summary of the upgrade
options is in Table 9.4. The upgrades will compete against other renewable options
when more renewables are required in future.

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
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
Table 9.4: Hydroelectric Upgrade Options

Resource Post
Falls
Monroe
Street/Upper
Falls
Long
Lake
Cabinet
Gorge
Incremental Capacity (MW) 22 80 68 110
Incremental Energy (MWh) 90,122 237,352 202,592 161,571
Incremental Energy (aMW) 10.3 27.1 23.1 9.2
Peak Credit (Winter/ Summer) 24/0 31/0 100/100 0/0
Capital Cost ($ Millions) $136 $193 $179 $286
Levelized Energy Cost ($/MWh) $159 $93 $112 $197

Long Lake Second Powerhouse
Avista studied adding a second powerhouse at Long Lake over 20 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 provide incremental capacity to meet peak
load growth.

The 2012 study focused on three alternatives. The first replaces the existing four-unit
powerhouse with four larger units to total 120 MW, increasing capacity by 32 MW. The
other two alternatives develop a second powerhouse with a penstock beginning from a
new intake structure just 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.

Post Falls Refurbishment
The Post Falls hydroelectric development is 109 years old. Three alternatives could
increase the existing capacity from 18 MW up to 40 MW. The first option is a new two-
unit 40 MW powerhouse on the south channel that replaces the existing powerhouse.
Alternative 2 retrofits the existing powerhouse with five 8.0 MW units (40 MW total). The
last alternative retrofits the existing powerhouse with six 5.6-MW units (33.6 MW total).
The cost differences between developing a new powerhouse in the south channel and
the smaller plant refurbishment is small. Studies of alternatives to address the aging
infrastructure of the plant will continue over the next decade.

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
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
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 110 MW underground powerhouse would benefit from an existing diversion tunnel
around the dam built during original 1952 construction.

Thermal Resource Upgrade Options
The 2013 IRP identified several thermal upgrade options for Avista’s fleet. 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 PowerPhaseLLC, the
technology increases airflow through a combustion turbine compressor increasing
machine output. This upgrade increases 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 2020 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 will be at the end of its expected life. At this time, Avista could
spend additional capital and upgrade the unit by 12 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.

Chapter 9- Generation Resource Options
Avista Corp 2015 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 3 percent of load and 3
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 10
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 10 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 investing in storage
technology. Today storage has a limited role due to cost and technology infancy. This
analysis studies the potential financial value storage brings to Avista’s power supply
costs based on 2012 actual data and average hydroelectric conditions. The study
includes several storage capacities with storage to peak ratio of three to one and 85
percent efficiency. Table 9.5 is the value 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 is optimizing the
battery to charge in low prices and discharging 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.

Chapter 9- Generation Resource Options
Avista Corp 2015 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. 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 this type of resource. 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%
Chapter 9- Generation Resource Options
Avista Corp 2015 Electric IRP
Table 9.5: Natural Gas-Fired Facilities Ancillary Service Value

Resource Type Capabilities Annual $/kW
Value
CCCT Load Following/ Spin5, 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

Currently, there is not a mature ancillary services market in the Northwest, so ancillary
service values are the costs of operating Avista’s system differently to provide more
ancillary services relative to traditional wholesale energy sales. The ancillary service
values of both storage and natural gas-fired technology were less than expected prior to
the analysis. Avista concluded that the results were reasonable for one primary reason:
having a large hydroelectric system, Avista’s system has a significant amount of
flexibility relative to its load variability in most periods. With as the addition of more
variable generation resources, the value of ancillary services capacity should rise.
Figure 9.5 details the significant surplus of ancillary service generation Avista’s system
contains. While the system can become constrained during peak load periods, the large
value in these periods is not as significant when averaged over the entire year.

Figure 9.5: Avista’s Monthly Up/Down Regulation Surplus

5 Fast start CCCTs may have some non-spin reserve capability.
0
100
200
300
400
500
600
700
800
900
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Me
g
a
w
a
t
t
s
Up Capacity Down Capacity
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
10. Market Analysis

Introduction
This section describes the electricity, natural gas, and other markets studied in the 2015
IRP. It contains price risks Avista considers to meet customer demands at the lowest
reasonable cost. The analytical foundation for the 2015 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
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 various sub-markets within the Western Interconnect and the changing
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 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), resource dispatch costs and
values, 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
 Natural gas, solar, and wind resources dominate new generation additions in
the Western Interconnect.
 Clean Power Plan regulation could cause large price and costs swings, but
without a final rule and state compliance plans, the impacts are unknown at
this time.
 The Expected Case forecasts a continuing reduction of Western Interconnect
greenhouse gas emissions due to coal plant closures brought on by federal
and state regulations and low natural gas prices.
Chapter 10- Market Analysis
Avista Corp 2015 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 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 2015 IRP relies on a load forecast for each zone of the Western Interconnect.
Avista uses other utilities’ resource plans and regional plans to quantify load growth
across the west. These estimates include energy efficiency, customer-owned
generation, plug-in electric vehicles, and demand response reductions within the
trajectory. Forecasting future energy use is difficult because of large uncertainties with
the long-term drivers of future energy use.

Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.2 shows regional load growth estimates. The total of the forecasts show
Western Interconnect loads rising nearly 1.1 percent annually over the next 20 years.
On a regional basis, the Northwest will grow at 0.73 percent, California at 1 percent, the
Rocky Mountain States at 1 percent, and the desert Southwest region is lower than
previous forecasts at 0.75 percent. The strongest projected growth area in the region
comes from Canada at 2 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 future 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 that cannot economically meet stricter air
emissions standards and compete with lower-cost natural gas-fired facilities.

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.
Remaining OTC natural gas-fired and nuclear facilities with more favorable economics
are candidates for retrofitting with new cooling technology. The IRP models the closure
of OTC plants with identified shutdown dates from their utility owners’ IRPs and news
releases. Elimination of OTC plants in California will eliminate older technology
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
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
presently used for reserves and high demand hours. Replacement plants will be
expensive for California customers, but a more modern and efficient generation fleet will
serve customers.

Coal-fired facilities face increasing regulatory scrutiny. In the Northwest, the Centralia
and Boardman coal plants will retire by the end of calendar years 2020 and 2025
respectively, for a reduction of 1,961 megawatts. Other coal-fired plants throughout the
Western Interconnect have announced plant closures, including Four Corners, Carbon,
Arapahoe, San Juan, Reid Gardner, and Corette. The Nevada legislature successfully
placed into law a plan to retire all in-state coal plants, and PacifiCorp appears poised to
retire many plants as indicated in its most recent IRP. Over the next 20 years, roughly
45 percent of the 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 29 gigawatts by 2035. Avista does not forecast any additional large
coal facility retirements in its Expected Case.

Figure 10.3: Resource Retirements (Nameplate Capacity)

New Resource Additions
New resource capacity is required to meet future load growth and replace retired power
plants over the next 20 years. To fill the gap, the model adds new resources in each
region to maintain a 5 percent Loss of Load Probability (LOLP). This means meeting all
system demand in 95 percent of simulated forecasts. The generation additions meet
capacity, energy, ancillary services, and renewable portfolio mandates. Only natural
GW
5 GW
10 GW
15 GW
20 GW
25 GW
30 GW
35 GW
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
Oil
Coal
Natural Gas
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
gas-fired peaking and CCCT plants, solar plants, and wind plants are in the plan. The
IRP does not include new nuclear or coal plants over the forecast horizon.

Many states have RPS requirements promoting renewable generation to reduce
greenhouse gas emissions, provide jobs, and diversify their 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 cannot model RPS levels explicitly.
Instead, Avista inputs RPS requirements into the model at levels sufficient to satisfy
state laws based on resource selection trends. Figure 10.4 illustrates new capacity and
RPS additions made in the modeling process. Nearly 112 GW will be required to meet
the renewable and capacity requirements for the system. Wind and solar facilities meet
most renewable energy requirements.

Geothermal, biomass, and hydroelectric resources provide limited RPS contributions.
Due to its low capacity factor, large quantities of solar capacity are necessary to make a
meaningful contribution. Renewable resource choices differ depending on state laws
and the local availability of renewable resources. For example, the Southwest will meet
RPS requirements with solar given policy choices by those states. The Northwest will
use a combination of wind, solar, and hydroelectric upgrades because the costs of
these resources are the lowest for the region. Rocky Mountain States will meet RPS
requirements predominately with wind.

Figure 10.4: Cumulative Generation Resource Additions (Nameplate Capacity)

In total, 45,000 MW of new utility and consumer-owned renewable generation will put
downward pressure on afternoon peak pricing and move peak load requirements later in
the day. Potential for oversupply in shoulder months in California will increase imports to
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
16 GW
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
To
t
a
l
Cu
m
u
l
a
t
i
v
e

Storage Biomass
Wind Geothermal
Hydro Net Meter
Solar Natural Gas
Cumulative
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
the Northwest and other markets. The forecast finds wind generation is no longer the
largest contributor of new renewable resources in the Western Interconnect; it
represents 6,000 MW, or 13 percent, of new renewable capacity. The largest resource
addition expected in the west is natural gas-fired generation. The technology likely will
be a combination of peakers and flexible combined cycle plants. A new entrant into the
resource forecast is storage technology. Given increasing government intervention in
the energy storage market in California, 1,300 MW of storage capacity is included in the
forecast. 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 2021. Utility resource size
requirements determine if the new plants are CCCTs or peakers. Based on market
simulation results, a 24 percent regional planning margin (including operating reserves)
is necessary to meet the 5 percent LOLP. The Northwest likely will continue to develop
wind to meet RPS requirements, but given the lower cost of solar, Avista expects some
utilities to move to solar to meet renewable requirements beginning in 2020. Table 10.2
shows the amount of new renewables added to the Northwest by the end of 2035 in the
Expected Case.

Table 10.2: Added Northwest Generation Resources

Resource Type Capacity (MW)
Wind 2,340
Utility- Solar 1,140
Customer- Solar 1,884
Other Renewables 225

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 limited siting opportunities.

Natural Gas
The natural gas industry continues its fundamental shift away from conventional gas to
hydraulic fracturing, or fracking. As fracking continues to become more efficient,
production increases at record pace. At the same time, growth in the residential,
commercial, and industrial markets is flat. Natural gas used for power generation is
growing due to its flexibility to support the variable output from renewable energy and as
a replacement resource for coal plant retirements caused by state and federal
regulations. Additionally, forecast adoption of natural gas for transportation and LNG
exports increases demand in later years of the forecast.

Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
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. Unconventional sources
reduce overall price levels and volatility, although it is unknown how much volatility will
exist in the future, as technology plays out against regulatory pressures and the
potential for new demand created by falling prices. Avista uses forward market prices
and a combination of two forecasts from prominent energy industry consultants to
develop the natural gas price forecast for this IRP. Based on these forecasts, the
levelized nominal price is $5.13 per dekatherm (Dth) at Henry Hub (shown in Figure
10.5 as the green bars). The pricing methodology to create a fundamental price forecast
is below, as follows:

 2016: 100 percent market;
 2017: 75 percent market, 25 percent consultant average;
 2018: 50 percent market, 50 percent consultant average; and
 2019-21: 25 percent market, 75 percent consultant average.

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. One change in recent years is the new
Ruby pipeline. It provides the west coast access to historically cheaper natural gas
supplies located in the Rocky Mountains. 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
$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
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
IRP Forecast
Consultant 1
Consultant 2
Forwards (12/04/14)
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
natural gas prices, monthly Stanfield price shapes are in Table 10.4 for selected
forecast years.

Table 10.3: Natural Gas Price Basin Differentials from Henry Hub

Stanfield 93% 94% 95% 97% 100%
Malin 98% 98% 98% 99% 101%
Sumas 90% 93% 93% 97% 100%
AECO 81% 83% 87% 92% 94%
Rockies 97% 96% 97% 98% 99%
Southern CA 103% 102% 102% 102% 103%

Table 10.4: Monthly Price Differentials for Stanfield from Henry Hub

Jan 97% 97% 98% 99% 103%
Feb 97% 96% 97% 98% 102%
Mar 96% 95% 96% 98% 101%
Apr 92% 94% 95% 96% 100%
May 91% 92% 93% 95% 99%
Jun 87% 88% 92% 94% 98%
Jul 87% 90% 93% 93% 98%
Aug 91% 93% 94% 95% 99%
Sep 93% 95% 95% 97% 100%
Oct 93% 95% 96% 98% 100%
Nov 95% 97% 97% 100% 102%
Dec 96% 97% 96% 99% 102%

Coal
This IRP models no new coal plants in the Western Interconnect, so coal price forecasts
affect only existing facilities. The average annual price increase over the IRP timeframe
is 3.6 percent based on data from the Energy Information Administration. For Colstrip
Units 3 and 4, Avista used escalation rates based on expectations from existing
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. This characteristic is valuable for meeting peak load, following general intra-day
load trends, shaping energy for sale during higher-valued hours, and integrating
variable 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
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
spanning 1928 to 2009. This IRP also includes BPA hydroelectric estimates for the 80-
year record for British Columbia and California.

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 17,370 aMW in 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 13,735 aMW (-21 percent) and a 90th
percentile water year of 20,340 aMW (+17 percent).

AURORAXMP maps each hydroelectric plant to a load zone, creating a similar energy
shape for all plants in that load zone. For Avista’s hydroelectric plants, AURORAXMP
uses the output from its own proprietary software with a better 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 move hydroelectric generation into peak load hours; this maximizes the
value of the system consistent with actual operations.

Figure 10.6: Northwest Expected Energy

0%
2%
4%
6%
8%
10%
12%
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
22
,
0
0
0
22
,
5
0
0
23
,
0
0
0
23
,
5
0
0
Pe
r
c
e
n
t

o
f

H
y
d
r
o

D
r
a
w
s
Average Megawatts
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Wind
New wind resources satisfy 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 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 risk for the electricity
industry because of its reliance on carbon-emitting power generation. Regulation may
require the reduction of carbon emissions at existing power plants, the construction of
low- and non-carbon-emitting technologies, and changes to existing resource
operations. Between 2008 and 2012, carbon emissions from electricity generation have
fallen by nearly 12 percent due to reduced loads and lower coal generation levels.

Future carbon emissions could fall due to fundamental market changes. In 2014, the
EPA released the draft CPP under section 111(d) of the CAA to reduce emissions from
existing plants. A description of the draft CPP is in Chapter 7 – Policy Considerations.
Use of compliance measures that do not rely on emission reductions solely from
covered fossil-fueled electric generating units, such as renewable energy and energy
efficiency standards, would not necessarily preclude emission increases from certain
sources, just an overall reduction in a statewide emission rate. If emissions from plants
covered under section 111(d) and newly constructed plants subject to section 111(b)
are not both subjected (at some point) to the same emission rate target established
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
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
under section 111(d), then newly constructed thermal facilities may increase emissions
even when complying with 111(b)’s emission performance standard.

The Expected Case makes assumptions about state and federal greenhouse gas
emissions policies. Avista’s 2013 IRP acknowledgement from the WUTC directed the
company to include a non-zero cost of carbon in the 2015 IRP. The acknowledgement
indicated that by not including a risk factor for this potential cost, the portfolio decision
does not include the potential risk of the added costs. The Expected Case in this IRP
includes a 10 percent probability of $12 per metric ton beginning in 2020. Beyond 2020,
the price increases 5 percent per year. This results in a levelized 2016-2035 cost of
$11.45 per metric ton, applied randomly in 10 percent of the modeled iterations.

The second carbon reduction assumption in the Expected Case is the Western
Interconnect meeting draft CPP goals by 2030. The CPP proposal was in draft form at
the time of IRP development. This regulation received the most comments on a
proposed rule in EPA history. The final rule, issued after the modeling was complete for
this IRP, differs from the draft. The IRP assumes meeting CPP state-by-state goals as a
whole in the Western Interconnect by 2030. The IRP assumes certain modifications to
the goals to conform to this modeling effort, including adjustments for plants located
outside the Western Interconnect, and adjusting Idaho’s goal to account for partial-year
operation of the Langley Gulch plant. The IRP assumes the Western Interconnect must
be below 801 pounds per MWh by 2030. Figure 10.8 shows adjusted state and regional
carbon intensity goals for CPP-regulated plants compared to the 2012 baseline.

Figure 10.8: 2030 Adjusted State Carbon Intensity CPP Goals

0
500
1,000
1,500
2,000
2,500
West AZ CA CO ID MT NM NV OR UT WA WY
EP
A

l
b
s

p
e
r

M
W
h
2030 Goal EPA Reduction
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
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 developed 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 among the most volatile of any traded commodity. Daily Stanfield
prices ranged between $1.72 and $24.36 per Dth between 2004 and 2014. Figure 10.9
shows average Stanfield monthly prices since January 2004. Prices retreated from 2008
highs to a monthly price of $2.26 per Dth in April 2015. Prices since 2009 are lower than
the previous five years, but continue to show volatility.

There are several methods to stochastically model natural gas prices. This study retains
the method from the 2011 IRP, with mean prices shown in Figure 10.5 as the starting
point. Prices vary using historical month-to-month volatility and a lognormal distribution.

Figure 10.9: 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
7/
1
/
2
0
0
4
1/
1
/
2
0
0
5
7/
1
/
2
0
0
5
1/
1
/
2
0
0
6
7/
1
/
2
0
0
6
1/
1
/
2
0
0
7
7/
1
/
2
0
0
7
1/
1
/
2
0
0
8
7/
1
/
2
0
0
8
1/
1
/
2
0
0
9
7/
1
/
2
0
0
9
1/
1
/
2
0
1
0
7/
1
/
2
0
1
0
1/
1
/
2
0
1
1
7/
1
/
2
0
1
1
1/
1
/
2
0
1
2
7/
1
/
2
0
1
2
1/
1
/
2
0
1
3
7/
1
/
2
0
1
3
1/
1
/
2
0
1
4
7/
1
/
2
0
1
4
1/
1
/
2
0
1
5
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.10 shows Stanfield natural gas price duration curves for 2016, 2025, and
2035. The chart illustrates a larger price range in the later years of the study, reflecting
less forecast certainty over time. Shorter-term prices are more certain due to additional
market information and the quantity of near term natural gas trading. Figure 10.11
shows another view of the forecast. The mean price in 2016 is $3.47 per Dth,
represented by the horizontal bar, and the levelized price over the 20 years is $4.97 per
MWh. The bottom and top of the bars represent the 10th and 90th percentiles. The bar
length indicates price uncertainty.

Figure 10.10: Stanfield Annual Average Natural Gas Price Distribution

0
100
200
300
400
500
$/
D
t
h
$5
/
D
t
h
$1
0
/
D
t
h
$1
5
/
D
t
h
$2
0
/
D
t
h
$2
5
/
D
t
h
$3
0
/
D
t
h
$3
5
/
D
t
h
It
e
r
a
t
i
o
n
2016 2025 2035
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.11: Stanfield Natural Gas Distributions
Regional Load Variation
Several factors drive load variability. The largest short-run driver is weather. Long-run
economic conditions like the recent 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 2013 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 2015 IRP. Statistics are relative
to the Northwest load area (Oregon, Washington, and Idaho). “NotSig” indicates that no
statistically valid correlation existed in the data. “Mix” indicates the relationship was not
consistent across the 2007 to 2013 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.1
1 The coefficient of determination is the standard deviation divided by the average.
$/Dth
$2/Dth
$4/Dth
$6/Dth
$8/Dth
$10/Dth
$12/Dth
$14/Dth
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
1
6
-
3
5
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Table 10.5: January through June Load Area Correlations

Area Jan Feb Mar Apr May Jun
Alberta Not Sig Not Sig Not Sig Mix Mix Mix
Arizona 14% 34% Mix Not Sig Mix 7%
Avista 89% 82% 81% 80% 43% 51%
British Columbia 87% 86% 72% 78% 50% 31%
California Not Sig Not Sig Mix Mix Mix 30%
CO-UT-WY -16% Mix Mix -24% -3% -6%
Montana 50% 43% 65% 57% Mix 7%
New Mexico Not Sig Mix Mix Mix Mix Not Sig
North Nevada 62% 22% 7% Not Sig Mix 25%
South Idaho 77% 75% 67% Mix Mix 32%
South Nevada 37% 59% Mix Not Sig Mix 7%

Table 10.6: July through December Load Area Correlations

Area Jul Aug Sep Oct Nov Dec
Alberta Not Sig Not Sig Not Sig Not Sig Not Sig Not Sig
Arizona Not Sig Not Sig Mix -7% Mix 8%
Avista 66% 75% 65% 77% 92% 92%
British Columbia 67% 47% 18% 80% 89% 84%
California 5% Not Sig Mix Not Sig Mix Not Sig
CO-UT-WY -9% Mix -2% -1% 19% Mix
Montana 14% 15% 8% 7% 76% 76%
New Mexico Not Sig Not Sig Mix -21% 36% Not Sig
North Nevada 48% 61% 32% Not Sig 75% 63%
South Idaho 40% 63% 32% Mix 86% 88%
South Nevada 7% 37% Mix -22% Mix 63%

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

Area Jan Feb Mar Apr May Jun
Alberta 5.4% 4.6% 5.2% 5.0% 5.5% 6.1%
Arizona 8.8% 8.3% 8.1% 12.3% 16.5% 18.6%
Avista 10.1% 8.8% 10.2% 9.8% 9.7% 11.1%
British Columbia 9.7% 8.7% 9.4% 9.3% 9.7% 9.9%
California 10.6% 10.5% 10.5% 10.8% 12.5% 14.2%
CO-UT-WY 8.6% 8.1% 8.6% 8.6% 10.0% 14.8%
Montana 8.5% 7.3% 8.0% 7.9% 8.2% 10.5%
New Mexico 9.4% 9.1% 9.3% 10.9% 14.5% 15.9%
Northern Nevada 6.3% 6.2% 6.3% 6.4% 7.6% 10.2%
Pacific Northwest 11.0% 9.8% 10.6% 10.1% 9.6% 9.9%
South Idaho 9.5% 8.6% 9.9% 10.5% 11.6% 16.3%
South Nevada 7.3% 6.6% 7.2% 12.5% 17.8% 20.1%

Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Table 10.8: Area Load Coefficient of Determination (Standard Deviation/Mean)

Area Jul Aug Sep Oct Nov Dec
Alberta 6.5% 6.2% 5.8% 5.6% 5.6% 5.5%
Arizona 16.4% 16.9% 18.1% 15.0% 8.7% 8.3%
Avista 13.9% 13.6% 11.5% 10.1% 11.1% 10.7%
British Columbia 10.8% 10.6% 10.3% 10.3% 11.3% 10.3%
California 14.9% 15.9% 16.0% 12.7% 11.2% 11.0%
CO-UT-WY 14.7% 14.3% 13.1% 9.5% 9.1% 9.3%
Montana 11.1% 10.9% 9.3% 8.4% 8.9% 9.0%
New Mexico 15.0% 14.7% 15.7% 12.2% 10.3% 10.0%
Northern Nevada 11.3% 10.9% 9.8% 6.8% 6.9% 7.3%
Pacific Northwest 11.8% 11.7% 10.8% 10.5% 12.0% 12.0%
South Idaho 12.2% 12.9% 13.5% 9.6% 10.4% 9.9%
South Nevada 17.9% 18.3% 20.0% 14.1% 7.8% 7.8%

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). There is
some debate in the Northwest over whether the hydroelectric record has year-to-year
correlation. Avista does not model year-to-year correlation after studying the data and
finding a modest 35 percent year-to-year correlation over the 80-year record.

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 is
varying. 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 it
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 it 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.

Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Avista believes an accurate representation of a wind shape across the West requires
meeting several conditions:

1. Data correlated between areas using historical data.
2. Data within load areas is auto-correlated.2
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,
attempts to adhere to 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, hour type (night or day), 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 four 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 2014 generation
recorded by BPA Transmission; in 2014, the average wind fleet in BPA’s balancing
authority had a 28.1 percent capacity factor.

2 Adjoining hours or groups of hours are correlated to each other.
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.12: Wind Model Output for the Northwest Region

Figure 10.13: 2014 Actual Wind Output BPA Balancing Authority3

There is speculation a correlation exists between wind and hydroelectric generation,
especially outside of the winter months where storm events bring both rain to the river
3 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
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
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
system and wind to the wind farms. This IRP does not correlate wind and hydroelectric
generation due to a lack of any historical wind data set large enough to test this
hypothesis. If correlation exists, it would be optimal to run the model using a large
dataset of historical wind and water years.

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 have a material impact on market prices and therefore 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.

Market Price Forecast
An optimal resource portfolio cannot ignore the extrinsic value inherent in its resource
choices. The 2015 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 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 individual
utilities in the region. This reflects the fact that Western regions can 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
the model does not itself select PRS resources. 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 analyses 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.

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
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
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 Western
Interconnect also has major trading hubs at the California/Oregon Border (COB), Four
Corners, in the northwestern corner of New Mexico, Palo Verde in central Arizona, SP-
15 in southern California, NP-15 in northern California, and Mead in southern Nevada.
The Mid-Columbia market is usually the lowest cost because of the significant amount
of hydroelectric generation assets at the hub, though other markets can be less
expensive when Rocky Mountain-area natural gas prices are low and natural gas-fired
generation is setting marginal power prices.

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 others. Each study simulates the entire
Western Interconnect hourly between 2016 and 2035. The analysis used 29 central
processing units (CPUs) linked to a SQL server, creating over 45 GB of data in 3,000
CPU-hours.

Figure 10.14 shows the Mid-Columbia stochastic market price results with horizontal
bars representing the 10th to 90th percentile range for annual prices, the diamonds show
the average prices, and the arrows represents the 95th percentile. The 20-year nominal
levelized price is $38.48 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 $7.78 per MWh over 20 years. The 2013 IRP annual average
nominal price was $44.08 per MWh. The reduction in pricing is a result of lower natural
gas prices, lower loads, and higher percentages of new lower-heat-rate natural gas
plants.

Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.14: Mid-Columbia Electric Price Forecast Range

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

2016 25.87 21.62 29.05
2017 27.27 23.03 30.47
2018 29.59 25.18 32.90
2019 31.40 26.83 34.82
2020 33.25 28.94 36.48
2021 34.54 30.21 37.79
2022 36.05 31.70 39.30
2023 36.43 32.17 39.64
2024 38.60 34.27 41.85
2025 39.42 35.18 42.59
2026 43.12 38.80 46.36
2027 44.72 40.23 48.08
2028 46.48 42.09 49.79
2029 48.01 43.51 51.39
2030 48.79 44.32 52.14
2031 51.23 46.52 54.76
2032 53.90 48.98 57.58
2033 54.98 49.95 58.74
2034 57.77 52.65 61.64
2035 59.33 54.12 63.24
$/MWh
$20/MWh
$40/MWh
$60/MWh
$80/MWh
$100/MWh
$120/MWh
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
1
6
-
3
5
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Greenhouse Gas Emission Levels
Greenhouse gas levels decline as natural gas prices decrease and coal plants react by
dispatching for fewer hours in the year or retire. This IRP includes a 10 percent
probability of a carbon price and includes reductions consistent with EPA’s CPP goal for
2030. This forecast also includes cap-and-trade costs in California and carbon taxes in
the Canadian provinces. Further discussion of carbon policy is in Chapter 7 – Policy
Considerations. Figure 10.15 shows historic and expected greenhouse gas emissions
for the Western Interconnect. Greenhouse gas emissions from electricity generation
decrease 6.4 percent between 2016 and 2035, and 2016 is 12 percent lower than 2012.
The figure also includes 10th and 90th percentile statistics from the 500-iteration dataset.
The higher and lower bands show where emissions could land depending on changes
in hydroelectric generation, load, resource availability, and other factors. The reduction
drivers are lower load forecasts, lower natural gas prices, higher RPS requirements in
some states, and forecasted coal-fired generation retirements due to federal and state
regulations, and carbon pricing. Further, emissions from plants covered under the CPP
fall by 28 percent as shown in the green line, but new plants emissions covered under
the CPP offset much of this reduction.

Figure 10.15: Western States Greenhouse Gas Emissions

Figure 10.16 illustrates the Expected Case emissions rate for EPA regulated plants
compared to EPA’s draft CPP goal for each year. The Expected Case estimates the
west will meet the 2030 goal by 2026; by 2035, the 681 lbs/MWh result is well below the
801 lbs/MWh CPP draft goal. Certain states, including Arizona, Colorado, Washington,
and Wyoming, likely will exceed the goal while other states witness falling emissions.
See Figure 10.17. If the final rule implements as the draft proposal, these state will need
to take additional action, as described later in this chapter.

0
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
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
111(d) plants
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.16: EPA’s CPP Annual Emissions Intensity for the West

Figure 10.17: EPA’s CPP 2030 State Goal vs. Modeling Result

0
200
400
600
800
1000
1200
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
EP
A

l
b
s

p
e
r

M
W
h
Emissions Rate
2030 Goal
801
702
537
1,108
324
1,763
1,048
647
372
1,322
215
1,666
681
926
311
1,112
289
1,693
688
542
324
1,236
236
1,711
0
500
1,000
1,500
2,000
West AZ CA CO ID MT NM NV OR UT WA WY
EP
A

l
b
s

p
e
r

M
W
h
2030 Goal
IRP Forecast
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Resource Dispatch
State-level RPS goals and greenhouse gas regulations change resource dispatch
decisions and affect future power prices. The Northwest already is witnessing the
market-changing effects of a more than 7,750 MW wind fleet. Figure 10.18 illustrates
how natural gas will increase its contribution as a percentage of Western Interconnect
generation, from 28 percent in 2016 to 42 percent 2035. The increase offsets coal-fired
generation, with coal dropping from 22 percent in 2016 to 10 percent in 2035. Utility-
owned solar and wind generation increase from 9 percent in 2016 to 14 percent by
2035. New renewable generation sources also reduce coal-fired generation, but natural
gas-fired generation is the primary resource meeting load growth.

Public policy changes encouraging renewable energy development may reduce
greenhouse gas emissions on a market scale, but they also change electricity
marketplace fundamentals. On the present trajectory, policy changes are likely to move
the generation fleet toward natural gas, with its currently low but historically volatile
prices. These policies will displace low-cost coal-fired generation with higher-cost
renewables and natural gas-fired generation having lower capacity factors (wind) and
higher marginal costs (natural gas). Stranded coal plant investments may increase the
overall cost of electricity. Further, wholesale prices likely will increase with the effects of
the changing resource dispatch driven by carbon emission limits and renewable
generation integration. New environmental policy-driven investments, combined with
higher market prices, will necessarily lead to higher than otherwise retail rates absent
greenhouse gas reduction policies.

Figure 10.18: Base Case Western Interconnect Resource Mix

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
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
Me
g
a
w
a
t
t

A
v
e
r
a
g
e
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Scenario Analysis
Scenario analysis evaluates the impact of changes in underlying market assumptions,
Avista’s generation portfolio and new generation resource options’ values. In addition to
the Expected Case, this IRP includes three stochastic analyses. The Benchmark Case
removes the carbon price and relaxes assumptions on meeting draft CPP goals. This
scenario provides data to calculate the impact of the environmental policies in the
Expected Case. The second scenario assumes all four Colstrip units retire by the end of
2026. This scenario uses a portfolio study to estimate impacts of an early closure at
Colstrip. The third scenario looks at the added costs and associated reductions in
greenhouse gas emissions if the social cost of carbon was included in the market price
analysis. Deterministic studies model impacts of state-by-state draft CPP compliance.

Benchmark Scenario
The Benchmark Scenario removes the carbon adder in 2020 and relaxes assumptions
in meeting the draft CPP targets. The flat levelized price for this scenario is $38.12 per
MWh, or a reduction of $0.39 per MWh from the Expected Case. Figure 10.19 shows
annual flat prices compared to the Expected Case. This scenario’s prices are similar to
the Expected Case. The levelized cost of the carbon adder in the Expected Case is
$1.15 per metric ton. While the emissions penalty was small in this case, Western
Interconnect emissions increase 2.3 percent by 2035. This scenario shows that the
lower emissions of the Expected Case are relatively modest, at a levelized $30 million
each year for the Western Interconnect. Figure 10.20 shows annual greenhouse gas
emissions for the Western Interconnect in the Benchmark Scenario.

Figure 10.19: Annual Mid-Columbia Flat Price Forecast Benchmark Scenario

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
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
Benchmark
Expected Case
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.20: Benchmark Scenario Annual Western U.S. Greenhouse Gas Emissions

No Colstrip Scenario
The No Colstrip Scenario models the implications of retiring Colstrip. 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 the site redevelops with several large CCCT plants upon retirement
in 2026. It does not attempt to represent the feasibility of this assumption, but rather
helps understand the impacts to the overall market place by replacing Colstrip with a
CCCT. Without Colstrip, regional market prices increase slightly as shown in Figure
10.21. There are small differences beginning in 2027 with a $0.93 per MWh annual
average price difference. 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. Further, without replacement capacity, price impacts and reliability
concerns increase. Beginning in 2027, the annual cost to all western customers
increases by $651 million with the closure of Colstrip, or 2.6 percent, in the No Colstrip
scenario. Without Colstrip, greenhouse gas emissions should decrease; in 2035
emissions in this scenario were 3.2 percent lower, or nearly 9.3 million metric tons per
year, as shown in Figure 10.22. Given the increased cost and associated emissions
reductions, the implied price of carbon reduction at Colstrip is $74.17 per metric ton in
2027; the average price between 2027 and 2035 is $73.18 per metric ton.

-
50
100
150
200
250
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
Mi
l
l
i
o
n
s

o
f

M
e
t
r
i
c

T
o
n
s
Benchmark
Expected Case
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.21: Annual Mid-Columbia Flat Price Forecast Colstrip Retires Scenario

Figure 10.22: No Colstrip Scenario Annual Western U.S. Greenhouse Gas Emissions

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
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
Colstrip Retires
Expected Case
-
50
100
150
200
250
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
Mi
l
l
i
o
n
s

o
f

M
e
t
r
i
c

T
o
n
s
Colstrip Retires
Expected Case
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Social Cost of Carbon Scenario
For the past several IRPs Avista has conducted carbon emission pricing scenarios. For
this IRP, the TAC recommended a Social Cost of Carbon case. The Social Cost of
Carbon study uses data from an EPA study. The prices from this study have different
ranges depending on the discount rate assumed and the point on the probability curve.
Avista chose the 5 percent discount rate study with a starting price of $11 per metric ton
in 2010 (2007 dollars). Figure 10.23 shows the nominal prices per metric ton. The
levelized price is $19.31 per metric ton, approximately 18 times the carbon cost
assumed in the Expected Case. These prices do not vary in each of the 500 iterations.

With a Social Cost of Carbon adder, the impact to Mid-Columbia prices is more
apparent. The levelized price increases to $45.46 per MWh, or 18 percent higher than
the Expected Case, as shown in Figure 10.24. The added pricing to emissions also
increases power costs by $3.6 billion annually (17.2 percent) across the U.S. west. In
exchange for the added costs, emissions fall 9.6 percent or 25 million metric tons by
2035. See Figure 10.25.

Figure 10.23: Social Cost of Carbon Scenario Emission Prices

$0
$5
$10
$15
$20
$25
$30
$35
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
$
p
e
r

M
e
t
r
i
c

T
o
n
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.24: Annual Mid-Columbia Flat Price Forecast Social Cost of Carbon Scenario

Figure 10.25: Social Cost of Carbon Scenario Western US Greenhouse Gas Emissions

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
$80/MWh
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
Social Cost of Carbon
Expected Case
-
50
100
150
200
250
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
Mi
l
l
i
o
n
s

o
f

M
e
t
r
i
c

T
o
n
s
Social Cost of Carbon
Expected Case
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Clean Power Plan Scenarios
The 2015 IRP analyzes implications of the draft CPP by first looking at its requirements
on a state-by-state compliance basis instead of a regional basis as is assumed in the
Expected Case. This scenario was studied on a deterministic basis rather than the full
500 stochastic iterations, because some of the stochastic variables have a large impact
on emissions. Because emissions are highly dependent on some of the stochastic
assumptions – for example, streamflows affect hydroelectric generation – a low water
year is tested. To meet the 2020 draft CPP goal, each state would have to change its
system. Any planned coal retirement beyond 2020 would accelerate to 2019. Some
states would need to increase conservation and renewable resource acquisitions. Many
states may need to implement a carbon emissions price. Northwest states would require
a carbon price of $1.25 per short ton in an average water year to reduce emissions,
even with the early closure of Centralia 1 & 2 and Boardman by the end of 2019. Other
states, such as Colorado and Arizona, would require prices near $20 per short ton.

Figure 10.26 shows the Mid-Columbia flat annual price in the state-by-state compliance
scenario. The levelized price is $39.06 per MWh, 1.6 percent higher than the Expected
Case’s deterministic study. This is not a large increase because the average price of
carbon across the west is actually lower than in the Expected Case, but since fewer
coal resources are available, the price is higher. In 2020, the year with the largest price
change, the difference is $1.59 per MWh, an increase of 4.7 percent.

Figure 10.26: Draft CPP as Proposed Scenario Flat Mid-Columbia Electric Prices

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
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
111(d)
Expected Case (Deterministic)
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
The cost of the Western electrical system requires review to understand the impacts of
a state-by-state draft CPP scenario. Levelized cost increases $342 million per year (1.7
percent), and cost is $1.2 billion (7.6 percent) in 2020. This added cost reduces
emissions from the Expected Case by 19 percent in 2020 and 9 percent in 2035, as
shown in Figure 10.27. The reduction from the Expected Case comes from earlier
retirement of coal resources. Reductions toward the end of the study are from additional
renewable resources and higher carbon emission prices. Emissions increase because
increased conservation offsets the need to reduce emissions from generation.

The draft CPP significantly affects the timing of new resources to replace retired coal
plants. It would require carbon pricing unless using other CPP building blocks. These
issues are minimal compared to a low water year in the Northwest. In low water years,
decreased hydroelectric production requires the region’s natural gas and coal-fired
resources to dispatch more and reduces regional exports and associated revenues.

Figure 10.27: Draft CPP as Proposed Scenario Western Greenhouse Gas Emissions

A low water year environment requires higher carbon prices to reduce emissions
compared to the average water year. To test this hypothesis, this study uses the water
conditions from 1941 to represent a lower 10th percentile water year. In this case, the
carbon prices required for the Northwest states are:

 Washington: $18/ton (2020), $18/ton (2030)
 Oregon: $19/ton (2020), $15/ton (2030)
 Idaho: $23/ton (2020), $14/ton (2030)
-
50
100
150
200
250
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
Mi
l
l
i
o
n
s

o
f

M
e
t
r
i
c

T
o
n
s
111d Proposed
Expected Case
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
The required carbon price changes as the amount of conservation increases, lowering
the reliance on the remaining generating fleet to meet the draft CPP goal. The cost
impact of this regulation in a lower water year can also be very high if the water
conditions are less than average beginning in 2020. For example, Figure 10.28
demonstrates the financial impact of the low water year; in 2020, the costs are $1.6
billion higher, or 9 percent, as compared to a low water year without the draft CPP
requirement. In 2030, as conservation ramps up and if a poor water year occurs, the
added costs decrease to $137 million or 0.4 percent higher. Electricity market prices at
the Mid-Columbia also have similar impacts. Figure 10.29 illustrates the increases of the
draft CPP in the 1941 water year and illustrates increases in prices compared to the
average water year from the Expected Case. In 2020, the added regulation increases
prices by $6.10 per MWh, or 17 percent, compared to the case without the regulation in
the poor water year. The impact decreases to approximately 5 percent in 2035. Given
that the future timing of low water years is unknown, the levelized price impact of $4.76
per MWh (10.3 percent) is the best indicator of the added price to the Northwest market.

Figure 10.28: Draft CPP as Proposed Scenario 1941 Water Year Annual Costs

$1.6 $1.3 $1.2 $1.1 $1.1 $0.9 $0.6 $0.6 $0.3 $0.3 $0.3 $0.2 $0.1 $0.1 $0.1 $0.1
$0
$5
$10
$15
$20
$25
$30
$35
$40
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
We
s
t
e
r
n

S
t
a
t
e
s

C
o
s
t

(
B
i
l
l
i
o
n
s
)
Expected Case Water Year 1941
111(d) as Proposed Water Year 1941
Cost Difference
Chapter 10- Market Analysis
Avista Corp 2015 Electric IRP
Figure 10.29: CPP as Proposed 1941 Water Year Scenario Mid-Columbia Electric Prices

$/MWh
$10/MWh
$20/MWh
$30/MWh
$40/MWh
$50/MWh
$60/MWh
$70/MWh
$80/MWh
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
Expected Case (Deterministic)
Expected Case (1941)
111d Proposal (1941)

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-1

11. Preferred Resource Strategy

Introduction
This chapter describes potential costs and financial risks of the new resource and
conservation strategy Avista plans to meet future requirements over the next 20 years.
It explains the decision making process used to select the PRS, and the resulting
avoided costs used to target future conservation.

The 2015 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 2013 plan include modestly less energy efficiency, the elimination of
demand response, and the elimination of a natural gas-fired peaking plant. The plan
also calls for upgrades to Avista’s thermal generating fleet. The strategy’s lower energy
efficiency acquisition is due to lower market prices and increased codes and standards
reducing some of the need for utility-sponsored acquisition. The reduction in natural
gas-fired resources results primarily from a lower retail load forecast. Demand response
is no longer in the PRS, as a third-party study found costs to be much higher than
estimated in the 2013 IRP. Like the prior plan, upgrades at certain existing facilities look
attractive as a resource alternative. Overall, the 2015 PRS performs better against the
efficient frontier than the 2013 strategy.

Supply-Side Resource Acquisitions
Avista began its shift away from coal-fired resources with the sale of its 210 MW share
of the Centralia coal plant in 2000. Natural gas-fired plants replaced it. See Figure 11.1.
Since the Centralia sale, Avista has made several generation acquisitions and
upgrades, including:

 25 MW Boulder Park natural gas-fired reciprocating engines (2002);
 7 MW Kettle Falls natural gas-fired CT (2002);
 35 MW Stateline wind power purchase agreement (2004 – 2014);
 56 MW (total) hydroelectric upgrades (through 2012);
 270 MW natural gas-fired Lancaster Generation Station tolling agreement
(2010 – 2026);
 105 MW Palouse Wind power purchase agreement (2012 – 2042); and
 16 MW Nine Mile Falls Upgrade (2016)
Section Highlights




Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-2

Figure 11.1: Resource Acquisition History

Resource Deficiencies
Avista uses a single-hour and an 18-hour peak event methodology to measure resource
adequacy. The 18-hour methodology assures energy-limited hydroelectric resources
can meet multiday extreme weather events.

Avista considers the regional power surplus in its planning, consistent with the NPCC’s
forecast, and does not intend to acquire long-term generation assets while the region is
significantly surplus. Current NPCC research indicates the region is long on capacity
through 2020 during the winter and forecasts no summer resource deficits.

Avista’s peak planning methodology includes operating reserves, regulation, load
following, wind integration, and a planning margin. Even with this planning methodology,
Avista currently projects having adequate resources between owned and contractually
controlled generation to meet physical energy and capacity needs until 2021.1 See
Figure 11.2 for Avista’s physical resource positions for annual energy, summer capacity,
and winter capacity. This figure accounts for the effects of new energy efficiency
programs on the load forecast. Absent energy efficiency, Avista would be deficient
earlier.

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
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
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-3

Figure 11.2: Physical Resource Positions (Includes Energy Efficiency)

Renewable Portfolio Standards
Washington voters approved the EIA in the November 2006 general election. The EIA
requires utilities with over 25,000 customers to meet 3 percent of retail load from
qualified renewable resources by 2012, 9 percent by 2016, 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, the Kettle Falls Generating Station and selective REC purchases.2 Table 11.1
provides a list of the qualifying generation projects and the associated expected output.
Figure 11.3 shows the forecast REC positions. The flexibility included in 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.

2 The RECs from Wanapum are not in WREGIS and are currently ineligible under the EIA requirements
for investor-owned utilities, but Avista is working with Grant County PUD to qualify the energy.

-600
-500
-400
-300
-200
-100
0
100
200
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
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 18 Hour Peak (MW)
Annual Energy (aMW)
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-4

Table 11.1: Qualifying Washington EIA Resources

Kettle Falls GS3 Biomass 1983 47.0 374,824 281,118
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 17.0 45,808 45,808
Cabinet Gorge 2 Hydro 2004 17.0 29,008 29,008
Cabinet Gorge 4 Hydro 2007 9.0 20,517 20,517
Wanapum Hydro 2008 0.0 22,206 22,206
Noxon Rapids 1 Hydro 2009 7.0 21,435 21,435
Noxon Rapids 2 Hydro 2010 7.0 7,709 7,709
Noxon Rapids 3 Hydro 2011 7.0 14,529 14,529
Noxon Rapids 4 Hydro 2012 7.0 12,024 12,024
Palouse Wind Wind 2012 105.0 349,726 419,671
Nine Mile 1 & 2 Hydro 2016 4.0 11,826 11,826

Figure 11.3: REC Requirements vs. Qualifying RECs for Washington State EIA

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
3 The Kettle Falls Generation Station becomes EIA qualified beginning in 2016. Clarification about old
growth fuel is required to determine the amount of energy to qualify for the law.

-
20
40
60
80
100
120
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
Av
e
r
a
g
e

M
e
g
a
w
a
t
t
s
Banking
Kettle Falls
Palouse Wind
Hydro Upgrades
Requirement
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-5

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 the 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, RPS, and
other requirements.

PRiSM
Avista staff developed the first version of PRiSM in 2002 to support resource decision
making in the 2003 IRP. Various 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.

Overview of the PRiSM model
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
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 The level of market reliance (surplus/deficit limits on energy, capacity and
RPS)
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. 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.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-6

Equation 11.1: PRiSM Objective Function

Minimize: (X1 * NPV2016-2040) + (X2 * NPV2016-2065)

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.4

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 risk Avista and its customers are willing to accept. No best point
on the curve exists, but Avista prefers points where small incremental cost additions
offer large 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 limit utility-scale resource options. Portfolios to the right of the curve
are less efficient as they have higher costs than a portfolio with the same level of risk.
The model does not meet deficits with market purchases or allow the construction of
resources in any incremental size.5 Instead, it uses the market to balance short-term
gaps and adds resources in sizes equal to the project sizes Avista could actually obtain.
4 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. 5 Market reliance, as identified in Section 2, is determined prior to PRiSM’s optimization.
Ri
s
k
Cost
Least Cost
Least Risk
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-7

Constraints
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 societal. The major resource constraints are capacity and
energy needs, Washington’s EIA, and greenhouse gas emissions performance
standard.

The PRiSM model selects from conservation, combined- and simple-cycle natural gas-
fired combustion turbines, natural gas-fired reciprocating engines, wind, solar, storage
batteries, and upgrades to existing thermal and hydroelectric resources.

Before the addition of an RPS obligation, the efficient frontier contained a least-cost
strategy on one axis, the least-risk strategy on the other axis, and all of the points in
between. Management used the efficient frontier to help determine where they wanted
to be on the cost-risk continuum. The least-cost strategy consists of natural gas-fired
peaking resources. Portfolios with less risk replace some of the natural gas-fired
peaking resources with wind generation, other renewables, combined cycle natural gas-
fired plants and/or coal-fired resources. Past IRPs identified resource strategies
including all of these risk-reducing resources. Added environmental and legislative
constraints reduce the number of resource choices available to reduce future costs
and/or risks.

Preferred Resource Strategy
The 2015 PRS consists of existing thermal resource upgrades, energy efficiency,
natural gas-fired peakers, and a natural gas-fired CCCT. A list of planned acquisitions is
in Table 11.2 and a graphic is in Figure 11.5. The first resource acquisition is 96 MW of
natural gas-fired peaking technology by the end of 2020. This resource acquisition fills
the capacity deficit created by the expiration of the 82-MW WNP-3 contract with the
BPA, the expiration of a 28 MW Douglas County PUD contract for a portion of its Wells
hydroelectric facility, and load growth. In this IRP evaluation, frame technology CTs are
the preferred gas-fired peaking technology. Given the relatively small cost differences
between the evaluated natural gas-fired peaker technologies, the future technology
decision will be determined in an RFP. Technological changes in efficiency and
flexibility may mean the Avista will need to revisit this resource choice closer to the
actual need. Since the long-term need is more than five years out, Avista will not
release an RFP in the next two years, but will begin a process to evaluate technologies
and potential locations prior to a RFP release, likely following the 2017 IRP.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-8

Table 11.2: 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 Efficiency 2016-2035 193 132
Distribution Efficiencies <1 <1
Total 193 132

Figure 11.5: New Resources Meets Winter Peak Loads

The next resource acquisitions in the PRS are upgrades to Avista’s thermal fleet. These
upgrades may be cost effective earlier depending upon negotiations with vendors. The
proposed 286 MW CCCT replaces the Lancaster tolling agreement expiring in October
2026. Avista could renegotiate the current agreement or find other mutual terms to
retain the plant for customers. If Avista does not retain Lancaster, it would need to build
or procure a similar-sized natural gas-fired unit. The new plant size could meet future
-
500
1,000
1,500
2,000
2,500
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
Me
g
a
w
a
t
t
s
Thermal Plant Upgrade NG Peaker
NG Combined Cycle CT Existing Resources
Load w/ Conservation + Cont.
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-9

load growth needs and delay or eliminate the need for later additional resource
acquisitions in this plan. Due to the uncertainty surrounding replacing Lancaster, this
IRP assumes the replacement is a new facility of similar size. More information and
replacement costs will be discussed in future IRPs as 2026 approaches.

The 2015 PRS is moderately different from the 2013 resource strategy shown in Table
11.3. Avista’s capacity needs have changed since the prior plan. The first need for new
resources has moved out one year, as Avista won an auction to purchase a share of the
output from Chelan County PUD’s hydroelectric projects. Lower loads compared to the
prior plan and new upgrade options eliminate the need for one of the peakers
forecasted in the prior plan.

Table 11.3: 2013 Preferred Resource Strategy

Resource By the End of
Year
ISO Conditions
(MW)
Winter Peak
(MW)
Energy
(aMW)
Simple Cycle CT 2019 83 86 76
Simple Cycle CT 2023 83 86 76
Combined Cycle CT 2026 270 281 248
Simple Cycle CT 2023 83 86 76
Rathdrum CT Upgrade 2028 6 2 5
Simple Cycle CT 2032 50 52 46
Total 575 594 527
Efficiency
Improvements
Acquisition
Range
Winter Peak
Reduction
Energy
(aMW)
Energy Efficiency 2014-2033 221 164
Demand Response 2022-2027 19 0
Distribution Efficiencies 2014-2017 <1 <1
Total 240 164

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 energy efficiency at below 110 percent
of generation alternative costs. Avista now models energy efficiency and supply side
options in a single optimization, a change from prior practice. This enhancement allows
PRiSM to select different conservation amounts along the efficient frontier instead of
one acquisition strategy across the entire curve.

Figure 11.6 shows the annual PRS conservation additions from the optimization
compared to the third party CPA. The PRiSM model selected nearly identical
conservation quantities each year and in total (132.5 aMW with PRiSM versus 132.1
with the CPA). Figure 11.7 shows the difference between the load forecast with and
without conservation. The 132 aMW of energy savings (including losses) represents 52
percent of potential load growth. Please refer to Chapter 5 – Energy Efficiency and
Demand Response for a detailed discussion of energy efficiency resources. That
chapter identifies 124.5 aMW, which is the 132 aMW minus 6 percent for losses.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-10

Figure 11.6: Energy Efficiency Annual Expected Acquisition Comparison6

Figure 11.7: Load Forecast with and without Energy Efficiency

6 Figure 11.6 includes 6.1 percent energy losses.

0
15
30
45
60
75
90
105
120
135
150
0
1
2
3
4
5
6
7
8
9
10
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
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
In
c
r
e
m
e
n
t
a
l

A
v
e
r
a
g
e

M
e
g
a
w
a
t
t
s
PRiSM
CPA
Cumulative PRiSM
-
200
400
600
800
1,000
1,200
1,400
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
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
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-11

Grid Modernization
Distribution feeder upgrades entered the PRS for the first time in the 2009 IRP. The grid
modernization process began with the Ninth and Central feeder in Spokane. The
decision to rebuild a feeder considers energy, operation and maintenance savings, the
age of installed equipment, reliability indexes, 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 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. Avista’s service territory has areas
with different combinations of benefits and costs for gas-fired generation. Locations in
Washington have higher generation costs because of natural gas fuel taxes and carbon
mitigation fees. However, Washington locations may benefit from their proximity to
natural gas pipelines and Avista’s transmission system, lower project elevations with
higher on-peak capacity contributions per investment dollar, and potential for water
rights to cool the facility more efficiently relative to air-cooled options. In Idaho, lower
taxes and fees decrease the cost of a potential facility, but fewer locations exist to site a
facility near natural gas pipelines, fewer low cost transmission interconnection points
are available, and fewer sites have available rights for cooling water. A 2013 IRP Action
Item was identification of a location for a future natural gas resource. Avista has studied
potential locations and concluded a site in Northern Idaho best fits customer needs.
Avista has yet to determine if a brownfield or a greenfield site is best. Given Avista’s
extended surplus position until the end of 2020, it will defer the decision while continuing
to pursue and evaluate sites.

Avista is not specifying a preferred peaking technology until a competitive bidding
process is completed. Given current assumptions, the resource strategy would include a
Frame CT machine. 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 has 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 low heat rates, and are highly flexible, but they
can be more expensive than other technologies. Given the expected number of
operating hours, the lowest cost option is the less efficient and less flexible Frame CT.
Increased flexibility requirements and greenhouse gas emissions costs could make a
hybrid plant or reciprocating engines preferable. Avista has enough resource flexibility
to meet customer needs to drive the strategy towards a lower cost peaker option, but
energy imbalance markets may provide enough revenues for a flexible peaker to offset
the higher costs.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-12

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 due to current or proposed environmental
regulations. The only significant carbon emitting lost resource is the expiration of the
Lancaster PPA in 2026. Figure 11.8 presents Avista’s expected greenhouse gas
emissions (excluding Kettle Falls Generating Station) with the addition of 2015 PRS
resources. Emissions should not change significantly prior to 2019 other than from year-
to-year fluctuations resulting from maintenance outages, market fluctuations, and
regional hydroelectric generation levels. Beginning in 2019, additional emissions will
come from new peaking resources, but these resources will not affect overall emissions
levels much due to low projected use. The estimates in Figure 11.8 do not include
emissions from purchased power or a reduction in emissions for off-system sales.
Avista expects its greenhouse gas emissions intensity from owned and controlled
generation to remain around 0.27 metric tons per MWh with the current resource mix
and the new generation identified in the PRS.

Figure 11.8: Avista Owned and Controlled Resource’s Greenhouse Gas Emissions

Capital Spending Requirements
This IRP assumes Avista will finance and own all new resources. This may or may not
be the result of competitive acquisition processes, but the overall result is unchanged by
assumed ownership structure. Using this assumption, and the resources identified in the
2015 PRS, the first capital addition to rate base is in 2021 for the first natural gas-fired
peaker. The development is likely to begin years earlier, but would likely enter rate base
-
0.13
0.25
0.38
0.50
Mil
1 Mil
2 Mil
3 Mil
4 Mil
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
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
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-13

January 1, 2021. Avista may begin making major capital investments for the addition in
2018 or earlier. The capital cash flows in Table 11.4 include AFUDC, transmission
investments for generation, tax incentives, and sales taxes. Over the 20-year IRP
timeframe, $682 million (nominal) in generation and related transmission expenditure is
required to support the PRS. A separate tariff rider funds energy efficiency.

Table 11.4: PRS Rate Base Additions from Capital Expenditures
(Millions of Dollars)

Year Investment Year Investment
2016 0.0 2026 8.2
2017 0.0 2027 398.9
2018 0.0 2028 98.7
2019 0.0 2029 0.0
2020 0.0 2030 0.0
2021 89.4 2031 0.0
2022 0.0 2032 0.0
2023 0.0 2033 0.0
2024 3.0 2034 4.2
2025 12.1 2035 68.1
2016-25 Total 104.5 2026-35 Totals 578.0

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 11.9 shows expected PRS costs through 2035 as the blue bar. In 2016,
costs are $26 per MWh. The chart shows a two-sigma cost range. Yellow diamonds
represent the lower range and orange triangles represent the upper range. The main
driver increasing power supply costs and volatility in future years is natural gas prices
and weather, which affects both hydroelectric generation levels and load variability.
Avista increases the volatility assumption of future natural gas prices, as the commodity
price has unknown future risks and a history of volatility.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-14

Figure 11.9: Power Supply Expense Range

Near Term Load and Resource Balance
Under Washington regulation (WAC 480-107-15), utilities expecting supply deficits
within three years of an IRP filing must file a RFP with the WUTC within 135 days after
filing the IRP. After WUTC approval, bids to meet the anticipated capacity shortfall are
issued within 30 days. In the 2013 IRP, an Action Item committed Avista to develop a
short-term capacity load and resource balance tool to monitor temporarily short
positions. Shortly after the filing of the 2013 plan, a Capacity Report was completed and
is consulted prior to the heating and cooling seasons. Chapter 6 – Long-term Position
discussed small deficits in 2015 and 2016. The company’s power supply department
filled those deficits due to monitoring of the Capacity Report. Table 11.5 shows the
latest position with the 2016 short-term capacity positions closed with market
purchases. In Table 11.6, the summer position is long in each of the next four years. As
described in Chapter 6, the region is long on summer capacity. Given this circumstance,
Avista is not planning to hold capacity for a planning margin and will utilize the surplus
in the wholesale market to meet load in extreme weather conditions or extended plant
outages.

$0
$10
$20
$30
$40
$50
$60
$70
$80
$90
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
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
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-15

Table 11.5: Avista Medium-Term Winter Peak Hour Capacity Tabulation

2016/17 2017/18 2018/19 2019/20
Load Obligations 1,718 1,725 1,737 1,748
Other Firm Requirements 239 89 59 8
Reserves Planning 376 374 376 381
Total Obligations 2,333 2,188 2,172 2,137

Firm Power Purchases 206 164 162 31
Owned & Contracted Hydro 1,014 1,029 996 1,001
Thermal & Storage Resources 1,137 1,142 1,142 1,141
Wind (at Peak) 0 0 0 0
Total Resources 2,357 2,335 2,300 2,173

Net Position 24 147 128 36

Table 11.6: Avista Medium-Term Summer 18-Hour Sustained Peak Capacity Tabulation

2016 2017 2018 2019
Load Obligations 1,515 1,529 1,542 1,554
Other Firm Requirements 189 89 89 59
Reserves Planning 165 164 166 166
Total Obligations 1,869 1,782 1,797 1,779

Firm Power Purchases 68 68 51 49
Owned & Contracted Hydro 823 818 806 781
Thermal Resources 984 988 988 988
Wind (at Peak) 0 0 0 0
Total Resources 1,875 1,874 1,845 1,818

Net Position 6 92 48 39

Efficient Frontier Analysis
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 met with
sets of intermediate portfolios between the least risk and least cost options.
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-16

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 given
current information.

Figure 11.10 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,
operating costs, and fuel expense; the y-axis displays the standard deviation of power
supply costs in 2027. It is necessary to move far enough into the future so load growth
provides PRiSM the opportunity to make new resource decisions. The year 2027 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 is not choosing to pursue the absolute least cost strategy in this IRP, as it relies
exclusively on natural gas-fired peaking facilities. A peakers-only strategy would include
more market risk than exists in the present portfolio because the portfolio would trade
diversity of the Lancaster CCCT for another peaker. Selecting the appropriate point on
the efficient frontier is not solvable through a mathematical formula.

Figure 11.10: 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
2
7

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 2016-2040
Least Cost
Preferred Resource Strategy
Least Risk
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-17

In the WUTC’s 2013 IRP acknowledgement, the Commission asked Avista to evaluate
the value of risk mitigation among competing resource strategies and provide
justification for its selection of the PRS over other portfolios along the efficient frontier.
Avista investigated several methods of measuring the benefits and costs of each
portfolio along the efficient frontier. Economic theory indicates all points on the curve
are the best portfolio for a given level of risk. Academic research suggests users of
efficient frontiers develop indifference curves to overlay against the efficient frontier to
help select the appropriate portfolio strategy. After researching this concept, it is no
different from finding what level of risk reduction a manager is accepting for each level
of risk. Avista investigated two other analytical methodologies to evaluate each portfolio
along the efficient frontier: risk adjusted PVRR and point-to-point derivatives.

The first step calculates risk adjusted PVRR for each portfolio. This calculation is the net
present value of the future revenue requirements, plus the present value of taking each
of the future year’s tail risk, calculated by 5 percent of the 95th percentile’s increase in
costs. This methodology assumes the lowest NPV should yield the best strategy. Figure
11.11 shows the results of this study of the efficient frontier. The lowest cost scenario,
including tail risk, is the Least Cost portfolio. This Risk-Adjusted PVRR methodology
suggests the Least Cost strategy would be the best choice. Before making this decision,
Avista considered additional analyses, given that this strategy built 527 MW of 11,000
Btu/kWh heat rate peakers. The strategy increases exposure to a potentially volatile
power and natural gas market as compared to today’s portfolio.

Figure 11.11: 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
PR
S
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 Porfolios
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-18

To illustrate this risk and the benefits of the PRS, Avista employed a second method. It
calculates point-to-point derivatives by analyzing the slope of the change in cost relative
to the change in costs. In this case, a greater slope indicates increasing benefits for
trading off risk reduction for higher portfolio costs; a higher slope indicates a better
tradeoff between cost and risk. Figure 11.12 illustrates the results of this study. The
PRS selected by PRiSM falls between Portfolios 3 and 4, indicating its results are valid.
Avista prefers the PRS relative to Portfolio 4 because it includes more efficiency
upgrades to its generation assets and a CCCT technology more closely aligned with our
expiring Lancaster CCCT facility contract.

Figure 11.12: Risk Adjusted PVRR of Efficient Frontier Portfolios

Other Efficient Frontier Portfolios
In addition to the PRS, the efficient frontier contains 16 additional resource portfolios.
The lower cost and higher risk portfolios contain primarily natural gas peakers, as
portfolio risk decreases, CCCT capacity increases. The amount of conservation varies
in these portfolios as it lowers risk, and as it fills deficiency gaps depending on the
resource selection. 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 larger needs with conservation in order to meet the load requirement. This
discussion continues in Chapter 12 – Portfolio Scenarios.

Toward the middle of the efficient frontier, PRiSM favors wind and solar to reduce risk
as additional conservation resources become more expensive. The lower half of the
efficient frontier includes portfolios with large capacity surpluses and renewable
-
1.00
2.00
3.00
4.00
5.00
6.00
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
Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-19

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.

Table 11.7: Alternative Resource Strategies along the Efficient Frontier (MW)

Portfolio
NG
Peaker
NG
CCCT Wind Solar
Thermal
Upgrade
Energy
Efficiency
Least Cost 527 - - - 38 128
2 524 - - - 41 135
3 239 286 - - 38 128
PRS 239 286 - - 41 132
4 143 341 - - 38 138
5 189 341 50 10 41 139
6 140 341 100 20 41 143
7 189 341 200 - 38 141
8 140 341 250 20 41 142
9 186 341 300 70 38 141
10 186 341 400 30 38 141
11 140 341 450 80 38 144
12 140 341 500 150 41 142
13 186 341 500 290 38 143
14 93 627 500 270 38 140
15 93 627 500 480 38 141
Least Risk 186 683 500 600 23 144

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 energy efficiency is within PRiSM, the prior IRP method of calculating avoided
costs is no longer required; but estimating these values is helpful in selecting
conservation measures in future 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 an hourly energy price to ensure
matching between savings and value. If the savings were the same each hour of the
year, it would receive the flat energy price, but if it were only saving energy in on-peak
hours, it would receive a higher price. In addition to energy prices, the 10 percent Power
Act adder and the value of loss savings are included.7 Reducing customer loads saves
future distribution and transmission capital and O&M costs, and is included in the
7 The Power Act adder refers to one aspect of federal law enacted in 1980 along with the creation of the
Northwest Power and Conservation Council.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-20

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 commodity energy sales in
the wholesale marketplace.

Equation 11.2 describes the avoided costs to evaluate conservation measures. This
equation is slightly different from the 2013 IRP. In prior IRPs, the capacity value
received the 10 percent Power Act benefit. Now with energy efficiency included in the
PRiSM model, the 10 percent adder cannot be included in the linear program as it
would create a non-linear solution. This change is consistent with the NPCC’s
methodology.

Equation 11.2: Conservation Avoided Costs

{(E + (E * L) + DC) * (1 + P)} + PCR

Where:
E = Market energy price. The price calculated by AURORAXMP is $38.48
per MWh assuming a flat load shape.
PCR = New resource capacity savings for the PRS selection point is
estimated to be $102 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 6.1 percent
based on Avista’s estimated system average losses.
DC = Distribution capacity savings. This value is approximately $12.30 per
kW-Year

Determining the Avoided Cost of New Generation Options
Avoided costs change as market prices, loads, and resources change. Table 11.8
shows avoided costs derived from the 2015 PRS, but they will change as Avista’s loads
and resources change. The prices represent the value of energy from a project making
equal deliveries over the year in all hours. In this case, a new resource, such as a
PURPA qualifying project, would not qualify for capacity payments until 2021. This is
because Avista does not need capacity resources until then. The capacity payments
included are tilted and levelized, meaning the actual capacity costs are linear and
increasing each year rather than the PRS’s actual declining cost curve for capacity. This
is similar to typical pricing in the marketplace.

Chapter 11 – Preferred Resource Strategy

Avista Corp 2015 Electric IRP 11-21

Table 11.8: Updated Annual Avoided Costs ($/MWh)

Year Flat
Energy
$/MWh
On-Peak
Energy
$/MWh
Off-Peak
Energy
$/MWh
Capacity
$/kW-Yr
2016 25.87 29.05 21.62 0.00
2017 27.27 30.47 23.03 0.00
2018 29.59 32.90 25.18 0.00
2019 31.40 34.82 26.83 0.00
2020 33.25 36.48 28.94 0.00
2021 34.54 37.79 30.21 145.00
2022 36.05 39.30 31.70 148.32
2023 36.43 39.64 32.17 151.72
2024 38.60 41.85 34.27 155.19
2025 39.42 42.59 35.18 158.75
2026 43.12 46.36 38.80 162.38
2027 44.72 48.08 40.23 166.10
2028 46.48 49.79 42.09 169.90
2029 48.01 51.39 43.51 173.80
2030 48.79 52.14 44.32 177.78
2031 51.23 54.76 46.52 181.85
2032 53.90 57.58 48.98 186.01
2033 54.98 58.74 49.95 190.27
2034 57.77 61.64 52.65 194.63
2035 59.33 63.24 54.12 199.09

Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-1

12. Portfolio Scenarios

Introduction
The PRS is Avista’s strategy to meet future loads. In case the future is different from the
IRP forecast, the strategy needs to be flexible enough to benefit customers under the
new future. 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 load forecasts, determines the impact of unit sizing, and the
selection of portfolios to the right of the efficient frontier. This chapter also identifies the
capital cost tipping points for solar, storage, and demand response options.

Mixed Integer versus Linear Programming
PRiSM is a mixed integer model that meets utility power supply deficits over the IRP
timeframe from a pre-defined set of resource options. The integer model selects only
commercially available resources. For example, if Avista is short 45.3 MW, the integer
model cannot select a 45.3 MW resource. Rather it must choose among unit sizes
actually for sale in the marketplace. This methodology creates lumpy resource
additions, meaning that by selecting a commercially available resource capable of fully
meeting the deficit, Avista likely will have some level of surplus. Figure 12.1 shows the
impact of lumpy resource acquisitions on the efficient frontier relative to a linear solution
not requiring lumpy additions. In this case, costs in the integer model average 0.5
percent higher than were Avista able to purchase resources exactly matching its deficits
in a linear model. In addition to higher costs, resources mixes on the efficient frontier
change when choices must match actual resources available in the marketplace. The
resources selected across the efficient frontier under a linear programming model are in
Table 12.1. This methodology creates a smoother transition of peakers to CCCTs and
energy efficiency increases at a smoother rate than the more realistic integer-based
model.

Chapter Highlights
 Lower or higher future loads do not materially change the resources strategy.
 Colstrip remains a cost-effective and reliable source of power to meet future
customer loads.
 Without Colstrip in 2027, customer bills increase $58 million.
 A $19 per metric ton social cost of carbon scenario increases customers’ costs
by $67 million per year levelized.
 Tipping point analysis suggests utility scale solar costs would need to decline
48 percent to be included in the PRS.

Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-2

Figure 12.1: Linear versus Integer Efficient Frontier Difference

Table 12.1: Efficient Frontier with Linear Programming

Portfolio NG
Peaker
NG
CCCT Wind Solar Thermal
Upgrade
Hydro
Upgrade
Energy
Efficiency
Least Cost 500 - - - 41 - 130
2 367 129 - - 41 - 133
3 222 274 - - 41 - 133
4 79 414 - - 41 - 135
5 58 429 60 - 41 - 139
6 56 431 132 - 41 - 139
7 48 439 202 - 41 - 139
8 41 445 276 - 41 - 139
9 41 445 352 - 41 - 140
10 30 456 400 46 40 - 140
11 29 458 478 50 38 - 141
12 6 480 500 143 38 - 141
13 - 515 500 282 38 - 141
14 - 549 500 446 38 - 141
15 - 674 500 523 12 - 144
Least Risk - 855 500 600 12 57 147

$ Mil
$10 Mil
$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
2
7

P
o
w
e
r

S
u
p
p
l
y

C
o
s
t
2016-2040 Levelized Power Supply Cost
Mixed Integer Model
Linear Model
PRS
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-3

Load Forecast Scenarios
The PRS meets the Expected Case energy load growth of 0.6 percent and winter peak
demand growth of 0.68 percent over the next 20 years. Chapter 3 – Economic and Load
Forecast provides details about three alternative load forecasts. Table 12.2 summarizes
the alternative growth assumptions. The high and low load scenarios use different
population growth assumptions than the Expected Case. The Increased DG Solar
scenario uses the same economic growth rate as the Expected Case, but assumes 10
percent of residential customers install rooftop solar with up to a 6 kW system by 2040.

Table 12.2: Load Forecast Scenarios (2016-2035)

Scenario Energy
Growth (%)
Winter
Peak
Growth (%)
Summer
Peak
Growth (%)
Expected Case 0.6 0.7 0.8
High Load 0.8 0.9 1.1
Low Load 0.2 0.6 0.7
Increased DG Solar 0.4 0.7 0.6

Table 12.3 shows changes to the PRS for each load scenario. In the High Load
scenario, 97 MW of additional natural gas-fired peakers meet added load growth, while
the Low Load scenario reduces peakers by 46 MW. The changes between the High and
the Low Load scenarios are not significant because expiring contracts is more of a
driver of Avista’s resource needs than load growth.

Table 12.3: Resource Selection for Load Forecast Scenarios

Resource
Expected
Case's
PRS
High
Loads
Low
Loads
Increased
DG Solar
NG Peaker 239 335 192 239
NG Combined Cycle CT 286 286 286 286
Wind 0 0 0 0
Solar 0 0 0 0
Demand Response 0 0 0 0
Thermal Upgrades 41 41 41 41
Hydro Upgrades 0 0 0 0
Total 565 662 519 565

The Increased DG Solar scenario provides interesting results. In this scenario, where
customer-supplied generation increases during summer peak-load periods, the PRS
does not change. The winter peak load drives Avista’s resource acquisition needs, so
this scenario does not change the resource strategy, as DG solar does not produce
energy between the hours of 5:00 pm and 7:00 pm in the winter. This results in the
same resource build, but with lower retail energy sales.

Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-4

Load forecast changes can also come in the form of new large loads or the loss of an
existing large load. In both cases, the change will likely be short notice. Avista likely
would meet these events by utilizing the energy market.

Colstrip Retirement Scenarios
The 2013 IRP acknowledgement letter from the Washington Commission (Docket UE-
121421) requested Avista continue assessing the impacts of a hypothetical portfolio
without Colstrip and provide the overall impacts on rates. TAC members requested
another scenario to analyze higher operating costs and shorter EPA compliance
timelines. Avista evaluated both continued operation and retirement of Colstrip under
each of these scenarios.

Modeling results for Colstrip in the Expected Case indicate Avista ownership interests in
the plant will remain cost effective for the next 20 years. The IRP assumes certain
capital investments will satisfy future state and federal regulations over the IRP
timeframe. The type, amount, and timing of capital expenditures are estimates used for
modeling purposes because exact dates and costs are unknown at this time. Future
IRPs will update assumptions as more and better information is available. The potential
capital investments include emerging requirements related to coal combustion residuals
(CCR) and Regional Haze-related controls. Other environmental regulations may drive
future investment requirements, such as ash pond improvements and the installation of
a system for NOX control. IRP modeling assumes that a default control system of a
selective catalytic reduction (SCR) will be required by the end of 2026, but the specific
target date or control type is unknown at this time.

Colstrip Retires in 2026 Scenario
This scenario assumes plant closure at the end of 2026 under the Expected Case’s
market forecast. This closure date eliminates capital spending for the SCR, accelerates
ash pond decommissioning, and alters ongoing capital and O&M spending at the plant.
This scenario assumes all costs related to existing and future capital spending would
fully depreciate five years after closure. It also assumes capital spending for ash pond
closure and no additional shutdown costs beyond the amount included in current
depreciation schedules for the plant. The scenario does not include any costs related to
employee retraining or relocation costs, payments to other owners, or costs to
decommission the plant beyond those included in current rates.

The results of the 2026 year-end closure scenario require 208 MW of new winter
capacity, assuming a replacement resource in Avista’s balancing area. Table 12.4
provides details about the resource strategy in this scenario. The strategy for this
scenario adds a second CCCT to replace the Colstrip capacity and serve future load
growth. Figure 12.2 shows a full efficient frontier analysis for this scenario. Levelized
power supply costs increase by $13.2 million or 3.6 percent per year across all years of
the IRP study. Portfolio risk increases by $12 million in 2027, or 16.6 percent. While the
3.6 percent cost impact appears to be modest due to the IRP’s method of levelizing
large future costs across the 20-year study timeframe, the annual cost increases in
Figure 12.3 are significant beginning in 2027.
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-5

Table 12.4: Colstrip Retires in 2026 Scenario Resource Strategy

Resource By End of
Year
ISO Conditions
(MW)
Natural Gas-Fired Peaker 2020 96
Thermal Upgrades 2021-2025 38
Natural Gas-Fired CCCTs 2026 627
Total 761

Conservation (w/ T&D losses) 2016-2035 130.7

Figure 12.2: Colstrip Retires Scenario Efficient Frontier Analysis

Between 2016 and 2021, customer costs increase due to accelerated recovery of
existing capital investments in the plant. In 2022-2026, the model assumes spending to
maintain and improve the plant continues at a lower rate, but most costs typically
classified as capital spending are expensed, leading to an earlier recovery of spending.
The elimination of the SCR offsets and lowers recovered Colstrip costs as high cost
investments are removed. The biggest cost to customers is replacement capacity. In
2027, this amounts to $58 million in added costs, or 13 percent. To put this into
perspective, Avista’s 2015 electric revenue requirement in that year is $900 million.
Assuming non-power supply costs increased at the rate of load growth, closing Colstrip
alone would increase customer rates by 5.7 percent the first year of closure.

$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550
20
2
7

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Levelized Annual Power Supply Cost (2016-40, Millions)
Efficient Frontier
Preferred Resource Strategy
Colstrip Retires 2026 Scenario- Efficient Frontier
Colstrip Replacement Resource Strategy
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-6

Figure 12.3: Colstrip Retires in 2026 Scenario Power Supply Cost Impact

Avista greenhouse gas emissions decline by an estimated 0.9 million metric tons per
year, or 32 percent. Figure 12.4 shows the change in emissions by year. In 2027, the
first year of closure in the scenario, the cost per saved metric ton of carbon is $66.

Figure 12.4: Colstrip Retires in 2027 Emissions

-$50
-$25
$0
$25
$50
$75
$100
$0
$100
$200
$300
$400
$500
$600
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
An
n
u
a
l

C
o
s
t

C
h
a
n
g
e

(
M
i
l
l
i
o
n
s
)
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
)
Delta
Preferred Resource Strategy
Colstrip Retires Scenario Resource Strategy
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
Me
t
r
i
c

T
o
n
s

(
M
i
l
l
i
o
n
s
)
Colstrip Retires Scenario Resource Strategy
Preferred Resource Strategy
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-7

High-Cost Colstrip Retention Scenario
The TAC proposed a second Colstrip case. The High-Cost Colstrip Retention scenario
assumes replacing existing SO2 scrubbers, converting the plant to dry ash handling,
landfill replacement, acceleration of SCR installation to 2022, and added O&M costs
due to the assumed closure of Colstrip Units 1 and 2 in 2017. While offering to perform
an analysis of High-Cost Colstrip Retention, Avista does not believe this scenario
represents a likely future for Colstrip and therefore has not vetted these assumptions
closely. The scenario provides a very high and unlikely case to test the viability of the
plant under much higher costs. A third scenario evaluates closing the plant in 2022 to
avoid the higher ongoing costs associated with the High-Cost Colstrip Retention case.
The resource strategy selected by PRiSM for this scenario is in Table 12.5; it is very
similar to the portfolio scenario with the plant retiring in 2027, but the scenario offsets
other plant requirements differently causing a small increase in capacity need (770 MW
versus 761 MW).

The High-Cost Colstrip scenario in Figure 12.5 uses the efficient frontier methodology to
measure cost and risk. It increases fixed costs by $18 million per year levelized
between 2016 and 2040 and risk levels do not change. Where Colstrip retires in 2022 to
avoid High-Cost Colstrip Retention costs, overall system cost increases $2 million per
year; risk increases by $11 million in 2027. The annual costs for the Colstrip scenarios
are in Figure 12.6 in 2023. The first year without Colstrip costs increase by $19 million
compared to the plant operating with the higher costs. This scenario shows with higher
operating costs, the plant is still marginally economic to continue operating.

Table 12.5: Colstrip Retires in 2022 Scenario Resource Strategy

Resource By End of
Year
ISO
Conditions
(MW)
Natural Gas Peaker 2020 56
Thermal Upgrades 2021-2035 41
Combined Cycle CTs 2023-2026 627
Natural Gas Peaker 2035 47
Total 770

Conservation (w/ T&D losses) 2016-2035 131

Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-8

Figure 12.5: High-Cost Colstrip Retention Scenario Efficient Frontier

Figure 12.6: High-Cost Colstrip Scenarios Annual Cost

$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550
20
2
7

S
t
d
e
v

(
M
i
l
l
i
o
n
s
)
Annual Levelized Portfolio Cost (Millions)
Expected Case- Efficient Frontier
High-Cost Colstrip Retention Scenario- Efficient Frontier
Colstrip Retires 2022 Efficient Frontier
Preferred Resource Strategy
PRS w/ High-Cost Colstrip Retention ScenarioColstrip Retires in 2022 Scenario Resource Strategy
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
PRS 252 276 288 298 313 331 341 348 366 382 399 454 484 487 488 504 523 523 545 564
PRS High Colstrip Costs 252 276 290 301 320 347 373 389 404 414 425 477 507 510 511 528 547 546 568 588
PRS Colstrip Retires 2022 260 284 299 325 333 336 351 408 425 432 444 490 498 500 501 514 534 536 551 572
$0
$100
$200
$300
$400
$500
$600
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
)
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-9

Social Cost of Carbon Market Scenarios
Chapter 10 describes alternative market scenarios. One modeled scenario was the
market impact of a social cost of carbon added to all carbon emissions. This section
describes the cost and portfolio impacts of such a market environment to Avista. Figure
12.7 is the efficient frontier of the Expected Case compared to the efficient frontier
developed for the Social Cost of Carbon market scenario. With the social cost of
carbon, the cost of the PRS increases by $67 million per year, or 17 percent. Risk also
increases by $4 million or 6 percent in 2027 for the same portfolio as the PRS.

Figure 12.7: Social Cost of Carbon Impact to Efficient Frontier

Colstrip Retires in 2027 with Social Cost of Carbon
Adding a fee to emit carbon will increase portfolio costs. This scenario analyzes the cost
effectiveness of keeping Colstrip open with the Social Cost of Carbon adder. The cost of
retiring Colstrip is approximately $6 million higher per year with the plant closed
compared to operating with the additional carbon pricing. Not only are system costs
higher with the closure of Colstrip in this scenario, but risk increases by 15 percent. See
Figure 12.8. This indicates Colstrip is still economic even with carbon pricing
approximately 10 times higher than in the Expected Case. The combination of the
Social Cost of Carbon with the assumptions from the High-Cost Colstrip Retention
scenario would find the plant marginally uneconomic, but as explained earlier, Avista
does not believe the assumptions of the High-Cost Colstrip Retention scenario are
realistic. The Social Cost of Carbon case reduces carbon emissions without Colstrip
$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550 $600
20
2
7

S
t
d
e
v

(
m
i
l
l
i
o
n
s
)
Annual Levelized Porfolio Cost (Millions)
Expected Case- Efficient Frontier
PRS (Expected Case)
Social Cost of Carbon Case- Efficient Frontier
PRS (Social Cost of Carbon)
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-10

retiring. In this scenario, emissions decline by 12 percent; if Colstrip retires, emissions
fall 24 percent in total (See Figure 12.9).

Figure 12.8: Colstrip Retires in 2027 Portfolio Efficient Frontier

Figure 12.9: Colstrip Retires in 2027 Portfolio Emissions

$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550 $600
20
2
7

S
t
d
e
v

(
m
i
l
l
i
o
n
s
)
Annual Levelized Porfolio Cost (Millions)
Social Cost of Carbon Case- Efficient Frontier
PRS (Social Cost of Carbon)
Social Cost of Carbon Case- Efficient Frontier- Colstrip Retires 2026
Colstrip Retires PRS (Social Cost of Carbon)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
Me
t
r
i
c

T
o
n
s

(
M
i
l
l
i
o
n
s
)
Expected Case (PRS)
Social Cost of Carbon (Colstrip Retires 2026)
Social Cost of Carbon (PRS)
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-11

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. They are similar to how Avista selected
resource strategies prior to its 2003 IRP and the adoption of more sophisticated
modeling tools such as PRiSM and Monte Carlo risk analysis. Figure 12.10 shows the
levelized cost and 2027 risk compared to the efficient frontier.

Figure 12.10: Other Resource Strategy Portfolio Cost and Risk (Millions)
Market and Conservation
The Market and Conservation portfolio shows the cost and risk if the utility chose not to
fill its capacity need with generation assets, instead dependeding on the wholesale
market for its future needs. This portfolio helps estimate the value of capacity in the
PRS. It assumes the same amount of conservation as the PRS. This portfolio’s cost is
$28 million per year levelized lower than the least cost portfolio, and the risk is $1
million higher in 2027. The cost difference between this portfolio and the least cost
represents the cost of capacity or the added cost of reliability. Given this strategy does
not meet reliability targets, it is not an acceptable portfolio. Utilities may lean toward this
type of portfolio when the market place is long on resources, which is not the case
beginning in 2021.

2013 Preferred Resource Strategy
This portfolio emulates the strategy selected in the 2013 IRP. The 2013 PRS portfolio
includes the resources described in Chapter 11, predominantly natural gas-fired
Market &
Conservation
2013 PRS
Renewables Meet All
Load Growth
Peakers & Hydro
Total Portfolio
Colstrip Retires 2027
PRS
Hydro Upgrades &
Peakers
$0
$20
$40
$60
$80
$100
$120
$350 $400 $450 $500 $550
20
2
7

S
t
d
e
v
Annual Levelized Portfolio Cost
Least Cost
Least Risk
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-12

peakers, a CCCT, and demand response. The portfolio reflects the current lower load
growth trajectory by eliminating a peaker from the previous strategy. This strategy’s
levelized cost is $3 million higher than the PRS, and risk is $1.5 million higher in 2027.
With the exception of the demand response, this portfolio is similar to the current PRS
results with similar metrics for cost and risk.

Renewables Meet All Load Growth
The Renewables Meet All Load Growth scenario is similar to a higher RPS scenario.
The objective is to meet all energy load growth with renewables along with meeting
capacity requirements. This scenario meets energy needs with newly acquired
renewable resources and natural gas-fired generation for capacity needs. The model
selected 250 MW of wind (87 aMW) with a 20 percent apprentice REC credit, plus an
upgrade to the Kettle Falls plant; with rollover ability, these renewables meet the 126
aMW requirement each year.

The added renewables, in addition to the capacity resources, add $18 million per year
to power supply expenses relative to the Expected Case, and lower risk in 2027 by $3
million. Avista could get the same amount of risk reduction by selecting a portfolio on
the efficient frontier with an annual $15 million reduction in cost.

Hydroelectric Upgrades and Peakers
This scenario uses a combination of peakers and hydroelectric upgrades to meet future
capacity needs. The scenario completes major upgrades at Long Lake and Monroe
Street during the IRP timeframe; natural gas-fired peakers meet all remaining capacity
needs. Costs increase by $6 million per year in this scenario, and risk increases by $4
million. An interesting result from the scenario is the increased risk metric. Typically,
more renewables reduce risk, but since hydro is highly correlated with the Northwest
marketplace, the upgrades actually increase risk relative to the PRS.

Peakers and Hydro Total Portfolio
A future with no coal or baseload natural gas resources is the premise of this scenario.
It retires Avista’s CCCTs and coal by 2027, replacing them with upgrades at
hydroelectric facilities and the construction of natural gas-fired peaking plants. In 2027,
when the retirements occur, the risk metric increases by $27 million; costs are $80
million higher compared to the PRS.

Risk-Adjusted PVRR
Avista believes efficient frontier analysis paired with robust analytics and data is a
superior method to measure tradeoffs between average costs and risk. Chapter 11
details the risk-adjusted PVRR methodology used to analyze the efficient frontier. Risk-
adjusted PVRR is helpful with measuring risk in handpicked portfolios that that do not
fall on the efficient frontier, or where the efficient frontier is not part of the IRP process.
Figure 12.11 shows the risk-adjusted PVRR analysis results for the other resource
strategy scenarios in this section. The portfolio with the lowest cost is the Market and
Conservation portfolio. This portfolio does not meet reliability objectives of the IRP, and
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-13

is not an acceptable option. The next lowest cost portfolio is the PRS, followed by the
2013 PRS.

Figure 12.11: Risk Adjusted PVRR (2016- 2035)
Resource Tipping Point Analyses
Recent Avista IRPs studied through tipping point analyses show how much capital costs
needed to change before different resource selections occurred in the PRS. The 2013
IRP included solar, nuclear, and IGCC coal tipping point analyses. This IRP includes
tipping point analyses for solar, energy storage, and demand response. As emerging
technology costs generally do not follow typical inflation, tipping point analyses are
important to understand at what point such technologies might affect the PRS.

Utility Scale Solar
The IRP assumes utility scale solar has a $1,500 per kW capital cost for fixed panel and
$1,600 per kW (2014 dollars) for single-axis tracking panel facilities. Avista estimates
solar costs will decline in real dollars by 27 percent over the 20-year planning horizon
and the 10 percent federal investment tax credit is available after 2016. Solar does not
provide winter on-peak capability. Therefore, the resource must be cost competitive with
wholesale market commodity prices.

The analysis decreases single axis solar capital costs in PRiSM until the model selects
the resource in the PRS. PRiSM selects solar in 2023 when its price falls 47 percent
below current projections, to $682 per kW in 2014-year dollars. Figure 12.11 shows the
solar cost curve and the point where solar becomes economic to Avista.

$3.8 Bil
$4.1 Bil $4.2 Bil $4.2 Bil $4.3 Bil $4.3 Bil $4.4 Bil
$ Bil
$1 Bil
$2 Bil
$3 Bil
$4 Bil
$5 Bil
Market &
Conservation
2015 IRP's
PRS
2013 IRP's
PRS
Hydro
Upgrades &
Peakers
Colstrip
Retires in
2027
Renewables
Meet All
Load Growth
Peakes &
Hydro Total
Portfolio
Ri
s
k

A
d
j
u
s
t
e
d

P
V
R
R
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-14

Figure 12.12: Utility Scale Solar Tipping Point Analysis (2014 $)

Utility Scale Energy Storage
Energy storage might become a commercial-scale resource for utilities and their
customers in the future. As the amount of intermittent generation grows, many believe
energy storage will help integrate these resources into the electricity grid. There are
many types of energy storage technologies, but this study remains agnostic to the
technology and only looks at how costs change, as long as each technology performs
similarly. Similar to solar generation, energy storage costs should decline as the
technology becomes more common. Unlike solar, energy storage can meet on peak
needs, but it consumes significant amounts of energy in the form of losses in the
process. The Expected Case assumes storage at $4,000 per kW in 2014. By the first
capacity need in 2021, utility scale energy storage is expects to be $2,736 per kW
(2014$) or $3,201/kW nominal. PRiSM first selects storage in 2021 with a price $770
per kW in 2014-year dollars, a 72 percent reduction in capital costs.

$/ kW
$200/ kW
$400/ kW
$600/ kW
$800/ kW
$1,000/ kW
$1,200/ kW
$1,400/ kW
$1,600/ kW
$1,800/ kW
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
Real ($2014)
Selected ($2014)
Chapter 12 – Portfolio Scenarios
Avista Corp 2015 Electric IRP 12-15

Figure 12.13: Utility Scale Storage Tipping Point Analysis (2014 $)

Demand Response
Demand response was part of the PRS in Avista’s 2013 IRP. At that time, the costs
were preliminary internal estimates; since then, Avista sponsored a study to determine
the demand response costs and quantities available. The results of the study showed
higher prices than the 2013 plan, and the higher costs meant demand response is not in
this plan. To make demand response attractive, costs must fall to $117 per kW-year
levelized between 2023 and 2035. This is a reduction of 46 percent.
$/ kW
$500/ kW
$1,000/ kW
$1,500/ kW
$2,000/ kW
$2,500/ kW
$3,000/ kW
$3,500/ kW
$4,000/ kW
$4,500/ kW
20
1
4
20
1
5
20
1
6
20
1
7
20
1
8
20
1
9
20
2
0
20
2
1
20
2
2
20
2
3
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
Real ($2014)
Selected ($2014)

Chapter 13–Action Items
Avista Corp 2015 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 2013 IRP Action Plan and provides the 2015 Action Plan.

Summary of the 2013 IRP Action Plan
The 2013 Action Plan included three categories: generation resource related analysis,
energy efficiency, and transmission planning.

2013 Action Plan and Progress Report

Generation Resource Related Analysis
 Consider Spokane and Clark Fork River hydroelectric upgrade options in the next
IRP as potential resource options to meet energy, capacity, and environmental
requirements.
o This IRP continues incorporating hydroelectric upgrades as resource
options in the PRS and scenario analysis. Chapter 9 – Generation
Resource Options provides details about the hydroelectric upgrades
evaluated for this IRP.

 Continue to evaluate potential locations for natural gas-fired resources identified to
be online by the end of 2019, including environmental reviews, transmission studies,
and potential land acquisition.
o The natural gas-fired peaker options included in this IRP assume both
greenfield and brownfield sites in Northern Idaho. Avista is currently
negotiating the purchase of property for a greenfield site. Information
about this site will not be available publically until after the close of the
potential transaction.

 Continue participation in regional IRP and regional planning processes, monitor
regional surplus capacity, and continue to participate in regional capacity planning
processes.
o Avista continues to monitor and review other Northwest IRP processes.
o The company continues to participate in regional processes including the
development of the Seventh Regional Power Plan, PNUCC studies, and
work by the Western Governors Association on energy issues.

 Commission a demand response potential and cost assessment of commercial and
industrial customers per its inclusion in the middle of the PRS action plan.
o Avista retained the services of AEG to study the amount and cost of
different types of demand response programs available in the service
Chapter 13–Action Items
Avista Corp 2015 Electric IRP
territory. A discussion about the scope of this study occurred with the TAC
during the first meeting on May 29, 2014, and the results presented at the
fourth TAC meeting on February 24, 2015. Both of these presentations are
available in Appendix A.
o The complete AEG demand response study is available in Appendix C.

 Continue monitoring state and federal climate change policies and report work from
Avista’s Climate Change Council.
o Several developments concerning state and federal climate change
policies have occurred since publication of the 2013 IRP. Most notably,
the CPP at the federal level and Washington Governor Inslee’s Executive
Order 14-04 concerning climate change and subsequent proposed
legislation concerning a cap and trade program at the state level.
o Details about the CPP proposal and Governor Inslee’s Executive Order
are available in Chapter 7 – Policy Considerations. Studies concerning
these areas are included in chapter 12 – Portfolio Scenarios. The original
presentations made to the TAC about these issues are in Appendix A.

 Review and update the energy forecast methodology to better integrate economic,
regional, and weather drivers of energy use.
o Please refer to Chapter 3 – Economic and Load Forecast for a detailed
account of changes made to the energy forecast methodology to better
integrate economic, regional, and weather drivers of energy use. Avista’s
chief economist presented the forecasting methodology updates at the
second TAC meeting on September 24, 2014. The presentation is
available in Appendix A

 Evaluate the benefits of a short-term (up to 24-months) capacity position report.
o Avista implemented a short-term capacity model in late 2013. The tool
assists in closing short capacity positions. An updated version of this tool
added long-term functionality to develop resource positions for this plan.

 Evaluate options to integrate intermittent resources.
o Avista completed development of the Avista Decision Support System
(ADSS); this tool can model the costs and benefits of intermittent
resources. A presentation about the model and the results of the value of
thermal resources assisting with ancillary services study occurred at the
May 19, 2015, Technical Advisory Committee meeting. This presentation
is located in Appendix A.

Energy Efficiency
 Work with NPCC, the UTC, and others to resolve adjusted market baseline issues
for setting energy efficiency target setting and acquisition claims in Washington.
o Avista hired AEG to conduct the biannual CPA. The study complied with
accepted NPPC methodologies where possible by using measure savings
Chapter 13–Action Items
Avista Corp 2015 Electric IRP
identified by the RTF or estimated by AEG. Where RTF unit energy
savings are utilized those savings will be symmetrically applied when
Avista claims the energy savings for the biennium. AEG is currently in the
process of updating inputs for the CPA to include indexing the CPA to the
forecast and other economic factors to address changing market
conditions.

 Study and quantify transmission and distribution efficiency projects as they apply to
EIA goals.
o Avista continues to invest in transmission and distribution projects
including efficiency upgrades. Chapter 8 contains details about completed
and announced projects.

 Assess energy efficiency potential on Avista’s generation facilities.
o Avista completed an energy audit on owned generating facilities. Chapter
5 – Energy Efficiency and Demand Response summarizes the results and
Appendix D includes the audit reports.

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.

2013 Action Plan and Progress Report – Supplemental
Avista submitted eight updated Action Items on January 27, 2014 in response to
comments made at the January 9, 2014 hearing with the WUTC. This section highlights
the work done in this IRP concerning the additional Action Items.

Generation Resource Related Analysis – Additional Updates

 Continue to evaluate scenarios related to Colstrip and how each scenario may
impact power supply costs.
o The 2015 IRP includes several Colstrip scenarios in Chapter 10 – Market
Analysis and Chapter 12 – Portfolio Scenarios.
Chapter 13–Action Items
Avista Corp 2015 Electric IRP
 Evaluate and explicitly document various options for quantifying carbon costs in the
IRP.
o Avista discussed different options concerning the quantification of the cost
of carbon in the Expected Case and in scenarios for the 2015 IRP. The
presentations made to the TAC are in Appendix A and the results of the
analyses are in chapters 10, 11 and 12.
 Work with TAC to determine which carbon quantification method should be
employed in the Expected Case of the 2015 IRP.
o Avista’s discussions with the TAC about different options for the
quantification of the cost of carbon in the Expected Case for the 2015 IRP
are in the presentations made to the TAC in Appendix A. The Expected
Case analysis concerning carbon emissions are in chapters 10 and 11.
 Use Avista’s new modeling capabilities to further evaluate the benefits of storage
resources to its generation portfolio, including the impacts on ancillary services
needs.
o Chapter 9 – Generation Resource Options and chapter 12 – Portfolio
Scenarios discuss the results of the evaluation of energy storage to
Avista’s generation portfolio.
 Revisit with the TAC the benefits and costs of the Company’s 2013 IRP planning
margin target to determine if a different level is warranted in the 2015 IRP.
o Avista discussed the planning margin target with the TAC. The
presentations concerning those discussions are in Appendix A. Chapter 6
– Long-Term Position has an extensive discussion about the choice of the
appropriate planning margin for the 2015 IRP.
 Evaluate with the TAC the impacts of different points along the efficient frontier.
o Avista discussed the evaluation of the impacts of choosing different points
along the efficient frontier with the TAC. The presentations concerning
those discussions are in Appendix A and details about the results in this
IRP are located in chapters 11 – Preferred Resource Strategy and 12 –
Portfolio Scenarios.

Energy Efficiency – Additional Updates

 Evaluate the impacts of targeting individual or groups of energy efficiency options
within PRiSM instead of targeting quantities using avoided cost.
o Avista developed and used a secondary methodology for identifying the
amount of achievable conservation potential using the PRiSM model.
Details about PRiSM co-optimization are in Chapter 5 – Energy Efficiency
and Demand Response.
 Work with TAC to determine if 2015 IRP should continue the historical method of
conservation quantification or if PRiSM should be used instead.
o The TAC meetings included discussions about the PRiSM co-optimization
methodology for identifying the amount of energy efficiency potential for
the 2015 IRP. Appendix A contains the presentation materials.
Chapter 13–Action Items
Avista Corp 2015 Electric IRP
2015 IRP Two Year Action Plan
Avista’s 2015 PRS provides direction and guidance for the type, timing, and size of
future resource acquisitions. The 2015 IRP Action Plan highlights the activities planned
for possible inclusion in the 2017 IRP. Progress and results for the 2015 Action Plan
items are reported to the TAC and the results will be included in Avista’s 2017 IRP. The
2015 Action Plan includes input from Commission Staff, Avista’s management team,
and the TAC.

Generation Resource Related Analysis
 Analysis of the continued feasibility of the Northeast Combustion Turbine due to its
age.
 Continue to review existing facilities for opportunities to upgrade capacity and
efficiency.
 Increase the number of manufacturers and sizes of natural gas-fired turbines
modeled for the PRS analysis.
 Evaluate the need for, and perform if needed, updated wind and solar integration
studies.
 Participate and evaluate the potential to join a Northwest EIM.
 Monitor regional winter and summer resource adequacy.
 Participate in state level implementation of the CPP.

Energy Efficiency
 Continue to study and quantify transmission and distribution efficiency projects as
they apply to EIA goals.
 Complete the assessment of energy efficiency potential on Avista’s generation
facilities.

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.

Chapter 13–Action Items
Avista Corp 2015 Electric IRP
Production Credits

Primary Avista 2015 Electric IRP Team

Individual Title Contribution
Clint Kalich Manager of Resource Planning & Analysis Project Manager
James Gall Senior Power Supply Analyst Analysis/Author
John Lyons Senior Resource Policy Analyst Research/Author/Editor
Grant Forsyth Senior Forecaster & Economist Load Forecast
Richard Maguire System Planning Engineer Transmission & Distribution

2015 Electric IRP Contributors

Name Title
Thomas Dempsey Manager, Generation Joint Projects
Leona Doege DSM Program Manager
Tom Pardee Natural Gas Planning Manager
Shane Pacini Manager Network Engineering
Eric Scott Natural Gas Resources Manager
Mike Dillon DSM Planning and Analytics Manager
Jeff Schlect Senior Manager of FERC Policy and Transmission Services
Dave Schwall Senior Engineer
Darrell Soyars Manager of Corporate Environmental Compliance
Xin Shane 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]