20170609Morehouse Exhibit 7.pdf

DAVID J. MEYER
VICE PRESIDENT AND CHIEF COUNSEL FOR
REGULATORY & GOVERNMENTAL AFFAIRS
AVISTA CORPORATION
P.O. BOX 3727
1411 EAST MISSION AVENUE
SPOKANE, WASHINGTON 99220-3727
TELEPHONE: [phone]
FACSIMILE: [phone]
[e-mail]

BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION

IN THE MATTER OF THE APPLICATION ) CASE NO. AVU-G-17-01
OF AVISTA CORPORATION FOR THE )
AUTHORITY TO INCREASE ITS RATES )
AND CHARGES FOR ELECTRIC AND )
NATURAL GAS SERVICE TO ELECTRIC ) Exhibit No. 7
AND NATURAL GAS CUSTOMERS IN THE )
STATE OF IDAHO ) JODY MOREHOUSE
)

FOR AVISTA CORPORATION

(NATURAL GAS ONLY)

2016
Natural Gas Integrated Resource Plan
August 31, 2016
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 1 of 162
Safe Harbor Statement

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

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 2 of 162
TABLE OF CONTENTS
0 Executive Summary………………………………………………..Page 1
1 Introduction………………………………………………………….Page 15
2 Demand Forecasts…………………………………………………Page 27
3 Demand Side Resources………………………………………….Page 45
4 Supply Side Resources……………………………………………Page 61
5 Integrated Resource Portfolio…………………………………….Page 83
6 Alternate Scenarios, Portfolios, and Stochastic Analysis……..Page 113
7 Distribution Planning………………………………………………Page 129
8 Action Plan…………………………………………………………Page 139
9 Glossary of Terms and Acronyms……………………………….Page 147

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 3 of 162
Executive Summary
Avista Corporation’s 2016 Natural Gas Integrated Resource Plan (IRP) identifies a
strategic natural gas resource portfolio to meet customer demand requirements over the
next 20 years. While the primary focus of the IRP is meeting customers’ needs under
peak weather conditions, this process also provides a methodology for evaluating
customer needs under normal or average conditions. The formal exercise of bringing
together customer demand forecasts with comprehensive analyses of resource options,
including supply-side resources and demand-side measures, is valuable to Avista, its
customers, regulatory agencies, and other stakeholders for long-range planning.

IRP Process and Stakeholder Involvement
The IRP is a coordinated effort by several Avista departments along with input from our
Technical Advisory Committee (TAC), which includes Commission Staff, peer utilities,
customers, and other stakeholders. The TAC is a vital component of our IRP process, as
it provides a forum for discussing multiple perspectives, identifies issues and risks, and
improves analytical planning methods. Topics discussed with the TAC include natural gas
demand forecasts, price forecasts, demand-side management (DSM), supply-side
resources, modeling tools, and distribution planning. The process results in a resource
portfolio designed to serve our customers’ natural gas needs while balancing cost and
risk.

Planning Environment
A long-term resource plan addresses the uncertainties inherent in any planning exercise.
Compared to prior planning cycles, there is relatively more certainty about the availability
of economically extractable natural gas. However, some of the future uses of this energy
resource are unknown. There are questions concerning an industrial renaissance,
liquefied natural gas (LNG) exports, natural gas vehicles, and power generation. We
continue to analyze key assumptions by evaluating multiple scenarios over a range of
possible outcomes to address the uncertainties.

Demand Forecasts
Avista defines eight distinct demand areas in this IRP structured around the pipeline
transportation and storage resources that serve them. Demand areas include Avista’s
four service territories (Washington/Idaho; Medford/Roseburg, Oregon; Klamath Falls,
Oregon and La Grande, Oregon) and then disaggregated by the pipelines that serve
them. The Washington/Idaho service territory includes areas served only by Northwest
Pipeline (NWP), only by Gas Transmission Northwest (GTN), and by both pipelines. The
Medford service territory includes an area served by NWP and GTN.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 4 of 162
Avista recognizes and accounts for weather, customer growth and use per customer as
the most significant demand influencing factors. Other demand influencing factors include
population, employment, age and income demographics, construction levels,
conservation technology, new uses (e.g. natural gas vehicles), and use-per-customer
trends.
Recognizing that customers may adjust consumption in response to price, Avista
analyzed factors that could influence natural gas prices and demand through price
elasticity. These factors include:

 Supply: shale gas, industrial use, and export LNG.
 Infrastructure: regional pipeline projects, national pipeline projects, and
storage.
 Regulatory: subsidies, market transparency/speculation, and carbon
legislation.
 Other: drilling innovation, thermal generation and energy correlations (i.e.
oil/gas, coal/gas, and liquids/gas).

Avista developed a historical-based reference case and conducted sensitivity analysis on
key demand drivers by varying assumptions to understand how demand changes. Using
this information, and incorporating input from the TAC, Avista created alternate demand
scenarios for detailed analysis. Table 1 summarizes these demand scenarios, which
represent a broad range of potential scenarios for planning purposes. The Average Case
represents Avista’s demand forecast for normal planning purposes. The Expected Case
is the most likely scenario for peak day planning purposes.

Table 1: Demand Scenarios
2016 IRP Demand Scenarios

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 5 of 162
The IRP process defines the methodology for the development of two primary types of
demand forecasts – annual average daily and peak day. The annual average daily
demand forecast is useful for preparing revenue budgets, developing natural gas
procurement plans, and preparing purchased gas adjustment filings. Forecasts of peak
day demand are critical for determining the adequacy of existing resources or the timing
for new resource acquisitions to meet our customers’ natural gas needs in extreme
weather conditions. The demand forecasts from the Average and Expected Cases
revealed the data shown in Table 2:

Table 2: Annual Average and Peak Day Demand Cases (Dth/day)
Year Annual Average Daily
Demand
Peak Day Demand Non-coincidental
Peak Day Demand

Annual Average Daily Demand – Expected average day, system-wide core demand
increases from an average of 94,164 dekatherms per day (Dth/day) in 2016 to 102,840
Dth/day in 2035. This is an annual average growth rate of 0.5 percent and is net of
projected conservation savings from DSM programs. Appendix 3.1 shows gross demand,
conservation savings and net demand.
Peak Day Demand – The peak day demand for the Washington/Idaho and La Grande
service territories is modeled on and around February 15 of each year. For the
southwestern Oregon service territories (Medford, Roseburg, Klamath Falls), the model
assumes this event on and around December 20 each year. Expected coincidental peak
day, or the sum of demand from each territories modeled peak, the system-wide core
demand increases from a peak of 361,901 Dth/day in 2016 to 425,144 Dth/day in 2035.
Forecasted non-coincidental peak day demand, or the sum of demand from the highest
single day including all forecasted territories, peaks at 331,820 Dth/day in 2016 and
increases to 387,742 Dth/day in 2035, a 0.8 percent average annual growth rate in peak
day requirements. This is also net of projected conservation savings from DSM programs.

Figure 1 shows forecasted average daily demand for the six demand scenarios modeled
over the IRP planning horizon.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 6 of 162
Figure 1: Average Daily Demand (Net of DSM Savings)

Figure 2 shows forecasted system-wide peak day demand for the six demand scenarios
modeled over the IRP planning horizon.

Figure 2: Peak Day Demand Scenarios (Net of DSM Savings)

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 7 of 162
Natural Gas Price Forecasts
Natural gas prices are a fundamental component of integrated resource planning because
the commodity price is a significant component of the total cost of a resource option. This
affects the avoided cost threshold for determining cost-effectiveness of conservation
measures. The price of natural gas also influences the consumption of natural gas by
customers. A price elasticity adjustment to use per customer reflects customer responses
to changing natural gas prices.
With more information known about the costs and volumes produced by shale gas there
appears to be market consensus that production costs will remain low for quite some time.
Avista expects continued low prices even with increased incremental demand for LNG
exports, transportation fuels, and increased industrial consumption.
The carbon legislation debate continues. Avista’s current estimate is that carbon
legislation will occur at both the federal level, through the Clean Power Plan, and on the
state level through a cap and trade or tax mechanism. Current IRP price forecasts include
a slightly higher carbon tax, occurring earlier than the 2014 IRP. Avista analyzed four
carbon sensitivities and their impact on demand forecasts to address the uncertainty
about carbon legislation.
Avista reviewed several price forecasts from credible sources and created a blended price
forecast to represent an expected price strip. A high and low price were developed via a
Monte Carlo simulation using historical daily cash price data at the Henry Hub trading
point dating back to 2009. These three price curves represent a reasonable range of
pricing possibilities for this IRP analysis. The range of prices provides necessary variation
for addressing uncertainty of future prices. Figure 3 depicts the price forecasts used in
this IRP.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 8 of 162
Figure 3: Low/Medium/High Henry Hub Forecasts (Real $/Dth)

Historical statistical analysis shows a long run consumption response to price changes.
In order to model consumption response to these price curves, Avista utilized an expected
elasticity response factor of -0.15 that was applied under various scenarios. Avista will
continue to monitor and research this assumption and make any necessary adjustments
as described in the Ongoing Activities section of Chapter 8 – Action Plan.

Existing and Potential Resources
Avista has a diversified portfolio of natural gas supply resources, including access to and
contracts for the purchase of natural gas from several supply basins; owned and
contracted storage providing supply source flexibility; and firm capacity rights on six
pipelines. For potential resource additions, Avista considers incremental pipeline
transportation, storage options, distribution enhancements, and various forms of LNG
storage or service. Beginning in Avista’s 2018 IRP and all future planning documents and
analysis thereafter, Avista intends to include conservation as a potential resource
addition.
Avista models aggregated conservation potential that reduces demand if the conservation
programs are cost-effective over the planning horizon. The identification and
incorporation of conservation savings into the SENDOUT® model utilizes projected
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 9 of 162
natural gas prices and the estimated cost of alternative supply resources. The operational
business planning process starts with IRP identified savings and ultimately determines
the near-term program offerings. Given current avoided costs, a limited number of DSM
programs are cost effective in Idaho, Oregon, and Washington. Avista actively promotes
cost-effective efficiency measures to our customers as one component of a
comprehensive strategy to arrive at mix of best cost/risk adjusted resources.

Resource Needs In the Average Case demand scenario, the analysis showed no resource deficiencies in
the 20-year planning horizon given Avista’s existing supply resources. The Expected
Case demand scenario, using the existing resources, determined when the first year peak
day demand would not be fully served. Figure 4 summarizes the results of this portfolio.
Avista is not resource deficient in the Expected Case in the 20-year planning horizon.

Figure 4: Expected Case – First Year Demand Not Met with Existing Resources

Figures 5 through 8 illustrate Avista’s peak day demand by service territory for both this
and the prior IRP. These charts compare existing peak day resources to expected peak
day demand by year and show the timing and extent of resource deficiencies, if any, for
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 10 of 162
the Expected Case. Based on this information, and more specifically where a resource
deficiency is nearly present as shown in Figure 6, Avista has time to carefully monitor,
plan and take action on potential resource additions as described in Ongoing Activities
section of Chapter 8 – Action Plan. Any underutilized resources will be optimized to
mitigate the costs incurred by customers until the resource is required to meet demand.
This management of long- and short-term resources provides the flexibility to meet firm
customer demand in a reliable and cost-effective manner as described in Supply Side
Resources – Chapter 4.

Figure 5: Expected Case – WA/ID Existing Resources vs. Peak Day Demand (Net of DSM)

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 11 of 162
Figure 6: Expected Case – Medford/Roseburg Existing Resources vs. Peak Day Demand
(Net of DSM)

Figure 7: Expected Case – Klamath Falls Existing Resources vs. Peak Day Demand
(Net of DSM)

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 12 of 162
Figure 8: Expected Case – La Grande Existing Resources vs. Peak Day Demand
(Net of DSM)

A critical risk is the slope of forecasted demand growth, which although increasing
continues to be almost flat in Avista’s current projections. This outlook implies that existing
resources will be sufficient within the planning horizon to meet demand. However, if
demand growth accelerates, the steeper demand curve could quickly accelerate resource
shortages by several years. Figure 9 conceptually illustrates this risk. In this hypothetical
example, a resource shortage does not occur until year eight in the initial demand case.
However, the shortage accelerates by five years under the revised demand case to year
three. This “flat demand risk” requires close monitoring of accelerating demand, as well
as careful evaluation of lead times to acquire the preferred incremental resource.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 13 of 162
Figure 9: Hypothetical Flat Demand Risk Example

Alternate Demand Scenarios
Avista performed the same analysis for five other demand scenarios: Average,
Expected/Low Price, High Growth/Low Price, Low Growth/High Price, and Coldest in 20
years. As expected, the High Growth/Low Price scenario has the most rapid growth and
is the only scenario with unserved demand. This “steeper” demand lessens the “flat
demand risk” discussed above, yet resource deficiencies occur very late in the planning
horizon. Figure 10 shows first year resource deficiencies under each scenario.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 14 of 162
Figure 10: Scenario Comparisons of First Year Peak Demand Not Met with Existing
Resources

Issues and Challenges
Even with the planning, analysis, and conclusions reached in this IRP, there is still
uncertainty requiring diligent monitoring of the following issues.

Demand Issues
Although the future customer growth trajectory in Avista’s service territory has slightly
increased compared to the 2014 IRP, the need in considering a range of demand
scenarios provides insight into how quickly resource needs can change if demand varies
from the Expected Case.
With an increase in natural gas supply and subsequent low costs, there is increasing
interest in using natural gas. Avista does not anticipate that traditional residential and
commercial customers will provide increased growth in demand. Power generation from
natural gas is increasingly being used to back up solar and wind technology as well as
replacing retired coal plants. There is also potential for increased natural gas usage in
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 15 of 162
other markets, such as transportation and as an industrial feedstock. Most of these
emerging markets will not be core customers of the LDC, however they will affect regional
gas infrastructure and could affect natural gas pricing.

Price Issues
Shale gas and drilling technology continues to change the face of North American gas
supply. The abundance of shale gas combined with lagging demand has created a near-
term supply glut that continues to suppress price levels. The mild winter of 2015-2016
brought decreased demand. This has led to a glut of natural gas in the market from both
record production and very high storage levels and will lead to a depressed market until
a supply and demand balance is found. These low prices are beneficial for customers,
but to address uncertainty in pricing, this plan includes high and low price scenarios along
with stochastic price analysis to capture a range of possible prices.

LNG Exports
The availability of natural gas in North America has changed global LNG dynamics.
Existing and new LNG facilities are looking to export low cost North American gas to the
higher priced Asian and European markets. In Canada, 20 LNG export projects are in
various stages of the permitting process. In the Northwest, there is one proposed terminal
in Oregon. How many of these terminals actually get approval and ultimately built is yet
to be determined. However, LNG exporting could alter the price, constrain existing
pipeline networks, stimulate development of new pipeline resources, and change flows of
natural gas across North America.

Action Plan
Avista’s 2017-2018 Action Plan outlines activities for study, development and preparation
for the 2018 IRP. The purpose of the action plan is to position Avista to provide the best
cost/risk resource portfolio and to support and improve IRP planning. The Action Plan
identifies needed supply and demand side resources and highlights key analytical needs
in the near term. It also highlights essential ongoing planning initiatives and natural gas
industry trends Avista will monitor as a part of its ongoing planning processes (Chapter 8
– Action Plan).
The price of natural gas has dropped significantly since the 2014 IRP. This is primarily
due to the amount of economically extractable natural gas in shale formations, efficient
drilling techniques and technology, and warmer than normal weather. Wells have been
drilled, but left uncompleted due to the poor market economics. This is depressing natural
gas prices and forcing many oil and gas companies into bankruptcy. Due to historically
low prices Avista will research market opportunities including: procuring a derivative
based contract, 10 year forward strip, and natural gas reserves.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 16 of 162
Avista’s 2018 IRP will contain a dynamic DSM program structure in its analytics. In prior
IRP’s it was a deterministic method based off of the Expected Case assumptions. In the
2018 IRP, each portfolio will have the ability to make a new selection of conservation to
meet unserved customer demand. Avista will explore methods to enable a dynamic
analytical process for the evaluation of conservation potential within individual portfolios.
Key ongoing components of the Action Plan include:
 Monitor actual demand for accelerated growth can address resource deficiencies
arising from exposure to “flat demand” risk. This will include providing Commission
Staff with IRP demand forecast-to-actual variance analysis on customer growth
and use per customer at least bi-annually.
 Continue to monitor supply resource trends including the availability and price of
natural gas to the region, LNG exports, methanol plants, supply and market
dynamics and pipeline and storage infrastructure availability.
 Monitor availability of resource options and assess new resource lead-time
requirements relative to resource need to preserve flexibility.
 Meet regularly with Commission Staff to provide information on market activities
and significant changes in assumptions and/or status of Avista activities related to
the IRP or natural gas procurement practices.
 Appropriate management of existing resources including optimizing underutilized
resources to help reduce costs to customers.

Conclusion
Slightly higher customer growth has been offset by lower use per customer and has
eliminated the need for Avista to acquire additional supply-side resources, therefore
appropriate management of underutilized resources to reduce costs until resources are
needed is essential. Additionally, the lower cost of natural gas continues to challenge the
cost-effectiveness of DSM programs. While Avista believes adoption of conservation is
the best strategy for minimizing costs to customers and meeting environmental goals, this
IRP shows a similar conservation potential as compared to the 2014 IRP, but is lower
than prior planning documents. The IRP has many objectives, but foremost is to ensure
that proper planning enables Avista to continue delivering safe, reliable, and economic
natural gas service to our customers.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 17 of 162
1: Introduction
Avista is involved in the production, transmission and distribution of natural gas and
electricity, as well as other energy-related businesses. Avista was founded in 1889 as
Washington Water Power and has been providing reliable, efficient and reasonably priced
energy to customers for over 125 years.

Avista entered the natural gas business with the purchase of Spokane Natural Gas
Company in 1958. In 1970, it expanded into natural gas storage with Washington Natural
Gas (now Puget Sound Energy) and El Paso Natural Gas (its interest subsequently
purchased by NWP) to develop the Jackson Prairie natural gas underground storage
facility in Chehalis, Washington. In 1991, Avista added 63,000 customers with the
acquisition of CP National Corporation’s Oregon and California properties. Avista
subsequently sold the California properties and its 18,000 South Lake Tahoe customers
to Southwest Gas in 2005. Figure 1.1 shows where Avista currently provides natural gas
service to approximately 334,000 customers in eastern Washington, northern Idaho and
several communities in northeast and southwest Oregon. Figure 1.2 shows the number
of natural gas customers by state.

Figure 1.1: Avista’s Natural Gas Service Territory

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 18 of 162
Figure 1.2: Avista’s Natural Gas Customer Counts

Avista manages its natural gas operation through two operating divisions – North and
South:
 The North Division covers about 26,000 square miles, primarily in eastern
Washington and northern Idaho. Over 840,000 people live in Avista’s
Washington/Idaho service area. It includes urban areas, farms, timberlands, and
the Coeur d’Alene mining district. Spokane is the largest metropolitan area with a
regional population of approximately 490,000 followed by the Lewiston,
Idaho/Clarkston, Washington and Coeur d’Alene, Idaho areas. The North Division
has about 73 miles of natural gas transmission pipeline and 9,100 miles in the
distribution system. The North Division receives natural gas at more than 40 points
along interstate pipelines and distributes it to over 235,000 customers.
 The South Division serves four counties in southwest Oregon and one county in
northeast Oregon. The combined population of these areas is over 480,000
residents. The South Division includes urban areas, farms and timberlands. The
Medford, Ashland and Grants Pass areas, located in Jackson and Josephine
Counties, is the largest single area served by Avista in this division with a regional
population of approximately 297,000 residents. The South Division consists of
about 51 miles of natural gas transmission main and 3,745 miles of distribution
pipelines. Avista receives natural gas at more than 20 points along interstate
pipelines and distributes it to almost 99,000 customers.
156,000
79,000 99,000
334,000
Wa
s
h
i
n
g
t
o
n
Id
a
h
o
Or
e
g
o
n
To
t
a
l
Total Natural Gas Customers
as of Dec. 31, 2015
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 19 of 162
Customers
Avista provides natural gas services to core and transportation-only customer classes.
Core or retail customers purchase natural gas directly from Avista with delivery to their
home or business under a bundled rate. Core customers on firm rate schedules are
entitled to receive any volume of natural gas they require. Some core customers are on
interruptible rate schedules. These customers pay a lower rate than firm customers
because their service can be interrupted. Interruptible customers are not considered in
peak day IRP planning.
Transportation-only customers purchase natural gas from third parties who deliver the
purchased gas to our distribution system. Avista delivers this natural gas to their business
charging a distribution rate only. Avista can interrupt the delivery service when following
the priority of service tariff. The long-term resource planning exercise excludes
transportation-only customers because they purchase their own gas and utilize their own
interstate pipeline transportation contracts. However, distribution planning exercises
include these customers.
Avista’s core or retail customers include residential, commercial and industrial categories.
Most of Avista’s customers are residential, followed by commercial and relatively few
industrial accounts (Figure 1.3).

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 20 of 162
Figure 1.3: Firm Customer Mix

The customer mix is more balanced between residential and commercial accounts on an
annual volume basis (Figure 1.4). Volume consumed by core industrial customers is not
significant to the total, partly because most industrial customers in Avista’s service
territories are transportation-only customers.

89.99%
9.91%0.10%
Customer Percentages
WA-ID
88.44%
11.54%0.03%
Customer Percentages
Oregon
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 21 of 162
Figure 1.4 Therms by Class

Core customer demand is seasonal, especially residential accounts in Avista’s service
territories with colder winters (Figure 1.5). Industrial demand, which is typically not
weather sensitive, has very little seasonality. However, the La Grande service territory
has several industrially classified agricultural processing facilities that produce a late
summer seasonal demand spike.

62.9%
35.7%
1.4%
Annual Demand
WA-ID
64.2%
35.5%
0.3%
Annual Demand
Oregon
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 22 of 162
Figure 1.5: Customer Demand by Service Territory

Integrated Resource Planning
Avista’s IRP involves a comprehensive analytical process to ensure that core firm
customers receive long-term reliable natural gas service at a reasonable price. The IRP
evaluates, identifies, and plans for the acquisition of an optimal combination of existing
and future resources using expected costs and associated risks to meet average daily
and peak-day demand delivery requirements over a 20-year planning horizon.

Purpose of the IRP
Avista’s 2016 Natural Gas IRP:
 Provides a comprehensive long-range planning tool;
 Fully integrates forecasted requirements with existing and potential resources;
 Determines the most cost-effective, risk-adjusted means for meeting future
demand requirements; and
 Meets Washington, Idaho and Oregon regulations, commission orders, and other
applicable guidelines.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 23 of 162
Avista’s IRP Process
The natural gas IRP process considers:
 Customer growth and usage:
 Weather planning standard;
 Conservation opportunities;
 Existing and potential supply-side resource options;
 Current and potential legislation/regulation;
 Risk; and
 Least cost mix of supply and conservation.

Public Participation
Avista’s TAC members play a key role and have a significant impact in developing the
IRP. TAC members include Commission Staff, peer utilities, public interest groups,
customers, academics, government agencies, and other interested parties. TAC
members provide important input on modeling, planning assumptions, and the general
direction of the process.
Avista sponsored four TAC meetings to facilitate stakeholder involvement in the 2016
IRP. The first meeting convened on January 21, 2016, and the last meeting occurred on
April 21, 2016. Meetings are at a variety of locations convenient for stakeholders and are
electronically available for those unable to attend in person. Each meeting included a
broad spectrum of stakeholders. The meetings focused on specific planning topics,
reviewing the progress of planning activities, and soliciting input on IRP development.
TAC members received a draft of this IRP in late May 2016 for their review. Avista
appreciates all of the time and effort TAC members gave to the IRP process; they
provided valuable input through their participation in the TAC process. A list of these
organizations can be found below (Table 1.1).

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 24 of 162
Table 1.1: TAC Member Participation
Cascade Natural Gas Northwest Industrial Gas
Users
Oregon Public Utility
Commission
Fortis Northwest Natural Gas Puget Sound Energy
Idaho Public Utilities
Commission
Williams - Northwest
Pipeline
TransCanada
Northwest Gas Association Citizens Utility Board of
Oregon
Washington Utilities and
Transportation
Commission

Preparation of the IRP is a coordinated endeavor by several departments within Avista
with involvement and guidance from management. We are grateful for their efforts and
contributions.

Regulatory Requirements Avista submits a natural gas IRP to the public utility commissions in Idaho, Oregon and
Washington on or before August 31 every two years as required by state regulation. There
is a statutory obligation to provide reliable natural gas service to customers at rates, terms
and conditions that are fair, just, reasonable and sufficient. Avista regards the IRP as a
means for identifying and evaluating potential resource options and as a process to
establish an Action Plan for resource decisions. Ongoing investigation, analysis and
research may cause Avista to determine that alternative resources are more cost effective
than resources reviewed and selected in this IRP. Avista will continue to review and refine
our understanding of resource options and will act to secure these risk-adjusted, least-
cost options when appropriate.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 25 of 162
Planning Model
Consistent with prior IRPs, Avista used the SENDOUT planning model to perform
comprehensive natural gas supply planning and analysis for this IRP. SENDOUT is a
linear programming-based model that is widely used to solve natural gas supply, storage
and transportation optimization problems. This model uses present value revenue
requirement (PVRR) methodology to perform least-cost optimization based on daily,
monthly, seasonal and annual assumptions related to the following:

 Customer growth and customer natural gas usage to form demand forecasts;
 Existing and potential transportation and storage options and associated costs;
 Existing and potential natural gas supply availability and pricing;
 Revenue requirements on all new asset additions;
 Weather assumptions; and
 Conservation.

Avista incorporated stochastic modeling by utilizing a SENDOUT module to simulate
weather and price uncertainty. The module generates Monte Carlo weather and price
simulations, running concurrently to account for events and to provide a probability
distribution of results that aid resource decisions. Some examples of the types of
stochastic analysis provided include:

 Price and weather probability distributions;
 Probability distributions of costs (i.e. system costs, storage costs, commodity
costs); and
 Resource mix (optimally sizing a contract or asset level of competing resources).

These computer-based planning tools were used to develop the 20-year best cost/risk
resource portfolio plan to serve customers.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 26 of 162
Planning Environment
Even though Avista publishes an IRP biannually, the process is ongoing with new
information and industry related developments. In normal circumstances, the process can
become complex as underlying assumptions evolve, impacting previously completed
analyses. Every planning cycle has challenges and uncertainties; this cycle was no
different. Widespread agreement on the availability of shale gas and the ability to produce
it at lower prices has increased interest in the use of natural gas for LNG exports and for
more transportation and industrial uses. However, there is uncertainty about the timing
and size of those markets.

IRP Planning Strategy
Planning for an uncertain future requires robust analysis that encompasses a wide range
of possibilities. Avista has determined that the planning approach needs to:

 Recognize that historical trends may be fundamentally altered:
 Critically review all assumptions;
 Stress test assumptions via sensitivity analysis;
 Pursue a spectrum of possible scenarios;
 Develop a flexible analytical framework to accommodate changes; and
 Maintain a long-term perspective.

With these objectives in mind, Avista developed a strategy encompassing all required
planning criteria. This produced a complete IRP that effectively analyzes risks and
resource options, which sufficiently ensures customers will receive safe and reliable
energy delivery services with the best-risk, lease-cost, long-term solutions. The following
chart summarizes significant changes from the 2014 IRP (Table 1.2).

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 27 of 162
Table 1.2: Summary of changes from the 2014 IRP
Chapter Issue 2016 Natural Gas IRP 2014 Natural Gas IRP
Demand Expected
Customer Growth
Expected Case customer
growth is 1.1% compounded
annually.
Expected Case customer
growth is 1% compounded
annually.
Environmental
Issues
Carbon Dioxide
Emission (Carbon)
Three sensitivities on level of
carbon tax ($/ton) were
compared. The expected case
has a probability of 2 sigma of
the likely policy. The
remainder of probability
equally assumed to Low and Washington State’s I-732 were
used to represent the tails in a
normal distribution. The base
carbon case is the expected
case. The high and low cases
help bracket the base case
results.
Three sensitivities on level of
carbon tax ($/ton) were
compared. The base carbon
case is the medium case.
The high and low cases help
bracket the base case
results.
Prices Price Curve Lower Price curve can drive
the conservation potential-
downward.
A higher price curve with
slightly higher conservation
potential.
Supply Side
Resources
Supply Side Scenarios The only case that identifies a resource deficiency is the High
Growth/Low Price scenario.
Avista solved this case by
using existing resources plus
added contracted capacity on
GTN for WA/ID. In Klamath
Falls, Medford and Roseburg
an upsized compressor would
be added on the Medford
lateral.
Evaluated three supply side scenarios on cases with
resource deficiencies.
Existing resources, Existing
plus Expected Available, and
GTN fully subscribed. Did
not solve for high growth/low
price unserved

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 28 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 29 of 162
2: Demand Forecasts

Overview
The integrated resource planning process begins with the development of forecasted
demand. Understanding and analyzing key demand drivers and their potential impact on
forecasts is vital to the planning process. Utilization of historical data provides a reliable
baseline, however past trends may not be indicative of future trends. This IRP mitigates
the uncertainty by considering a range of scenarios to evaluate and prepare for a broad
spectrum of outcomes.

Demand Areas
Avista defined eight demand areas, structured around the pipeline transportation
resources that serve them, within the SENDOUT model (Table 2.1). These demand
areas are aggregated into four service territories and further summarized as North or
South divisions for presentation throughout this IRP.

Table 2.1 Geographic Demand Classifications

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 30 of 162
Demand Forecast Methodology
Avista uses the IRP process to develop two types of demand forecasts – annual and peak
day. Annual average demand forecasts are useful for preparing revenue budgets,
developing natural gas procurement plans, and preparing purchased gas adjustment
filings. Peak day demand forecasts are critical for determining the adequacy of existing
resources or the timing for acquiring new resources to meet customers’ natural gas needs
in extreme weather conditions.
In general, if existing resources are sufficient to meet peak day demand, they will be
sufficient to meet annual average day demand. Developing annual average demand first
and evaluating it against existing resources is an important step in understanding the
performance of the portfolio under normal circumstances. It also facilitates
synchronization of modeling processes and assumptions for planning purposes.
Peak weather analysis aids in assessing resource adequacy and any differences in
resource utilization. For example, storage may be dispatched differently under peak
weather scenarios.

Demand Modeling Equation
Developing daily demand forecasts is essential because natural gas demand can vary
widely from day-to-day, especially in winter months when heating demand is at its highest.
In its most basic form, natural gas demand is a function of customer base usage (non-
weather sensitive usage) plus customer weather sensitive usage. Basic demand takes
the formula in Table 2.2:

Table 2.2: Basic Demand Formula

SENDOUT® requires inputs as expressed in the Table 2.3 format to compute daily
demand in dekatherms.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 31 of 162
Table 2.3: SENDOUT® Demand Formula

SENDOUT performs this calculation daily for each customer class and each demand
area. The base and weather sensitive usage (heating degree day usage) factors use
customer demand coefficients developed outside the SENDOUT model, and the
coefficients capture a variety of demand usage assumptions. This is discussed in more
detail in the Use-per-Customer Forecast section below. The number of daily degree days
is simply heating degree days (HDDs), which are discussed in the Weather Forecast
section later in this chapter.

Customer Forecasts
Avista’s customer base includes residential, commercial and industrial categories. For
each of the customer categories, Avista develops customer forecasts incorporating
national economic forecasts and then drilling down into regional economies. U.S. GDP
growth, national and regional employment growth, and regional population growth
expectations are key drivers in regional economic forecasts and are useful in estimating
the number of natural gas customers. A detailed description of the customer forecast is
found in Appendix 2.1 – Economic Outlook and Customer Count Forecast. Avista
combines this data with local knowledge about sub-regional construction activity, age and
other demographic trends, and historical data to develop the 20-year customer forecasts.
Several Avista departments use these forecasts, but Finance, Accounting, Rates, and
Gas Supply are the primary users. The natural gas distribution engineering group utilizes
the forecast data for system optimization and planning purposes (see discussion in
Chapter 7 – Distribution Planning).
Forecasting customer growth is an inexact science, so it is important to consider different
forecasts. Two alternative growth forecasts were developed for this IRP. Avista developed
High and Low Growth forecasts to provide potential paths and test resource adequacy.
Appendix 2.1 contains a description of how these alternatives were developed.
Figure 2.1 shows the three customer growth forecasts. Due to a change in forecasting
methodology for customer growth, the expected case customer counts are lower than the
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 32 of 162
last IRP. This has impacted forecasted demand from both the average and peak day
perspective. Detailed customer count data by region and class for all three scenarios is
in Appendix. 2.2 – Customer Forecasts by Region.

Figure 2.1: Customer Growth Scenarios

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 33 of 162
Use-per-Customer Forecast
The goal for a use-per-customer forecast is to develop base and weather sensitive
demand coefficients that can be combined and applied to HDD weather parameters to
reflect average use-per-customer. This produces a reliable forecast because of the high
correlation between usage and temperature as depicted in the example scatter plot in
Figure 2.2.

Figure 2.2: Example Demand vs. Average Temperature – WA/ID

The first step in developing demand coefficients was gathering daily historical gas flow
data for all of Avista’s city gates. The use of city gate data over revenue data is due to
the tight correlation between weather and demand. The revenue system does not capture
data on a daily basis and, therefore, makes a statistical analysis with tight correlations on
a daily basis virtually impossible. Avista reconciles city gate flow data to revenue data to
ensure that total demand is properly captured.
The historical city gate data was gathered, sorted by service territory/temperature zone,
and then by month. As in the last IRP, Avista used three years of historical data to derive
the use-per-customer coefficients, but also considered varying the number of years of
historical data. When comparing five years of historical use-per-customer to three years
of data, the five-year data had slightly higher use-per-customer, which may overstate use
0
50,000
100,000
150,000
200,000
250,000
-20020406080100
Dth
Farenheit
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 34 of 162
as efficiency and use-per-customer-per-HDD have been on a downward trend since
2006. Three years struck a balance between historical and current customer usage
patterns. Figure 2.3 illustrates the annual demand differences between the three and five-
year use-per-customer with normal and peak weather conditions.

Figure 2.3: Annual Demand – Demand Sensitivities 3-Year vs. 5-Year Use-per-Customer

The base usage calculation used three years of July and August data to derive
coefficients. Average usage in these months divided by the average number of customers
provides the base usage coefficient input into SENDOUT. This calculation is done for
each area and customer class based on customer billing data demand ratios.
To derive weather sensitive demand coefficients for each monthly data subset, Avista
removed base demand from the total and plotted usage by HDD in a scatter plot chart to
verify correlation visually. The process included the application of a linear regression to
the data by month to capture the linear relationship of usage to HDD. The slopes of the
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 35 of 162
resulting lines are the monthly weather sensitive demand coefficients input into
SENDOUT. Again, this calculation is done by area and by customer class using
allocations based on customer billing data demand ratios.

Weather Forecast
The last input in the demand modeling equation is weather (specifically HDDs). This
started with the most current 20 years of daily weather data from the National Oceanic
Atmospheric Administration (NOAA), converted to HDDs, and is used to compute an
average for each day to develop the weather forecast. The Oregon weather input used
four weather stations, corresponding to the areas where Avista provides natural gas
services. HDD weather patterns between these areas are uncorrelated. Weather data for
the Spokane Airport is used for the eastern Washington and northern Idaho portions of
the service area, as HDD weather patterns within that region are correlated.
The NOAA 20-year average weather serves as the base weather forecast to prepare the
annual average demand forecast. The peak day demand forecast includes adjustments
to average weather to reflect a five-day cold weather event. This consists of adjusting the
middle day of the five-day cold weather event to the coldest temperature on record for a
service territory, as well as adjusting the two days on either side of the coldest day to
temperatures slightly warmer than the coldest day. For the Washington/Idaho and La
Grande service territories, the model assumes this event on and around February 15 each
year. For the southwestern Oregon service territories (Medford, Roseburg, Klamath
Falls), the model assumes this event on and around December 20 each year. The
following section provides details about the coldest days on record for each service
territory.
The Washington/Idaho service areas coldest day on record was an 82 HDD for Spokane
and occurred on Dec. 30, 1968. This is equal to an average daily temperature of -17
degrees Fahrenheit. Only one 82 HDD has been experienced in the last 48 years for this
area; however, within that same time period, 80, 79 and 74 HDD events occurred on Dec.
29, 1968, Dec. 31,1978 and Jan. 5, 2004, respectively.
Medford experienced the coldest day on record, a 61 HDD, on Dec. 9, 1972. This is equal
to an average daily temperature of 4 degrees Fahrenheit. Medford has experienced only
one 61 HDD in the last 43 years; however, it has also experienced 59 and 58 HDD events
on Dec. 8, 1972 and Dec. 21, 1990, respectively.
The other three areas in Oregon have similar weather data. For Klamath Falls, a 72 HDD
occurred on Dec. 8, 2013; in La Grande a 74 HDD occurred on Dec. 23, 1983; and a 55
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 36 of 162
HDD occurred in Roseburg on Dec. 22, 1990. As with Washington/Idaho and Medford,
these days are the peak day weather standard for modeling purposes.
Utilizing a peak planning standard of the coldest temperature on record may seem
aggressive given a temperature experienced rarely, or only once. Given the potential
impacts of an extreme weather event on customers’ personal safety and property damage
to customer appliances and Avista’s infrastructure, it is a prudent regionally accepted
planning standard. While remote, peak days do occur, as on Dec. 8, 2013, when Avista
matched the previous peak HDD in Klamath Falls.
Avista analyzes an alternate planning standard using the coldest temperature in the last
twenty years. The Washington/Idaho service area uses a 76 HDD, which is equal to an
average daily temperature of -11 degrees Fahrenheit. In Medford, the coldest day in 20
years is a 54 HDD, equivalent to an average daily temperature of 11 degrees Fahrenheit.
In Roseburg, the coldest day in 20 years is a 48 HDD, equivalent to an average daily
temperature of 17 degrees Fahrenheit. In Klamath Falls, the coldest day in 20 years is a
72 HDD, equivalent to an average daily temperature of -7 degree Fahrenheit. In La
Grande, the coldest day in 20 years is a 74 HDD, equivalent to an average daily
temperature of -9 degree Fahrenheit. The HDDs by area, class and day entered into
SENDOUT® are in Appendix 2.4 – Heating Degree Day Data.

Developing a Reference Case
To adjust for uncertainty, Avista developed a dynamic demand forecasting methodology
that is flexible to changing assumptions. To understand how various alternative
assumptions influence forecasted demand Avista needed a reference point for
comparative analysis. For this, Avista defined the reference case demand forecast shown
in Figure 2.4. This case is only a starting point to compare other cases.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 37 of 162
Figure 2.4: Reference Case Assumptions
1. Customer Compound Annual Growth Rates
Area Residential Commercial Industrial
Washington/ Idaho 1.1% 0.6% 0.0%
Klamath Falls 1.3% 0.9% 0.0%
La Grande 0.6% 0.4% 0.1%
Medford 1.3% 1.0% 0.0%
Roseburg 1.1% 0.2% 0.0%

2. Use-Per-Customer Coefficients
Flat Across All Classes 3-year Average Use per Customer per HDD by Area/Class
3. Weather
20-year Normal – NOAA (1996-2015)
4. Elasticity
None
5. Conservation
None

Dynamic Demand Methodology

The dynamic demand planning strategy examines a range of potential outcomes. The
approach consists of:

 Identifying key demand drivers behind natural gas consumption;
 Performing sensitivity analysis on each demand driver;
 Combining demand drivers under various scenarios to develop alternative
potential outcomes for forecasted demand; and
 Matching demand scenarios with supply scenarios to identify unserved demand.

Figure 2.5 represents Avista’s methodology of starting with sensitivities, progressing to
portfolios, and ultimately selecting a preferred portfolio.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 38 of 162
Figure 2.5: Sensitivities and Preferred Portfolio Selection

Sensitivity Analysis
In analyzing demand drivers, Avista grouped them into two categories based on:
 Demand Influencing Factors directly influencing the volume of natural gas
consumed by core customers.
 Price Influencing Factors indirectly influencing the volume of natural gas consumed
by core customers through a price elasticity response.
After identifying demand and price influencing factors, Avista developed sensitivities to
focus on the analysis of a specific natural gas demand driver and its impact on forecasted
demand relative to the Reference Case when modifying the underlying input
assumptions.
Sensitivity assumptions reflect incremental adjustments not captured in the underlying
Reference Case forecast. Avista analyzed 18 demand sensitivities to determine the
results relative to the Reference Case. Table 2.4 lists these sensitivities. Detailed
information about these sensitivities is in Appendix 2.6 – Demand Forecast Sensitivities
and Scenarios Descriptions.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 39 of 162
Table 2.4: Demand Sensitivities

Figure 2.6 shows the annual demand from each of the sensitivities modeled for this IRP.

Figure 2.6: 2016 IRP Demand Sensitivities

Scenario Influence Weather Growth Use per Customer Price Curve Carbon Adder LNG Adder DSM New Markets Elasticity
Reference Case Direct Normal Expected 3 year Expected No No No No No
Reference Case plus Peak
Weather Direct Peak Expected 3 year Expected No No No No No
High Growth Case Direct Peak High 3 year Expected No No No No No
Low Growth Case Direct Peak Low 3 year Expected No No No No No
Alternate Use per Customer Direct Peak Expected 5 year Expected No No No No No
CNG/NGV Case Direct Peak Expected 3 year Expected No No No Yes No
DSM Direct Normal Expected 3 year Expected No No No No No
Peak plus DSM Direct Peak Expected 3 year Expected No No Yes No No
Alternate Weather Planning Standard Direct Coldest in 20 Expected 3 year Expected No No Yes No No
Expected Elasticity Indirect Peak Expected 3 year Expected No No No No Yes
Low Price Indirect Peak Expected 3 year Low No No No No No
High Price Indirect Peak Expected 3 year High No No No No No
Carbon Legislation - Expected Indirect Peak Expected 3 year Expected Yes No No No No
Carbon Legislation - Low Indirect Peak Expected 3 year Expected Yes No No No No
Carbon Legislation - High Indirect Peak Expected 3 year Expected Yes No No No No
Exported LNG Indirect Peak Expected 3 year Expected No Yes No No No
Table 3.4 - Demand Sensitivities
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 40 of 162
Scenario Analysis
After testing the sensitivities, Avista grouped them into meaningful combinations of
demand drivers to develop demand forecasts representing scenarios. Table 2.5 identifies
the scenarios developed for this IRP. The Average Case represents the case used for
normal planning purposes, such as corporate budgeting, procurement planning, and
PGA/General Rate Cases. The Expected Case reflects the demand forecast Avista
believes is most likely given peak weather conditions. The Expected Case/Low Price
represents the Expected Case assumptions combined with the low price curve. The High
Growth/Low Price and Low Growth/High Price cases represent a range of possibilities for
customer growth and future prices. The Alternate Weather Standard case utilizes the
coldest day in Avista’s service territories in the last 20 years. Each of these scenarios
provides a “what if” analysis given the volatile nature of key assumptions, including
weather and price. Appendix 2.6 lists the specific assumptions within the scenarios while
Appendix 2.7 contains a detailed description of each scenario.

Table 2.5: Demand Scenarios
2016 IRP Demand Scenarios
Average Case
Expected Case
High Growth, Low Price
Low Growth, High Price
Alternate Weather Standard
Expected Case, Low Price

Price Elasticity
The economic theory of price elasticity states that the quantity demanded for a good or
service will change with its price. Price elasticity is a numerical factor that identifies the
relationship of a consumer’s consumption change in response to a price change.
Typically, the factor is a negative number as consumers normally reduce their
consumption in response to higher prices or will increase their consumption in response
to lower prices. For example, a price elasticity factor of negative 0.15 for a particular good
or service means a 10 percent price increase will prompt a 1.5 percent consumption
decrease and a 10 percent price decrease will prompt a 1.5 percent consumption
increase.
Complex relationships influence price elasticity and given the current economic
environment, Avista questions whether current behavior will become normal or if
customers will return to historic usage patterns. Furthermore, complex regulatory pricing
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 41 of 162
mechanisms shield customers from price volatility, thereby dampening price signals and
affecting price elastic responses. For example, budget billing averages a customer’s bills
into equal payments throughout the year. This popular program helps customers manage
household budgets, but does not send a timely price signal. Additionally, natural gas cost
adjustments, such as the Purchased Gas Adjustment (PGA), annually adjusts the
commodity cost which shields customers from daily gas price volatility. These
mechanisms do not completely remove price signals, but they can significantly dampen
the potential demand impact.
When considering a variety of studies on energy price elasticity, a range of potential
outcomes was identified, including the existence of positive price elastic adjustments to
demand. One study looking at the regional differences in price elasticity of demand for
energy found that the statistical significance of price becomes more uncertain as the
geographic area of measurement shrinks.1 This is particularly important given Avista’s
geographically diverse and relatively small service territories.
Avista acknowledges changing price levels can and do influence natural gas usage, so
this IRP includes a price elasticity of demand factor of -0.15 into the modeling
assumptions to allow use-per-customer to vary as the natural gas price forecast changes.
Recent usage data indicates that even with declines in the retail rate for natural gas, long
run use-per-customer continues to decline. This likely includes a confluence of factors
including high regional unemployment, increased investments in energy efficiency
measures, building code improvements, behavioral changes, and heightened focus of
consumers’ household budgets.

Results
During 2016, the Average Case demand forecast indicates Avista will serve an average
of 334,000 core natural gas customers with 33,219,431 Dth of natural gas. By 2035,
Avista projects 412,000 core natural gas customers with an annual demand of over
36,154,721 Dth. In Washington/Idaho, the projected number of customers increases at
an average annual rate of 1.10 percent, with demand growing at a compounded average
annual rate of 0.36 percent. In Oregon, the projected number of customers increases at
an average annual rate of 1.18 percent, with demand growing 0.70 percent per year.
During 2016, the Expected Case demand forecast indicates Avista will serve an average
of 334,000 core natural gas customers with 34,369,993 Dth of natural gas. By 2035,
1 Bernstein, M.A. and J. Griffin (2005). Regional Differences in Price-Elasticity of Demand for Energy,
Rand Corporation.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 42 of 162
Avista projects 412,000 core natural gas customers with an annual demand of 37,536,603
Dth.
Figure 2.7 shows system forecasted demand for the demand scenarios on an average
daily basis for each year.2
Figure 2.7: Average Daily Demand – 2016 IRP Scenarios

2 Appendix 2.1 shows gross demand, conservation savings and net demand.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 43 of 162
Figure 2.8 shows system forecasted demand for the Expected, High and Low Demand
cases on a peak day basis for each year relative to the Average Case average daily winter
demand. Detailed data for all demand scenarios is in Appendix 2.8 – Demand Before and
After DSM.

Figure 2.8: February 15th – Peak Day – 2016 IRP Demand Scenarios

The IRP balances forecasted demand with existing and new supply alternatives. Since
new supply sources include conservation resources, which act as a demand reduction,
the demand forecasts prepared and described in this section include existing efficiency
standards and normal market acceptance levels. The methodology for modeling DSM
initiatives is in Chapter 3 – Demand-Side Resources.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 44 of 162
Alternative Forecasting Methodologies
There are many forecasting methods available and used throughout different industries.
Avista uses methods that enhance forecast accuracy, facilitate meaningful variance
analysis, and allows for modeling flexibility to incorporate different assumptions. Avista
believes the IRP statistical methodology to be sound and provides a robust range of
demand considerations. The methodology allows for the analysis of different statistical
inputs by considering both qualitative and quantitative factors. These factors come from
data, surveys of market information, fundamental forecasts, and industry experts. Avista
is always open to new methods of forecasting natural gas demand and will continue to
assess which, if any, alternative methodologies to include in the dynamic demand
forecasting methodology.

Key Issues
Demand forecasting is a critical component of the IRP requiring careful evaluation of the
current methodology and use of scenario planning to understand how changes to the
underlying assumptions will affect the results. The evolution of demand forecasting over
recent years has been dramatic, causing a heightened focus on variance analysis and
trend monitoring. Current techniques have provided sound forecasts with appropriate
variance capabilities. However, Avista is mindful of the importance of the assumptions
driving current forecasts and understands that these can and will change over time.
Therefore, monitoring key assumptions driving the demand forecast is an ongoing effort
that will be shared with the TAC as they develop.

Flat Demand Risk
Forecasting customer usage is a complex process because of the number of underlying
assumptions and the relative uncertainty of future patterns of usage with a goal of
increasing forecast accuracy. There are many factors that can be incorporated into these
models, assessing which ones are significant and improving the accuracy are key. Avista
continues to evaluate economic and non-economic drivers to determine which factors
improve forecasting accuracy. The forecasting process will continue to review research
on climate change and the best way to incorporate the results of that research into the
forecasting process.
For the last few planning cycles, the TAC has discussed the changing slope of forecasted
demand. Growth has slowed due to the recent recession and declining use-per-customer.
Use-per-customer seems to have stabilized, but customer growth in Avista’s service
territory may not return to pre-recession levels.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 45 of 162
This reduced demand pushes the need for resources beyond the planning horizon, which
means no new investment in resources is necessary. However, should assumptions
about lower customer growth prove to be inaccurate and there is a rebound in demand,
new resource needs will occur sooner than expected. Therefore, careful monitoring of
demand trends in order to identify signposts of accelerated demand growth is critical to
the identification of new resource needs coming earlier than expected.
Action Plan
 Monitor actual demand for accelerated growth can address resource deficiencies
arising from exposure to “flat demand” risk. This will include providing Commission
Staff with IRP demand forecast-to-actual variance analysis on customer growth
and use per customer at least bi-annually.

Emerging Natural Gas Demand
The shale gas revolution has fundamentally changed the long-term availability and price
of natural gas. This revolution has prompted an evolution in the increased use of natural
gas. An ever growing demand for natural gas-fired generation to integrate variable wind
and solar resources along with an increasing demand from coal retirements and fuel
switching has developed over the last few years. This demand is expected to increase
due to the availability of natural gas combined with its lower carbon emissions. Other
areas of emerging demand include everything from methanol plants to food processors,
and interest in industrial processes using natural gas as a feedstock is growing.

Conclusion
Recessionary impacts have significantly reduced Avista’s outlook for customer growth
and reduced the long-term demand forecasts. Avista’s dynamic demand methodology
provides a means to assess the individual and collective demand impact of a variety of
economic and non-economic drivers. The results of this comprehensive analysis provides
a better understanding of the possible outcomes with respect to core consumption of
natural gas and helps drive resource decisions based on changing consumer needs.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 46 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 47 of 162
3: Demand-Side Resources
Overview
Avista is committed to offering natural gas DSM portfolios to residential, commercial and
industrial customers when it is feasible to do so in a cost-effective manner as prescribed within
each jurisdiction. Improved drilling and extraction techniques of natural gas has led to declines
in natural gas prices in recent years which has made offering cost-effective DSM programs
challenging. In May 2012, Avista proposed to suspend its Washington and Idaho natural gas
DSM programs due to decreased natural gas prices. The WUTC guided utilities to continue
natural gas programs using the Utility Cost Test (UCT), Oregon provided cost-effectiveness
exemptions for certain measures using the Total Resource Cost Test (TRC) and Avista
requested and was given approval to suspend Avista’s Idaho natural gas DSM programs
under the TRC. During 2015, the Avista DSM group reviewed the current composition and
components of natural gas avoided costs and compared them with both other regional and
national utilities. The research and proposed additions to Avista’s avoided cost were
presented to Avista’s DSM Advisory Group for feedback on August 19th and August 20th,
2015 to ensure these were appropriate changes and to seek advice about other future
avoided cost components analyses the company should perform. After the review of Avista’s
avoided cost methodology and with an IPUC ruling that allows companies to emphasize the
UCT when seeking prudence for their DSM programs, Avista filed for and was approved to
reinstate its Idaho Natural Gas DSM programs as of January 1, 2016.
As part of the settlement for the Avista 2015 Oregon General Rate case, the company will
begin to transition the regular income DSM programs to the Energy Trust of Oregon (ETO) in
2016, with the full program delivery and administration from the ETO January 1st, 2017. Avista
will continue to administer the Oregon Low-Income conservation programs, which are not
offered through the ETO.
During 2015, the proposed and accepted changes to avoided cost methodology were to
include the additional costs associated with bringing natural gas from the wellhead to the
customer meter beyond the firm transmission variable costs. Avista contracts for enough
natural gas pipeline transportation to provide firm transportation capacity for a peak day. A
large majority of the transaction costs are in reserving the capacity with a small component
for the amount of natural gas that actually flows at a given time. Only including the variable
portion of the transportation contracts does not accurately represent all of the costs of
transporting natural gas from the wellhead to the customer meter and when excess capacity
is available, the company releases or optimizes excess capacity for the direct benefit of
customers. Transportation costs are built into the demand portion of Washington and Idaho
schedule 150 which is a variable cost to customers that encompasses the net fixed costs to
Avista to deliver natural gas to customer’s meters. In Washington, a $10/ton carbon cost
starting 2020 was included to account for the potential carbon reduction approaches currently
occurring in the state.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 48 of 162
The company has also been working to quantify the deferred distribution capacity benefits
from natural gas conservation. Natural gas measures have two primary savings distribution:
level annual therms and heating driven therm savings. For level annual therms for every 365
therms saved approximately one therm of peak load is reduced and the equivalent figure for
space heating driven therms is 140 therms (~Average Day Dth/Peak Day Dth * 365) and on
average a customer uses about 11.5 therms on a peak day. Using these figures, Avista DSM
programs in Washington and Idaho since 2001 have offset the peak load of approximately
8,380 customers. The company is looking to quantifying the value of a peak day therm
reduction and has a couple of potential approaches. The first is utilizing the cost per service
as a value for the equivalent offset of capital expenditures. This approach benefits from data
that is currently tracked and reported, but might not be the most accurate approach. The
second approach looks at the costs associated with the recent and future reinforcement and
capacity upgrade projects and calculating either the increased system capacity or the number
of potential low pressure customers avoided under the design conditions due to the capacity
upgrade. This approach would provide a more direct correlation, however the company does
not currently track the upgraded system capacity or avoided low pressure customers and an
exact calculation might be difficult due to the interconnected nature of the natural gas
distribution system. Since the benefits of deferred capacity are a one-time cost, the benefits
would be spread over the life of a typical distribution upgrade (35 years) as an avoided
payment. This component of avoided costs was not included in this IRP and will be discussed
with Avista’s DSM Advisory Group as well as the TAC to determine the correct approach for
the 2018 IRP.
Conservation Potential Assessment Methodology Overview
Avista issued an RFP and Applied Energy Group (AEG) was chosen to perform an external
independent evaluation of the technical, economic and achievable conservation potential
within each of Avista’s three jurisdictions over a 20-year planning horizon. This process
involves indexing existing nationally recognized Conservation Potential Assessment (CPA)
models to the Avista service territory load forecast, housing stock, end-use saturations and
other key characteristics. Additional consideration of the impact of energy codes and
appliance standards for end-use equipment at both the state and national level are
incorporated into the projection of energy use and the baseline for the evaluation of efficiency
options. The modeling process also utilizes ramp rates for the acquisition of efficiency
resources over time in a manner generally consistent with the assumptions used by the
Northwest Power and Conservation Council (NPCC).
The process described above defines an Avista-specific supply curve for conservation
resources. Simultaneously, the avoided cost of natural gas consistent with serving the full
forecasted demand was defined as part of the SENDOUT® modeling of the Avista system.
The preliminary cost-effective conservation potential is determined by applying the stream of
annual natural gas avoided costs to the Avista-specific supply curve for conservation
resources. This quantity of conservation acquisition is then decremented from the load which
the utility must serve and the SENDOUT® model is rerun against the modified (reduced) load
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 49 of 162
requirements. The resulting avoided costs are compared to those obtained from the previous
iteration of SENDOUT® avoided costs. This process continues until the differential between
the avoided cost streams of the most recent and the immediately previous iteration becomes
immaterial. The cost-effectiveness test used for Washington and Idaho was the UCT, and
Oregon continued to utilize the TRC to determine conservation selection.
Integrating the DSM portfolio into the IRP process by equilibrating the avoided costs in this
iterative process is useful since Avista’s DSM acquisition is small relative to the total western
natural gas market used to establish the commodity prices driving the avoided cost stream.
Therefore, few iterations are necessary to reach a stable avoided cost. Additionally, it provides
some assurance, at least at the aggregate level, that the quantity of DSM resource selected
will be cost-effective when the final avoided cost stream is used in retrospective portfolio
evaluation.
Conservation Potential Assessment Findings
Prior to the development of potential conservation estimates, AEG created a baseline end-
use forecast to quantify the use of natural gas by end use in the base year, and projections
of consumption in the future in the absence of utility programs and naturally occurring
conservation. The end-use forecast includes the relatively certain impacts of codes and
standards that will unfold over the study timeframe. All such mandates defined as of January
2015 are included in the baseline. The baseline forecast is the foundation for the analysis of
savings from future DSM programs as well as the metric against which potential savings are
measured.
Inputs to the baseline forecast include current economic growth forecasts (e.g. customer
growth and income growth), natural gas price forecasts, trends in fuel shares and equipment
saturations developed by AEG, existing and approved changes to building codes and
equipment standards, and Avista’s internally developed load forecast.
According to the CPA, the residential sector natural gas consumption for all end uses and
technologies increases primarily due to the projected 1.1 percent annual growth in the number
of households and the slight increase in the average home size. Other heating, which includes
unit wall heaters and miscellaneous loads, have a relatively high growth rate compared to
other loads. However, at the end of the 20-year planning period these loads represent only a
small part of overall use.
For the commercial and industrial sectors, natural gas use continues to grow slowly over the
20-year planning horizon as new construction increases the overall square footage in this
sector. Growth in heating, ventilation and air conditioning (HVAC) and water heating end uses
is moderate. Food preparation, though a small percentage of total usage, grows at a higher
rate than other end uses. Consumption by miscellaneous equipment and process heating are
also projected to increase.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 50 of 162
Table 3.1 illustrates the baseline consumption broken out by state and sector for selected
years over the 20-year planning horizon. The overall baseline consumption is expected to
increase 21 percent over the 20-year planning horizon corresponding to an annualized growth
of 1.1 percent. The forecast projects steady growth over the next 20 years with growth in the
residential sector making up for the flat sales in the industrial sector. Idaho is projected to
experience the highest level of growth with Washington having the next highest level of
growth.

Table 3.1: Baseline Forecast Summary (Dth)
Sector 2015 2017 2018 2021 2026 2036 %
Change
(’17-
’36)
Avg.
Growth
Rate
(’17-’36)
Residential 17,796,844 19,617,372 19,846,006 20,458,537 21,702,908 24,462,944 25% 1.1%
Commercial 10,890,632 12,092,364 12,119,571 12,265,926 12,651,011 13,816,504 14% 0.7%
Industrial 507,024 546,799 556,102 582,284 632,932 748,791 37% 1.6%
Total 29,194,500 32,256,535 32,521,679 33,306,747 34,986,852 39,028,239 21% 1.0%

State 2015 2017 2018 2021 2026 2036 %
Change
(’17-’36)
Avg.
Growth
Rate (’17-’36)
Washington 15,192,109 16,571,868 16,714,623 17,138,164 18,008,011 20,090,687 21% 1.0%
Idaho 6,948,564 7,610,218 7,683,310 7,918,057 8,399,424 9,552,526 26% 1.1%
Oregon 7,053,827 8,074,449 8,123,746 8,250,527 8,579,416 9,385,026 16% 0.8%
Total 29,194,500 32,256,535 32,521,679 33,306,747 34,986,852 39,028,239 21% 1.0%

The next step in the study is the development of the three types of potential: technical,
economic and achievable. Technical potential is the theoretical upper limit of conservation
potential. This assumes that all customers replace equipment with the most efficient option
available and adopt the most efficient energy use practices possible at every opportunity
without regard to cost-effectiveness. Economic potential represents the adoption of all cost-
effective conservation measures based on the TRC test in Oregon and the UCT test in
Washington and Idaho. The achievable potential takes into account market maturity,
customer preferences for energy efficiency technologies and expected program participation.
Achievable potential establishes a realistic target for conservation savings that a utility can
expect to achieve through its efficiency programs.
DSM measures that achieve generally uniform year round energy savings independent of
weather are considered base load measures. Examples include high efficiency water heaters,
cooking equipment and front load clothes washers. Weather sensitive measures are those
which are influenced by heating degree day factors and include higher efficiency furnaces,
ceiling/wall/floor insulation, weather stripping, insulated windows, duct work improvements
(tighter sealing to reduce leaks) and ventilation heat recovery systems (capturing chimney
heat). Weather sensitive measures are often referred to as winter load shape measures and
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 51 of 162
are typically valued using a higher avoided cost (due to summer to winter natural gas pricing
differentials) while base load measures often called annual load shape measures are valued
at a lower avoided cost.
Conservation measures are offered to residential, non-residential and low-income1
customers. Measures offered to residential customers are almost universally on a prescriptive
basis, meaning they have a fixed incentive for all customers and do not require individual pre-
project analysis by the utility. Low-income customers are treated with a more flexible approach
through cooperative arrangements with participating Community Action Agency partnerships.
Non-residential customers have access to various prescriptive and site-specific conservation
measures. Site-specific measures are customized to specific applications and have cost and
therm savings that are unique to the individual facility.
In Oregon, some conservation measures are required by law and therefore their costs and
benefits are incorporated into the portfolio without being subject to cost-effectiveness testing.
These measures, for example, include energy audits that do not in and of themselves
generate energy savings absent customer action and the timing and cost-effectiveness of the
action(s) taken by the customer are uncertain.
See Table 3.2 for residential, commercial and industrial measures evaluated in this study for
all three states.

1 For purposes of tables, figures and targets, low income is a subset of residential class.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 52 of 162
Table 3.2: Conservation Measures

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 53 of 162
Conservation Potential Assessment Results
Based upon the previously described methodology and baseline forecasts, AEG developed
technical, economic and achievable potentials by jurisdiction and segment over a full 20-year
horizon.
The technical potential for the overall Avista service territory for the full 20-year IRP horizon
period ultimately reaches 27.5 percent of the baseline end-use forecast.
Economic potential applies the cost-effectiveness metric appropriate to each jurisdiction to
DSM measures identified within the technical potential and quantify the impact of the adoption
of only those DSM measures that are cost-effective. By the end of the 20-year timeframe this
represents 16.4 percent of the baseline energy forecast. The significant difference between
the technical and economic potential is a reflection of the economic impact of falling natural
gas avoided costs as well as the conservation market saturation that was achieved in previous
years with higher prevailing natural gas avoided costs. Past adoption of the most cost-
effective measures leads to progressively higher costs for the remaining measures. At the
same time the avoided cost value of these future adoptions is falling.
The achievable potential across the residential, commercial and industrial sectors,
incorporating ramp rates derived from the NPCC and past company performance, is 9.1
percent of the baseline energy forecast by the end of 2036.
Tables 3.3 and 3.4 summarize cumulative conservation for each potential type for selected
years across the 20-year CPA and IRP horizon. Initially the large commercial sector provides
a relatively higher percentage of the achievable savings compared with its share of sales but
over time this situation reverses such that the residential sector’s share of savings is the
greatest due to growth in residential customer count. For more specific detail, please refer to
the natural gas CPA provided in Appendix 3.1.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 54 of 162
Table 3.3: Summary of Cumulative Achievable, Economic and Technical Conservation
Potential (Dth)
Washington 2017 2018 2021 2026 2036
Baseline Forecast (DTh) 16,571,868 16,714,623 17,138,164 18,008,011 20,090,687
Potential Forecasts (DTh)
Achievable Potential 16,522,957 16,604,429 16,774,904 17,128,936 18,033,128
Economic Potential 16,376,621 16,324,360 16,158,725 16,036,550 16,301,340
Technical Potential 16,272,909 16,117,023 15,652,845 15,062,160 14,504,805
Cumulative Savings (DTh)
Achievable Potential 48,911 110,194 363,259 879,075 2,057,559
Economic Potential 195,247 390,263 979,438 1,971,461 3,789,348
Technical Potential 298,959 597,600 1,485,318 2,945,852 5,585,883
Energy Savings (% of
Baseline)

Achievable Potential 0.3% 0.7% 2.1% 4.9% 10.2%
Economic Potential 1.2% 2.3% 5.7% 10.9% 18.9%
Technical Potential 1.8% 3.6% 8.7% 16.4% 27.8%

Idaho 2017 2018 2021 2026 2036
Baseline Forecast (DTh) 7,610,218 7,683,309 7,918,056 8,399,424 9,552,525
Potential Forecasts (DTh)
Achievable Potential 7,590,455 7,638,632 7,771,409 8,044,198 8,677,000
Economic Potential 7,531,121 7,525,047 7,520,628 7,589,742 7,909,518
Technical Potential 7,477,826 7,418,444 7,255,838 7,070,733 6,949,808
Cumulative Savings (DTh)
Achievable Potential 19,764 44,677 146,648 355,226 875,525
Economic Potential 79,098 158,262 397,428 809,682 1,643,007
Technical Potential 132,392 264,865 662,219 1,328,691 2,602,717
Energy Savings (% of
Baseline)

Achievable Potential 0.3% 0.6% 1.9% 4.2% 9.2%
Economic Potential 1.0% 2.1% 5.0% 9.6% 17.2%
Technical Potential 1.7% 3.4% 8.4% 15.8% 27.2%

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 55 of 162
Oregon 2017 2018 2021 2026 2036
Baseline Forecast (DTh) 8,074,448 8,123,746 8,250,527 8,579,416 9,385,026
Potential Forecasts (DTh)
Achievable Potential 8,059,948 8,092,232 8,156,777 8,341,963 8,759,342
Economic Potential 8,030,116 8,035,068 8,027,229 8,120,949 8,427,077
Technical Potential 7,934,077 7,846,087 7,566,082 7,227,445 6,836,516
Cumulative Savings (DTh)
Achievable Potential 14,501 31,514 93,750 237,453 625,684
Economic Potential 44,332 88,678 223,297 458,467 957,949
Technical Potential 140,371 277,659 684,445 1,351,972 2,548,510
Energy Savings (% of
Baseline)

Achievable Potential 0.2% 0.4% 1.1% 2.8% 6.7%
Economic Potential 0.5% 1.1% 2.7% 5.3% 10.2%
Technical Potential 1.7% 3.4% 8.3% 15.8% 27.2%

The overall achievable potential is presented first by state and by sector in the following
table.
Table 3.4: Summary of Cumulative Achievable Potential by State and Sector (Dth)
Cumulative Savings 2017 2018 2021 2026 2036
Washington 48,911 110,194 363,259 879,075 2,057,559
Idaho 19,764 44,677 146,648 355,226 875,525
Oregon 14,501 31,514 93,750 237,453 625,684
Total 83,176 186,385 603,657 1,471,754 3,558,768

Cumulative Savings 2017 2018 2021 2026 2036
Residential 45,243 101,737 332,135 800,034 2,024,490
Commercial 37,171 83,084 267,342 662,485 1,510,525
Industrial 762 1,564 4,180 9,235 23,754
Total 83,176 186,385 603,657 1,471,754 3,558,768

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 56 of 162
Figure 3.1 illustrates the impact of the conservation potential forecast upon the end-use
baseline absent of any conservation acquisition.

Figure 3.1 - Conservation Potential Energy Forecast (1000 therms)

Potential Results – Residential Single-family homes represent 61 percent of Avista’s residential natural gas customers, but
account for 65 percent of the sector’s consumption in the study base year 2015. In the current
IRP residential provides the largest opportunity for cumulative savings over the next 20 years.
Table 3.5 provides a distribution of achievable potential by state for the residential sector.

-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
35,000,000
40,000,000
45,000,000
2015 2018 2021 2024 2027 2030 2033 2036
En
e
r
g
y

C
o
n
s
u
m
p
t
i
o
n

(
D
T
h
)
Baseline Forecast
Achievable Potential
Economic Potential
Technical Potential
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 57 of 162
Table 3.5 Residential Cumulative Achievable Potential by State, Selected Years
2017 2018 2021 2026 2036
Baseline projection Dth
Washington 10,067,567 10,191,025 10,534,588 11,198,918 12,661,346
Idaho 4,741,736 4,802,813 4,992,555 5,366,588 6,213,091
Oregon 4,808,069 4,852,168 4,931,394 5,137,402 5,588,507
Total 19,617,372 19,846,006 20,458,537 21,702,908 24,462,943
Natural Gas Cumulative Savings Dth
Washington 27,598 62,492 207,653 497,074 1,226,734
Idaho 11,138 25,406 85,812 208,875 536,817
Oregon 6,507 13,839 38,671 94,086 260,939
Total 45,243 101,737 332,135 800,034 2,024,490
% of Total Residential Savings
Washington 61% 61% 63% 62% 61%
Idaho 25% 25% 26% 26% 27%
Oregon 14% 14% 12% 12% 13%

The bulk of the residential potential exists in space heating end-uses followed by water
heating applications. Appliances and miscellaneous end-use loads contribute a small
percentage of potential. Based on a measure-by-measure finding of the potential study the
greatest sources of residential achievable potential across all three jurisdictions are:
 Shell measures and insulation;
 High efficiency furnaces;
 Thermostats and home energy monitoring systems;
 Water-saving devices (low-flow showerheads and faucet aerators); and
 Water heater tank blankets and pipe insulation.
Conservation Potential Results – Commercial and Industrial
The large commercial sector provides the next biggest opportunities for savings. Although
potential as a percentage of baseline use varies between sectors, results do not vary greatly
among the three states under the TRC test; Washington and Idaho have relatively higher
savings due to using the UTC cost effectiveness test. Table 3.6 below details the achievable
conservation potential by sector for selected years.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 58 of 162
Table 3.6 Commercial Achievable Potential by Selected Years
2017 2018 2021 2026 2036
Baseline projection Dth
Washington 6,220,478 6,236,027 6,305,231 6,490,547 7,066,197
Idaho 2,656,853 2,664,007 2,695,763 2,776,753 3,021,253
Oregon 3,215,033 3,219,537 3,264,933 3,383,711 3,729,054
Total 12,092,364 12,119,571 12,265,926 12,651,011 13,816,504
Natural Gas Cumulative Savings Dth
Washington 20,930 46,926 153,614 377,951 822,411
Idaho 8,320 18,631 59,027 141,940 324,991
Oregon 7,921 17,527 54,701 142,594 363,123
Total 37,171 83,084 267,342 662,485 1,510,525
% of Total Commercial Savings
Washington 56% 56% 57% 57% 54%
Idaho 22% 22% 22% 21% 22%
Oregon 21% 21% 20% 22% 24%

Table 3.7 Industrial Cumulative Achievable Potential by Selected Years
2017 2018 2021 2026 2036
Baseline projection Dth
Washington 283,824 287,571 298,345 318,546 363,144
Idaho 211,629 216,490 229,739 256,083 318,182
Oregon 51,346 52,041 54,200 58,303 67,465
Total 546,799 556,101 582,284 632,933 748,791
Natural Gas Cumulative Savings Dth
Washington 383 777 1,993 4,050 8,414
Idaho 306 641 1,809 4,411 13,717
Oregon 73 147 379 773 1,622
Total 762 1,564 4,180 9,235 23,754
% of Total Industrial Savings
Washington 50% 50% 48% 44% 35%
Idaho 40% 41% 43% 48% 58%
Oregon 10% 9% 9% 8% 7%

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 59 of 162
Most of the commercial and industrial conservation potential exists within space heating and
water heating applications. Food preparation, process and miscellaneous represents a
smaller proportion of potential. Primary sources of commercial and industrial sector
achievable savings are:
 Energy management systems and programmable thermostats:
 Retro-commissioning:
 Boiler operating measures such as maintenance;
 Hot water reset and efficient circulation;
 Equipment upgrades for furnaces, boilers and unit heaters; and
 Food service equipment.

Aggregate Conservation Potential Results
The following three tables provide the 2017-2018 CPA identified conservation opportunity
for Idaho, Oregon and Washington, respectively.
Table 3.8: Idaho Natural Gas Target (2017-2018)
11,138 14,268
8,626 10,376
Total

Table 3.9: Oregon Natural Gas Target (2017-2018)
6,507 7,332
7,994 9,680
Total

Table 3.10: Washington Natural Gas Target (2017-2018)
27,598 34,894
21,313 26,390
Total
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 60 of 162
Uses and Applications of the CPA
It is useful to place the IRP process and the CPA component of that process into the larger
perspective of Avista’s efforts to acquire all available cost-effective conservation resources.
Activities outside the immediate scope of the IRP process include the formal annual business
planning and annual cost-effectiveness and acquisition reporting processes in addition to the
ongoing management of the DSM portfolio.
The IRP leads to the establishment of a 20-year avoided cost stream that is essential to
determining the quantity of DSM resources that are cost-effective when compared to the CPA-
identified conservation supply curve and the management of the DSM portfolio between the
two year IRP cycles. The many related and coordinated processes all contribute to the
planning and management of the DSM portfolio towards meeting its cost-effectiveness and
acquisition goals.
The relationship between the CPA and the annual business planning process is of particular
note. The CPA is regarded as a high-level tool that is useful for establishing aggregate targets
and identifying general target markets and target measures. However the CPA of necessity
must make certain broad assumptions regarding key characteristics that are fine-tuned as
part of the creation of an operational business plan. Some of the assumptions that are most
frequently modified include market segmentation, customer eligibility, measure definition,
incentive level, interaction between measures and the opportunities for packaging measures
or coordinated the delivery of measures.
One issue that inevitably arises as part of moving from the CPA analysis to the business
planning process is the treatment of market segments. The CPA defines market segments
(e.g. by residential building type or vintage) to appropriately define the cost-effective potential
for efficiency options and to ensure consistency with system loads and load forecasts.
However, it is often infeasible to recognize these distinctions on an operational basis. This
may result in aggregations of market segments into programs that could lead to more or less
operationally achievable savings.
A second issue that often arises is the “clumpiness” that often occurs with large Commercial
and Industrial projects. Large natural gas conservation projects typically have long lead times
with multiple years between the original customer contact and design of a project to the final
completion with any required measurement and verification. These projects can lead to over
or underperforming targets in individual years but typically average out over the 20-year time
frame of an IRP.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 61 of 162
Conservation Action Plan
The analytical process for the CPA is based on a deterministic model as compared to the
assumptions within the Expected Case. In order to further enhance our analytical
methodology, Avista will focus on the following:
 Explore methods to enable a dynamic analytical process for the evaluation of
conservation potential within individual portfolios.

Conclusion
Avista has a long-term commitment to responsibly pursuing all available and cost-effective
efficiency options as an important means to reduce our customer’s energy cost. Cost-effective
demand-side management options are a key element in our strategy to meet those
commitments. Falling avoided costs and lower growth in customer demand have led to a
reduced role for conservation in the overall natural gas portfolio compared with IRPs done
prior to 2012 however, a regulatory shift to utilizing the UCT in Washington and Idaho DSM
programs will continue to provide a vital role in offsetting future natural gas load growth. The
company has begun to transition our regular income Oregon DSM programs to the ETO with
the ETO being the sole administrator beginning January 1, 2017. Avista is continuing to
adaptively manage our DSM programs in response to the ever shifting economic climate.
Perhaps of most importance in the long-term are the Company’s ongoing efforts to work with
key regional players to develop a regional natural gas market transformation organization and
portfolio. NEEA has begun to execute the first stages of their 2015 – 2019 Natural Gas Market
Transformation Business Plan and we look forward to the conservation opportunities that
arise out of their work in the coming years.
Market transformation is not itself called out within the CPA since the CPA focuses upon
conservation potential without regard to how that potential is achieved. The prospect for a
regional market transformation entity will potentially bring a valuable tool to bear in working
towards the achievement of the cost-effective conservation opportunities identified within the
natural gas CPA.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 62 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 63 of 162
4: Supply-Side Resources
Overview
Avista analyzed a range of future demand scenarios and possible cost-effective
conservation measures to reduce demand. This chapter discusses supply options to meet
net demand. Avista’s objective is to provide reliable natural gas to customers with an
appropriate balance of price stability and prudent cost under changing market conditions.
To achieve this objective, Avista evaluates a variety of supply-side resources and
attempts to build a diversified natural gas supply portfolio. The resource acquisition and
commodity procurement programs resulting from the evaluation consider physical and
financial risks, market-related risks, and procurement execution risks; and identifies
methods to mitigate these risks.
Avista manages natural gas procurement and related activities on a system-wide basis
with several regional supply options available to serve core customers. Supply options
include firm and non-firm supplies, firm and interruptible transportation on six interstate
pipelines, and storage. Because Avista’s core customers span three states, the diversity
of delivery points and demand requirements adds to the options available to meet
customers’ needs. The utilization of these components varies depending on demand and
operating conditions. This chapter discusses the available regional commodity resources
and Avista’s procurement plan strategies, the regional pipeline resource options available
to deliver the commodity to customers, and the storage resource options available to
provide additional supply diversity, enhanced reliability, favorable price opportunities, and
flexibility to meet a varied demand profile. Non-traditional resources are also considered.

Commodity Resources

Supply Basins Avista is fortunate to be located near the two largest natural gas producing regions in
North America – the Western Canadian Sedimentary Basin (WCSB), located in the
Canadian provinces of Alberta and British Columbia, and the Rocky Mountain (Rockies)
gas basin, located in parts of Wyoming, Utah and Colorado. Avista sources most of its
natural gas supplies from these two basins.
Several large pipelines connect the WCSB and Rockies gas basins to the Pacific
Northwest, Southwest, Midwest and Northeast sections of the continent. Historically,
natural gas supplies from the WCSB and Rockies cost less relative to other parts of the
country. Shale gas production from the Northeast has altered flow dynamics and helped
sustain the regional pricing discount. Forecasts show a long-term regional price
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 64 of 162
advantage for WCSB and Rockies gas basins as the need for these supplies in the East
diminishes as more shale gas supply develops in the East.
Increased availability of North American natural gas has prompted a change in the global
LNG landscape. Excess supply has prompted LNG developers to consider exporting
natural gas to capture higher prices in the Asian and European markets. The oil markets
continued oversupply has changed the fortunes of LNG. Since oil prices maintain a
depressed state, the expected fuel switching has not taken place. Switching can occur
when oil is high enough to force users to want to make the switch to an alternate form of
fuel. LNG was expected to be the primary beneficiary of fuel switching. Regionally there
is only one proposed project in Oregon - Jordan Cove. Jordan Cove has received their
FERC export authorization, but in 2016 FERC turned down its supply source known as
Pacific Connector pipeline due to the perception there was insufficient demand. An
updated filing showing demand is expected to take place in the near term, but the results
are unknown at this time. There are 17 announced export LNG projects in British
Columbia. While there is much uncertainty about the number of completed facilities, the
bigger question is the impact of exports on regional infrastructure and prices.

Regional Market Hubs
There are numerous regional market hubs where natural gas is traded extending from the
two primary basins. These regional hubs are typically located at pipeline interconnects.
Avista is located near, and transacts at, most of the Pacific Northwest regional market
hubs, enabling flexible access to several supply points. These supply points include:
 AECO – The AECO-C/Nova Inventory Transfer market center located in Alberta is
a major connection region to long-distance transportation systems, which take
natural gas to points throughout Canada and the United States. Alberta is the
major Canadian exporter of natural gas to the U.S. and historically produced 90
percent of Canada's natural gas.
 Rockies – This pricing point represents several locations on the southern end of
the NWP system in the Rocky Mountain region. The system draws on Rocky
Mountain natural gas-producing areas clustered in areas of Colorado, Utah and
Wyoming.
 Sumas/Huntingdon – This pricing point at Sumas, Washington, is on the
U.S./Canadian border where the northern end of the NWP system connects with
Spectra Energy’s Westcoast Pipeline and predominantly markets Canadian
natural gas from Northern British Columbia.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 65 of 162
 Malin – This pricing point is at Malin, Oregon, on the California/Oregon border
where the pipelines of TransCanada Gas Transmission Northwest (GTN) and
Pacific Gas & Electric Company connect.
 Station 2 – Located at the center of the Spectra Energy/Westcoast Pipeline
system connecting to northern British Columbia natural gas production.
 Stanfield – Located near the Washington/Oregon border at the intersection of the
NWP and GTN pipelines.
 Kingsgate – Located at the U.S./Canadian (Idaho) border where the GTN pipeline
connects with the TransCanada Foothills pipeline.

Given the ability to transport natural gas across North America, natural gas pricing is often
compared to the Henry Hub price. Henry Hub, located in Louisiana, is the primary natural
gas pricing point in the U.S. and is the trading point used in NYMEX futures contracts.
Figure 4.1 shows historic natural gas prices for first-of-month index physical purchases
at AECO, Sumas, Rockies and Henry Hub. The figure illustrates the usually tight
relationship among the regional market hubs; however, there have been periods where
one or more price points have disconnected.
Figure 4.1: Monthly Index Prices

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 66 of 162
Northwest regional natural gas prices typically move together; however, the basis
differential can change depending on market or operational factors. This includes
differences in weather patterns, pipeline constraints, and the ability to shift supplies to
higher-priced delivery points in the U.S. or Canada. By monitoring these price shifts,
Avista can often purchase at the lowest-priced trading hubs on a given day, subject to
operational and contractual constraints.
Liquidity is generally sufficient in the day-markets at most Northwest supply points. AECO
continues to be the most liquid supply point, especially for longer-term transactions.
Sumas has historically been the least liquid of the four major regional supply points
(AECO, Rockies, Sumas and Malin). This illiquidity contributes to generally higher relative
prices in the high demand winter months.
Avista procures natural gas via contracts. Contract specifics vary from transaction-to-
transaction, and many of those terms or conditions affect commodity pricing. Some of the
terms and conditions include:
 Firm vs. Non-Firm: Most term contracts specify that supplies are firm except for
force majeure conditions. In the case of non-firm supplies, the standard provision
is that they may be cut for reasons other than force majeure conditions.
 Fixed vs. Floating Pricing: The agreed-upon price for the delivered gas may be
fixed or based on a daily or monthly index.
 Physical vs. Financial: Certain counterparties, such as banking institutions, may
not trade physical natural gas, but are still active in the natural gas markets. Rather
than managing physical supplies, those counterparties choose to transact
financially rather than physically. Financial transactions provide another way for
Avista to financially hedge price.
 Load Factor/Variable Take: Some contracts have fixed reservation charges
assessed during each of the winter months, while others have minimum daily or
monthly take requirements. Depending on the specific provisions, the resulting
commodity price will contain a discount or premium compared to standard terms.
 Liquidated Damages: Most contracts contain provisions for symmetrical penalties
for failure to take or supply natural gas.

For this IRP, the SENDOUT® model assumes natural gas purchases under a firm,
physical, fixed-price contract, regardless of contract execution date and type of contract.
Avista pursues a variety of contractual terms and conditions to capture the most value for
customers. Avista‘s natural gas buyers actively assess the most cost-effective way to
meet customer demand and optimize unutilized resources.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 67 of 162
Avista’s Natural Gas Procurement Plan
No company can accurately predict future natural gas prices, but market conditions and
experience help shape the overall approach to procurement. Avista’s natural gas
procurement plan process seeks to acquire natural gas supplies while reducing exposure
to short-term price volatility. The procurement strategy includes hedging, storage
utilization and index purchases. Although the specific provisions of the procurement plan
will change based on ongoing analysis and experience, the following principles guide
Avista’s procurement plan.
Avista employs a time, location and counterparty diversified hedging strategy. It is
appropriate to hedge over a period and establish hedge periods when portions of future
demand are physically and/or financially hedged. Avista views hedging as an appropriate
part of a diversified procurement plan with a mission to provide a diversified portfolio
of reliable supply with a level of price certainty in volatile markets. Hedges may not
be at the lowest possible price, but they still protect customers from price volatility. With
access to multiple supply basins, Avista transacts with the lowest priced basin at the time
of the hedge. Furthermore, Avista transacts with a range of counterparties to spread
supply among a wider range of market participants.
Avista evaluates market opportunities as they become available. The abundance of
shale gas, combined with recent lagging weather related demand, has created a near-
term supply glut that continues to suppress price levels. Because of this oversupply,
many oil and natural gas companies have been shedding their assets due to high debts
incurred in the exploration and production side of the business. These companies need
cash to pay debts, so many of these assets can be purchased at a favorable rate
providing a potential market opportunity for Avista to hedge natural gas into the future at
depressed prices.
These market opportunities can include physical/financial instruments or contracts used
to trade or reduce risk or exposure in our current procurement plan portfolio. These
opportunities are primarily market driven with a very short timeframe. A derivative based
contract, a ten-year forward strip or natural gas reserves are some examples of these
market opportunities. Avista has written procedures and guidelines as to whether the
specific opportunity might fit within our portfolio, with a detailed analysis being used to vet
each opportunity when appropriate.
Avista uses a disciplined, but flexible hedging approach. Avista’s hedging strategy
includes the prompt year as part of our short term hedging combined with our long term
hedging of two, three and four future winter periods. In addition to establishing periods
when hedges are to be completed, Avista also sets upper and lower pricing points. This
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 68 of 162
reduces Avista’s exposure to extreme price spikes in a rising market and encourages
capturing the benefit associated with lower prices.
Avista regularly reviews its procurement plan in light of changing market
conditions and opportunities. Avista’s plan is open to change in response to ongoing
review of the procurement plan assumptions. Even though the initial plan establishes
various targets, policies provide flexibility to exercise judgment to revise targets in
response to changing conditions.
Avista utilizes a number of tools to help mitigate financial risks. Avista purchases gas in
the spot market and forward markets. Spot purchases are for the next day or weekend.
Forward purchases are for future delivery. Many of these tools are financial instruments
or derivatives that can provide fixed prices or dampen price volatility. Avista continues to
evaluate how to manage daily demand volatility, whether through option tools from
counterparties or through access to additional storage capacity and/or transportation.

Market-Related Risks and Risk Management
There are several types of risk and approaches to risk management. The 2016 IRP
focuses on two areas of risk: the financial risk of the cost of natural gas to supply
customers will be unreasonably high or volatile, and the physical risk that there may not
be enough natural gas resources (either transportation capacity or the commodity) to
serve core customers.
Avista’s Risk Management Policy describes the policies and procedures associated with
financial and physical risk management. The Risk Management Policy addresses issues
related to management oversight and responsibilities, internal reporting requirements,
documentation and transaction tracking, and credit risk.
Two internal organizations assist in the establishment, reporting and review of Avista’s
business activities as they relate to management of natural gas business risks:
 The Risk Management Committee includes corporate officers and senior-level
management. The committee establishes the Risk Management Policy and
monitors compliance. They receive regular reports on natural gas activity and meet
regularly to discuss market conditions, hedging activity and other natural gas-
related matters.
 The Strategic Oversight Group coordinates natural gas matters among internal
natural gas-related stakeholders and serves as a reference/sounding board for
strategic decisions, including hedges, made by the Natural Gas Supply
department. Members include representatives from the Gas Supply, Accounting,
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 69 of 162
Regulatory, Credit, Power Resources, and Risk Management departments. While
the Natural Gas Supply department is responsible for implementing hedge
transactions, the Strategic Oversight Group provides input and advice.

Transportation Resources
Although proximity to liquid market hubs is important from a cost perspective, supplies
are only as reliable as the pipeline transportation from the hubs to Avista’s service
territories. Capturing favorable price differentials and mitigating price and operational risk
can also be realized by holding multiple pipeline transport options. Avista contracts for a
sufficient amount of diversified firm pipeline capacity from various receipt and delivery
points (including storage facilities), so that firm deliveries will meet peak day demand.
This combination of firm transportation rights to Avista’s service territory, storage facilities
and access to liquid supply basins ensure peak supplies are available to serve core
customers.

The major pipelines servicing the region include:

 Williams - Northwest Pipeline (NWP)
A natural gas transmission pipeline serving the Pacific Northwest moving natural
gas from the U.S./Canadian border in Washington and from the Rocky Mountain
region of the U.S.
 TransCanada Gas Transmission Northwest (GTN): A natural gas transmission
pipeline originating at Kingsgate, Idaho, (Canadian/U.S. border) and terminating
at the California/Oregon border close to Malin, Oregon.
 TransCanada Alberta System: This natural gas gathering and transmission
pipeline in Alberta, Canada, delivers natural gas into the TransCanada Foothills
pipeline at the Alberta/British Columbia border.
 TransCanada Foothills System: This natural gas transmission pipeline delivers
natural gas between the Alberta, British Columbia, border and the Canadian/U.S.
border at Kingsgate, Idaho.
 TransCanada Tuscarora Gas Transmission: This natural gas transmission
pipeline originates at Malin, Oregon, and terminates at Wadsworth, Nevada.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 70 of 162
 Spectra Energy - Westcoast Pipeline: This natural gas transmission pipeline
originates at Fort Nelson, British Columbia, and terminates at the Canadian/U.S.
border at Huntington, British Columbia/Sumas, Washington.
 El Paso Natural Gas - Ruby pipeline: This natural gas transmission pipeline
brings supplies from the Rocky Mountain region of the U.S. to interconnections
near Malin, Oregon.

Avista has contracts with all of the above pipelines (with the exception of Ruby Pipeline)
for firm transportation to serve core customers. Table 4.1 details the firm
transportation/resource services contracted by Avista. These contracts are of different
vintages with different expiration dates; however, all have the right to be renewed by
Avista. This gives Avista and its customers available capacity to meet existing core
demand now and in the future.

Table 4.1: Firm Transportation Resources Contracted (Dth/Day)

Avista defines two categories of interstate pipeline capacity. Direct-connect pipelines
deliver supplies directly to Avista’s local distribution system from production areas,
storage facilities or interconnections with other pipelines. Upstream pipelines deliver
natural gas to the direct-connect pipelines from remote production areas, market centers
and out-of-area storage facilities. Firm storage - max deliverability is specifically tied to
Avista’s withdrawal rights at the Jackson Prairie storage facility and is based on our one
Firm Transportation/Resources Contracted*
Dth/Day

Firm Transportation Winter Summer Winter Summer
NWP TF-1 157,869 157,869 42,699 42,699
GTN T-1 100,605 75,782 42,260 20,640
NWP TF-2 91,200 2,623
Total 349,674 233,651 87,582 63,339
Firm Storage Resources - Max Deliverability
Jackson Prairie
(Owned and
Contracted)346,667 54,623
Total 346,667 54,623
* Represents original contract amounts after releases expire.
Avista Avista
North South
Table 5.1
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 71 of 162
third ownership rights. This number only indicates how much we can withdraw from the
facility as transport on NWP is needed to move it from the facility itself. Figure 4.2
illustrates the direct-connect pipeline network relative to Avista’s supply sources and
service territories.1

Figure 4.2: Direct-Connect Pipelines

Supply-side resource decisions focus on where to purchase natural gas and how to
deliver it to customers. Each LDC has distinctive service territories and geography relative
to supply sources and pipeline infrastructure. Solutions that deliver supply to service
territories among regional LDCs are similar but are rarely generic.
The NWP system, for the most part, is a fully-contracted system. With the exception of
La Grande, Avista’s service territories lie at the end of NWP pipeline laterals. The
Spokane, Coeur d’Alene and Lewiston laterals serve Washington/Idaho load, and the
Grants Pass lateral serves Roseburg and Medford. Capacity expansions of these laterals
1 Avista has a small amount of pipeline capacity with TransCanada Tuscarora Gas Transmission, a natural gas transmission pipeline originating at Malin, Oregon, to service a small number of Oregon
customers near the southern border of the state.
Roseburg
Medford
Stanfield
Washington / Idaho
SUMAS AECO
ROCKS
La Grande
MALIN
Klamath
Falls Roseburg &
Medford
Stanfield
NWP GTN
Washington/Idaho
LaGrande
JP
Storage
Malin
Klamath Falls
AECO
Kingsgate Station 2
Sumas
Rockies
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 72 of 162
would be lengthy and costly endeavors which Avista would likely bear most of the
incremental costs.
The GTN system currently has ample unsubscribed capacity. This pipeline runs directly
through or near most of Avista’s service territories. Mileage based rates provide an
attractive option for securing incremental resource needs.
Peak day planning aside, both pipelines provide an array of options to flexibly manage
daily operations. The NWP and GTN pipelines directly serve Avista’s two largest service
territories, providing diversification and risk mitigation with respect to supply source, price
and reliability. The NWP system (a bi-directional, fixed reservation fee-based pipeline)
provides direct access to Rockies and British Columbia supply and facilitates optionality
for storage facility management. The Stanfield interconnect of the two lines is also
geographically well situated to Avista’s service territories.
The rates used in the planning model start with filed rates currently in effect (See
Appendix 4.1 – Current Transportation/Storage Rates and Assumptions). Forecasting
future pipeline rates is challenging. Assumptions for future rate changes are the result of
market information on comparable pipeline projects, prior rate case experience, and
informal discussions with regional pipeline owners. Pipelines will file to recover costs at
rates equal to the GDP with adjustments made for specific project conditions.
NWP and GTN also offer interruptible transportation services. Interruptible transportation
is subject to curtailment when pipeline capacity constraints limit the amount of natural gas
that may be moved. Although the commodity cost per dekatherm transported is the same
as firm transportation, there are no demand or reservation charges in these transportation
contracts. As the marketplace for release of transportation capacity by the pipeline
companies and other third parties has become more prevalent, the use of interruptible
transportation services has diminished. Avista does not rely on interruptible capacity to
meet peak day core demand requirements.
Avista's transportation acquisition strategy is to contract for firm transportation to serve
core customers on a peak day in the planning horizon. Since contracts for pipeline
capacity are often lengthy and core customer demand needs can vary over time,
determining the appropriate level of firm transportation is a complex analysis. The
analysis includes the projected number of firm customers and their expected annual and
peak day demand, opportunities for future pipeline or storage expansions, and relative
costs between pipelines and upstream supplies. This analysis is on an annual basis and
through the IRP. Active management of underutilized transportation capacity through the
capacity release market and engaging in optimization transactions offsets some
transportation costs. Timely analysis is also important to maintain an appropriate time
cushion to allow for required lead times should the need for securing new capacity arise
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 73 of 162
(See Chapter 5 – Integrated Resource Portfolio for a description of the management of
underutilized pipeline resources).
Avista manages existing resources through optimization to mitigate the costs incurred by
customers until the resource is required to meet demand. The recovery of costs is often
market based with rules governed by the FERC. Avista is recovering full costs on some
resources and partial costs on others. The management of long- and short-term
resources ensures the goal to meet firm customer demand in a reliable and cost-effective
manner. Unutilized resources like supply, transportation, storage and capacity can
combine to create products that capture more value than the individual pieces. Avista has
structured long-term arrangements with other utilities that allow available resources
utilization and provide products that no individual component can satisfy. These products
provide more cost recovery of the fixed charges incurred for the resources. Another
strategy to mitigate transportation costs is to participate in the daily market to assess if
unutilized capacity has value. Avista seeks daily opportunities to purchase natural gas,
transport it on existing unutilized capacity, and sell it into a higher priced market to capture
the cost of the natural gas purchased and recover some pipeline charges. The recovery
is market dependent and may or may not recover all pipeline costs, but mitigates pipeline
costs to customers.

Storage Resources
Storage is a valuable strategic resource that enables improved management of a highly
seasonal and varied demand profile. Storage benefits include:

 Flexibility to serve peak period needs;
 Access to typically lower cost off-peak supplies;
 Reduced need for higher cost annual firm transportation;
 Improved utilization of existing firm transportation via off-season storage injections;
and.
 Additional supply point diversity.

While there are several storage facilities available in the region, Avista’s existing storage
resources consist solely of ownership and leasehold rights at the Jackson Prairie Storage
facility.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 74 of 162
Avista optimizes storage as part of its asset management program. This helps to ensure
a controlled cost mechanism is in place to manage the large supply found within the
storage facility. An example of this storage optimization is selling today at a cash price
and buying a forward month contract. Since forward months have risks or premiums built
into the price the result is Avista locking in a given spread. All optimization of assets go
directly to the customer to reduce their monthly billing.

Jackson Prairie Storage
Avista is one-third owner, with NWP and Puget Sound Energy (PSE), of the Jackson
Prairie Storage Project for the benefit of its core customers in all three states. Jackson
Prairie Storage is an underground reservoir facility located near Chehalis, Washington
approximately 30 miles south of Olympia, Washington. The total working natural gas
capacity of the facility is approximately 25 Bcf. Avista’s current share of this capacity for
core customers is approximately 8.5 Bcf and includes 398,667 Dth of daily deliverability
rights. Besides ownership rights, Avista leased an additional 95,565 Dth of Jackson
Prairie capacity with 2,623 Dth of deliverability from NWP to serve Oregon customers.

Incremental Supply-Side Resource Options
Avista’s existing portfolio of supply-side resources provides a mix of assets to manage
demand requirements for average and peak day events. Avista monitors the following
potential resource options to meet future requirements in anticipation of changing demand
requirements. When considering or selecting a transportation resource, the appropriate
natural gas supply pairs with the transportation resource and the SENDOUT® model
prices the resources accordingly.

System Enhancements
Distribution planning plays a role in the IRP, but is not the primary focus. Distribution
works with supply to meet customer demand on average and peak days. Modifications,
enhancements or upgrades occur on the distribution system that are routine projects,
enhancing system reliability. However, in certain instances, Avista can facilitate additional
peak and base load-serving capabilities through a modification or upgrade of distribution
facilities. These projects enable more takeaway capacity from the interstate pipelines.
When resource deficiencies are identified, Gas Supply works with Distribution
Engineering to assess if the distribution system can facilitate additional take away. These
opportunities are geographically specific and require case-by-case study. Costs of these
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 75 of 162
types of enhancements are included in the context of the IRP. A description of routine
and non-routine system enhancements are in Chapter 7 – Distribution Planning.

Capacity Release Recall
Pipeline capacity not utilized to serve core customer demand is available to sell to other
parties or optimized through daily or term transactions. Released capacity is generally
marketed through a competitive bidding process and can be on a short-term (month-to-
month) or long-term basis. Avista actively participates in the capacity release market with
short-term and long-term capacity releases. Avista assesses the need to recall capacity
or extend a release of capacity on an on-going basis. The IRP process evaluates if or
when to recall some or all long-term releases.

Existing Available Capacity
In some instances, there is available capacity on existing pipelines. NWP’s mainline is
fully subscribed; however, GTN mainline has available capacity. There is some
uncertainty about the future capacity availability as the demand needs of utilities and end-
users vary across the region. Avista models access to the GTN capacity as an option to
meet future demand needs.

GTN Backhauls
The GTN interconnection with the Ruby Pipeline has enabled GTN the physical capability
to provide a limited amount of firm back-haul service from Malin with minor modifications
to their system. Fees for utilizing this service are under the existing Firm Rate Schedule
(FTS-1) and currently include no fuel charges. Additional requests for back-haul service
may require additional facilities and compression (i.e., fuel).
This service can provide an interesting solution for Oregon customers. For example,
Avista can purchase supplies at Malin, Oregon and transport those supplies to Klamath
Falls or Medford. Malin-based natural gas supplies typically include a higher basis
differential to AECO supplies, but are generally less expensive than the cost of forward-
haul transporting traditional supplies south and paying the associated demand charges.
The GTN system is a mileage-based system, so Avista pays only a fraction of the rate if
it is transporting supplies from Malin to Medford and Klamath Falls. The GTN system is
approximately 612 miles long and the distance from Malin to the Medford lateral is only
about 12 miles.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 76 of 162
New Pipeline Transportation
Additional firm pipeline transportation resources are viable and attractive resource
options. However, determining the appropriate level, supply source and associated
pipeline path, costs and timing, and if existing resources will be available at the
appropriate time, make this resource difficult to analyze. Firm pipeline transportation
provides several advantages; it provides the ability to receive firm supplies at the
production basin, it provides for base-load demand, and it can be a low-cost option given
optimization and capacity release opportunities. Pipeline transportation has several
drawbacks, including typically long-dated contract requirements, limited need in the
summer months (many pipelines require annual contracts), and limited availability and/or
inconvenient sizing/timing relative to resource need.
Pipeline expansions are typically more expensive than existing pipeline capacity and
often require long-term contracts. Even though expansions may be more expensive than
existing capacity, this approach may still provide the best option given that some of the
other options require matching pipeline transportation. Matching pipeline transportation is
creating equivalent volumes on different pipelines from the basin to the delivery point in
order to fully utilize subscribed capacity. Expansions may also provide increased reliability
or access to supply that cannot be obtained through existing pipelines. This is the case
with the Pacific Connector pipeline being proposed as the connecting feedstock for the
Jordan Cove LNG facility in Oregon. The pipeline’s current path connects into Northwest
Pipelines Grants Pass Lateral where capacity is limited. The Pacific Connector pipeline
would add an additional 50,000 Dth/day of capacity along that lateral and would push up
from a south-to-north direction.
Several specific projects have been proposed for the region. The following summaries
describe these projects while Figure 4.3 illustrates their location.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 77 of 162

Figure 4.3: Proposed Pipeline Locations

 NWP Washington Expansion
NWP continues to explore options to expand service from Sumas, Wash., to
markets along the Interstate-5 corridor. Looping sections of 36-inch diameter
pipeline with the existing pipeline and additional compression at existing
compressor stations can add incremental capacity. Actual miles of pipe and
incremental compression will determine the amount of capacity created, but it can
scale to meet market demand. This project is currently under FERC review.
 Blue Bridge/Palomar Expansion
NWP began working with Palomar Gas Transmission (a partnership between NW
Natural and TransCanada) to develop the Cascade (eastern) section of the
previously proposed Palomar gas transmission line in conjunction with an
expansion of NWP’s existing system. The proposed 106-mile, 30-inch-diameter
pipeline would extend from TransCanada’s GTN’s mainline to NW Natural’s
system near Molalla, Oregon. It would be a bi-directional pipeline with an initial
Source: Northwest Gas Association

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 78 of 162
capacity of up to 300 MMcf/d expandable to 750 MMcf/d. In 2011, Palomar Gas
Transmission withdrew its application for this pipeline, yet remains prepared if
natural gas demand rebounds.
 Spectra/FortisBC System Enhancement
FortisBC and Spectra Energy are considering a 100-mile, 24-inch expansion
project from Kingsvale to Oliver, British Columbia, to expand service to the Pacific
Northwest and California markets. Removing constraints will allow expansion of
Spectra’s T-South enhanced service offering, which provides shippers the options
of delivering to Sumas or the Kingsgate market. Expanding the bi-directional
Southern Crossing system would increase capacity at Sumas during peak demand
periods. Initial capacity from the Spectra system to Kingsgate would be 300
MMcf/d, expandable to 450 MMcf/d. An expanded east-to-west flow will increase
delivery of supply to Sumas by an additional 150 MMcf/d. Currently, there is no
plan to construct this pipeline, but it would be available if demand was sufficient.
 Pacific Connector
Veresen and The Williams Company are currently attempting to acquire approval
for a 232-mile, 36-inch diameter pipeline designed to transport up to 1 billion cubic
feet of natural gas per day from interconnects near Malin, Oregon, to the Jordan
Cove LNG terminal in Coos Bay, Oregon. The pipeline would deliver the feedstock
to the LNG terminal providing natural gas to international markets, but also to the
Pacific Northwest. The pipeline will connect with Williams’ Northwest Pipeline on
the Grants Pass lateral. This ties in directly within Avista’s service territory and will
bring in an additional 50,000 Dth/day of capacity into that area. This new option
could provide Avista’s customers in the area new capacity for growth and supply
diversity. In order to show support of this project, Avista signed a 10,000 Dth/day
non-binding contract for capacity in a display of its support to the project.

Avista supports proposals that bring supply diversity and reliability to the region. Supply
diversity provides a diverse supply base in the procurement of goods and services. Since
there are few options in the Northwest, supply diversity provides options and security
when constraints or high demand are present. Avista engages in discussions and analysis
of the potential impact of each regional proposal from a demand serving and
reliability/supply diversity perspective. In most cases, for Avista to consider them a viable
incremental resource to meet demand needs would require combining them with
additional capacity on existing pipeline resources. However, the IRP considers a generic
expansion that represents a new pipeline build to Avista’s service territories.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 79 of 162

In-Ground Storage
In-ground storage provides advantages when natural gas from storage can be delivered
to Avista’s city-gates. It enables deliveries of natural gas to customers during peak cold
weather events. It also facilitates potentially lower-cost supply for customers by capturing
peak/non-peak pricing differentials and potential arbitrage opportunities within individual
months. Although additional storage can be a valuable resource, without deliverability to
Avista’s service territory, this storage cannot be an incremental firm peak serving
resource.

Jackson Prairie
Jackson Prairie is a potential resource for expansion opportunities. Any future storage
expansion capacity does not include transportation and therefore cannot be considered
an incremental peak day resource. However, Avista will continue to look for exchange
and transportation release opportunities that could fully utilize these additional resource
options. When an opportunity presents itself, Avista assesses the financial and reliability
impact to customers. Due to the fast paced growth in the region, and the need for new
resources, a future expansion is possible, though a robust analysis would be required to
determine feasibility. Currently, there are no plans for immediate expansion of Jackson
Prairie.

Other In-Ground Storage Other regional storage facilities exist and may be cost effective. Additional capacity at
Northwest Natural’s Mist facility, capacity at one of the Alberta area storage facilities,
Questar’s Clay Basin facility in northeast Utah, Ryckman Creek in Uinta County, Wyo.,
and northern California storage are all possibilities. Transportation to and from these
facilities to Avista’s service territories continues to be the largest impediment to these
options. Avista will continue to look for exchange and transportation release opportunities
while monitoring daily metrics of load, transport and market environment.

LNG and CNG
LNG is another resource option in Avista’s service territories and is suited for meeting
peak day or cold weather events. Satellite LNG uses natural gas that is trucked to the
facilities in liquid form from an offsite liquefaction facility. Alternatively, small-scale
liquefaction and storage may also be an effective resource option if natural gas supply
during non-peak times is sufficient to build adequate inventory for peak events. Permitting
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 80 of 162
issues notwithstanding, facilities could be located in optimal locations within the
distribution system.
CNG is another resource option for meeting demand peaks and is operationally similar to
LNG. Natural gas could be compressed offsite and delivered to a distribution supply point
or compressed locally at the distribution supply point if sufficient natural gas supply and
power for compression is available during non-peak times.
LNG and CNG supply resource options for LDCs are becoming more attractive as the
market for LNG and CNG as alternative transportation fuels develops. The combined
demand for peaking and transportation fuels can increase the volume and utilization of
these resource assets thus lowering unit costs for the benefit of both market segments.
Estimates for LNG and CNG resources vary because of sizing and location issues. This
IRP uses estimates from other facilities constructed in the area and from conversations
with experts in the industry. Avista will monitor and refine the costs of developing LNG
and CNG resources while considering lead time requirements and environmental issues.

Plymouth LNG NWP owns and operates an LNG storage facility at Plymouth, Wash., which provides
natural gas liquefaction, storage and vaporization service under its LS-1 and LS-2F tariffs.
An example ratio of injection and withdrawal rates show that it can take more than 200
days to fill to capacity, but only three to five days to empty. As such, the resource is best
suited for needle-peak demands. Incremental transportation capacity to Avista’s service
territories would have to be obtained in order for it to be an effective peaking resource.
With available capacity, Plymouth LNG was considered in our supply side resource
modeling but was not selected.

Avista-Owned Liquefaction LNG
Avista could construct a liquefaction LNG facility in the service area. Doing so could use
excess transportation during off-peak periods to fill the facility, avoid tying up
transportation during peak weather events, and it may avoid additional annual pipeline
charges.
Construction would depend on regulatory and environmental approval as well as cost-
effectiveness requirements. Preliminary estimates of the construction, environmental,
right-of-way, legal, operating and maintenance, required lead times, and inventory costs
indicate company-owned LNG facilities have significant development risks. Avista
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 81 of 162
modeling included LNG, but it was not selected as a resource when compared to existing
resources.

Biogas
Biogas typically refers to a mixture of gases produced by the biological breakdown of
organic matter in the absence of oxygen. Biogas can be produced by anaerobic digestion
or fermentation of biodegradable materials such as woody biomass, manure or sewage,
municipal waste, green waste and energy crops. This type of biogas is primarily methane
and carbon dioxide.
Biogas is a renewable fuel, so it may qualify for renewable energy subsidies. Avista is not
aware of any current subsidies, but future stimulus or state or federal energy policies
could lead to some form of financial incentives.
Biogas projects are unique, so reliable cost estimates are difficult to obtain. Project
sponsorship has many complex issues, and the more likely participation in such a project
is as a long-term contracted purchaser. Avista did not consider biogas as a resource in
this planning cycle, since they are small and relatively insignificant compared to demand,
but remains receptive to such projects as they are proposed.

Supply Scenarios
The 2016 IRP includes two supply scenarios. Additional details about the results of the
supply scenarios are in Chapters 5 and 6.
 Existing Resources: This scenario represents all resources currently owned or
contracted by Avista.
 Existing + Expected Available: In this scenario, existing resources plus supply
resource options expected to be available when resource needs are identified. This
includes currently available south and north bound GTN, capacity release recalls,
NWP expansions and satellite LNG.

Supply Issues
The abundance and accessibility of shale gas has fundamentally altered North American
natural gas supply and the outlook for future natural gas prices. Even though the supply
is available and the technology exists to access it, there are issues that can affect the
cost and availability of natural gas.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 82 of 162
Hydraulic Fracturing
Improvements in hydraulic fracturing, a 60-year-old technique used to extract oil and
natural gas from shale rock formations, coupled with horizontal drilling has enabled
access to previously uneconomic resources. However, the process does not come
without challenges. The publicity caused by movies, documentaries and articles in
national newspapers about “fracking” has plagued the natural gas and oil industry. There
is concern that hydraulic fracturing is contaminating aquifers, increasing air pollution and
causing earthquakes. The wide-spread publicity generated interest in the production
process and caused some states to issue bans or moratoriums on drilling until further
research was conducted.
Government, industry and universities engaged in studies to understand the actual and
potential impacts of hydraulic fracturing. Industry has been working to refute these claims
by focusing on ensuring companies use best practices for well drilling, disclosing the fluids
used in the hydraulic fracturing processing, and implementing “green completions” for
wells. In the past, wells either flared off the initial natural gas or released an excess
amount of natural gas into the air. Green completions is now a standard for well drilling
and captures the natural gas at the well head instead of releasing it.

Pipeline Availability The Pacific Northwest has efficiently utilized its relatively sparse network of pipeline
infrastructure to meet the region’s needs. As the amount of renewable energy increases,
future demand for natural gas-fired generation will increase. Pipeline capacity is the link
between natural gas and power.
Adding additional pressure to existing pipeline resources is the announcement of three
proposed methanol plants in the region. The plants use large amounts of natural gas as
a feedstock for creating methanol, which is used to make other chemicals and as a fuel.
To date, the Port of Kalama is gaining ground in its approval process and is looking like
the most probable of the three methanol plants and will take around 300,000 Dth/day in
a region already constrained by pipeline deliverability.
LDCs will have to compete with power generators, LNG exporters and other large end
users for limited pipeline capacity. The new mix could alter current pipeline operations
and the potential availability of infrastructure to the region. This future competition of
pipeline capacity should have little impact on Avista’s existing portfolio of pipeline
contracts unless the overall cost of the new pipeline capacity is less than the current cost
with the same deliverability. In general, new pipeline capacity will have higher costs than
operational pipelines, though a thorough analysis is needed prior to making this
determination.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 83 of 162
Ongoing Activity
Without resource deficiencies or a need to acquire incremental supply-side resources to
meet peak day demands over the next 20 years, Avista will focus on normal activities in
the near term, including:
 Continue to monitor supply resource trends including the availability and price of
natural gas to the region, LNG exports, supply dynamics and marketplace, and
pipeline and storage infrastructure availability.
 Monitor availability of resource options and assess new resource lead-time
requirements relative to resource need to preserve flexibility.
 Appropriate management of existing resources including optimizing underutilized
resources to help reduce costs to customers.

Conclusion
Avista is committed to providing reliable supplies of natural gas to its customers. Avista
procures supplies with a diversified plan that seeks to acquire natural gas supplies while
reducing exposure to short-term price volatility through a strategy that includes hedging,
storage utilization and index purchases. The supply mix includes long-term contracts for
firm pipeline transportation capacity from many supply points and ownership and leasing
of firm natural gas storage capacity sufficient to serve customer demand during peak
weather events and throughout the year.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 84 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 85 of 162
5: Integrated Resource Portfolio

Overview
This chapter combines the previously discussed IRP components and the model used to
determine resource deficiencies during the 20-year planning horizon. This chapter also
provides an analysis of potential resource options to meet resource deficiencies as
exhibited in the High Growth, Low Prices scenario.
The foundation for integrated resource planning is the criteria used for developing
demand forecasts. Avista uses the coldest day on record as its weather-planning
standard for determining peak-day demand. This is consistent with past IRPs as
described in Chapter 2  Demand Forecasts. This IRP utilizes coldest day on record and
average weather data for each demand region. Avista plans to serve expected peak day
in each demand region with firm resources. Firm resources include natural gas supplies,
firm pipeline transportation and storage resources. In addition to peak requirements,
Avista also plans for non-peak periods such as winter, shoulder and summer demand.
The modeling process includes a daily optimization for every day of the 20-year planning
period.
It is assumed that on a peak day all interruptible customers have left the system in order
to provide service to firm customers. Avista does not make firm commitments to serve
interruptible customers. Therefore, IRP analysis of demand-serving capabilities only
includes the residential, commercial and firm industrial classes. Using coldest day on
record weather criteria, a blended price curve developed by industry experts, and an
academically backed customer forecast all work together to develop stringent planning
criteria.
Forecasted demand represents the amount of natural gas supply needed. In order to
deliver the forecasted demand, the supply forecast needs to be increased between 1.0
percent and 3.0 percent on both an annual and peak-day basis to account for additional
supplies that are purchased primarily for pipeline compressor station fuel. The 1.0 percent
to 3.0 percent, known as fuel, varies depending on the pipeline. The FERC and National
Energy Board approved tariffs govern the percentage of required additional fuel supply.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 86 of 162
SENDOUT® Planning Model
The SENDOUT® Gas Planning System from Ventyx performs integrated resource
optimization modeling. Avista purchased the SENDOUT® model in April 1992 and has
used it to prepare all IRPs since then. Avista has a long-term maintenance agreement
with Ventyx for software updates and enhancements. Enhancements include software
corrections and improvements driven by industry needs.
SENDOUT® is a linear programming model widely used to solve natural gas supply and
transportation optimization questions. Linear programming is a proven technique to solve
minimization/maximization problems. SENDOUT® analyzes the complete problem at one
time within the study horizon, while accounting for physical limitations and contractual
constraints.
The software analyzes thousands of variables and evaluates possible solutions to
generate a least cost solution. The model uses the following variables:

 Demand data, such as customer count forecasts and demand
coefficients by customer type (e.g., residential, commercial and
industrial).
 Weather data, including minimum, maximum and average
temperatures.
 Existing and potential transportation data which describes the network
for physical movement of natural gas and associated pipeline costs.
 Existing and potential supply options including supply basins, revenue
requirements as the key cost metric for all asset additions and prices.
 Natural gas storage options with injection/withdrawal rates, capacities
and costs.
 Conservation potential.

Figure 5.1 is a SENDOUT® network diagram of Avista’s demand centers and resources.
This diagram illustrates current transportation and storage assets, flow paths and
constraint points.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 87 of 162
Figure 5.1 SENDOUT® Model Diagram

The SENDOUT® model provides a flexible tool to analyze scenarios such as:

 Pipeline capacity needs and capacity releases;
 Effects of different weather patterns upon demand;
 Effects of natural gas price increases upon total natural gas costs;
 Storage optimization studies;
 Resource mix analysis for conservation;
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 88 of 162
 Weather pattern testing and analysis;
 Transportation cost analysis;
 Avoided cost calculations; and
 Short-term planning comparisons.

SENDOUT® also includes Monte Carlo capabilities, which facilitates price and demand
uncertainty modeling and detailed portfolio optimization techniques to produce probability
distributions. More information and analytical results are located in Chapter 6 – Alternate
Scenarios, Portfolios and Stochastic Analysis.
Resource Integration
The following sections summarize the comprehensive analysis bringing demand
forecasting and existing and potential supply and demand-side resources together to form
the 20-year, least-cost plan.

Demand Forecasting
Chapter 2 - Demand Forecasts describes Avista’s demand forecasting approach.
Avista forecasts demand in the SENDOUT® model in eight service areas given the
existence of distinct weather and demand patterns for each area and pipeline
infrastructure dynamics. The SENDOUT® areas are Washington/Idaho (disaggregated
into three sub-areas because of pipeline flow limitations); Medford (disaggregated into
two sub-areas because of pipeline flow limitations); and Roseburg, Klamath Falls and La
Grande. In addition to area distinction, Avista also models demand by customer class
within each area. The relevant customer classes are residential, commercial and firm
industrial customers.
Customer demand is highly weather-sensitive. Avista’s customer demand is not only
highly seasonable, but also highly variable. Figure 5.2 captures this variability showing
monthly system-wide average demand, minimum demand day observed by month,
maximum demand day observed in each month, and winter projected peak day demand
for the first year of the Expected Case forecast as determined in SENDOUT®.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 89 of 162
Figure 5.2: Total System Average Daily Load (Average, Minimum and Maximum)

Natural Gas Price Forecasts
Natural gas prices are a fundamental component of the IRP. The commodity price is a
significant component of the total cost of a resource option. This affects the avoided cost
threshold for determining cost-effectiveness of conservation measures. The price of
natural gas influences consumption, so price elasticity is part of the demand evaluation
shown in Chapter 2 – Demand Forecasts.
The natural gas price outlook has changed dramatically in recent years in response to
several influential events and trends affecting the industry. The recent recession, shale
gas production, greenhouse gas issues, and renewable energy standards creating the
potential for more natural gas-fired generation impact the natural gas outlook. The rapidly
changing environment and uncertainty in predicting future events and trends, requires
modeling a range of forecasts.
Many additional factors influence natural gas pricing and volatility, such as regional
supply/demand issues, weather conditions, hurricanes/storms, storage levels, natural
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 90 of 162
gas-fired generation, infrastructure disruptions, and infrastructure additions (e.g. new
pipelines and LNG terminals).
Even though Avista continually monitors these factors, we cannot accurately predict
future prices for the 20-year horizon of this IRP. This IRP reviewed several price forecasts
from credible sources. Figure 5.3 depicts the price forecasts considered in the IRP
analyses.

Figure 5.3: Henry Hub Forecasted Price (Real $/Dth)

Selecting the price curves can be more art than science. With the assistance of the TAC,
Avista selected high, expected and low price curves to consider possible outcomes and
their impact on resource planning. The expected curve was a blended price derived from
two consulting services subscriptions along with the NYMEX forward strip on January 7,
2016. The high and low price curves were derived via a Monte Carlo simulation of 500
draws where a high and low price were selected from these draws. The selected price
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 91 of 162
curves have variation and provide reasonable upper and lower bounds, consistent with
stretching modeling assumptions to address uncertainty in the planning environment.
These curves are in real dollars in Figure 5.4. Additionally, stochastic modeling of natural
gas prices is also completed. The results from that analysis are in Chapter 6 – Alternate
Scenarios, Portfolios and Stochastic Analysis.

Figure 5.4 Henry Hub Forecasts for IRP Low/ Medium/ High Forecasted Price – Real $/Dth

Each of the price forecasts above are for Henry Hub, which is located in Louisiana just
onshore from the Gulf of Mexico. Henry Hub is recognized as the most important pricing
point in the U.S. because of its proximity to a large portion of U.S. natural gas production
and the sheer volume traded in the daily or spot market, as well as the forward markets
via the New York Mercantile Exchange’s (NYMEX) futures contracts. Consequently, all
other trading points tend to be priced off of the Henry Hub.
The primary physical supply points at Sumas, AECO and the Rockies (and other
secondary regional market hubs) determine Avista’s costs. Prices at these points typically
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 92 of 162
trade at a discount, or negative basis differential, to Henry Hub because of their proximity
to the two largest natural gas basins in North America (the WCSB and the Rockies).
Table 5.1 shows the Pacific Northwest regional prices from the consultants, historic
averages and the prior IRP as a percent of Henry Hub price, along with three-year
historical comparisons.

Table 5.1: Regional Price as a Percent of Henry Hub Price

89.9% 98.8% 95.4% 101.4% 100.4%
85.3% 94.2% 96.7% 98.6% 96.8%
86.8% 97.2% 97.1% 99.6% 97.5%
82.5% 90.8% 88.9% 94.5% 92.1%

This IRP used monthly prices for modeling purposes because of Avista’s winter-weighted
demand profile. Table 5.2 depicts the monthly price shape used in this IRP. A slight
change to the shape of the pricing curve occurred since the last IRP. Driven primarily by
supply availability, the forecasted differential between winter and summer pricing has
decreased to some extent compared to historic data.

Table 5.2: Monthly Price as a Percent of Average Price
Jan Feb Mar Apr May Jun
Consult1 104.7% 104.2% 96.8% 95.9% 96.6% 98.2%
Consult2 101.0% 101.6% 101.5% 98.9% 98.8% 98.5%
2014 IRP 102.0% 101.5% 98.5% 98.0% 98.5% 100.5%
Jul Aug Sep Oct Nov Dec
Consult1 99.2% 99.7% 98.9% 99.4% 101.0% 105.2%
Consult2 99.3% 99.3% 100.3% 99.3% 100.5% 101.1%
2014 IRP 101.5% 102.0% 98.5% 98.5% 99.0% 103.0%

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 93 of 162
Avista selected a blend of Consultant 1 and Consultant 2’s forecast of regional prices and
monthly shapes. Appendix 5.1 – Monthly Price Data by Basin contains detailed monthly
price data behind the summary table information discussed above.
Carbon Policy
To help address carbon scenarios within our jurisdictions and at a federal level, Avista
included multiple sensitivities and analysis around carbon policy and legislation. The
expected price was derived from a consultants forecast beginning in 2026 to 2035.
Avista’s expectation of a carbon policy begins earlier, from 2018-2025, and includes a
form of cap and trade policy in several of our jurisdictions. The blending occurred in a
way to ensure no double counting of price adders as there are no cross over years
between policies.
Avista models carbon as an incremental price adder to address any potential policy.
Carbon adders increase the price of a dekatherm of natural gas and can impact resource
selections and demand through expected elasticity (Chapter 2 – Demand Forecasts,
Price Elasticity). The starting price was assumed to be similar to California’s cap and
trade system where the initial floor was set at $10 per metric ton of CO2. A blending of
the likely policy as an assumed two sigma, or 95.45 percent, of expected outcomes. The
remaining distribution was equally divided into the remaining likelihood between the high
case, Washington State’s I-732, and the low case of no carbon adder. The final, Expected
Case, incremental adder to our Henry Hub pricing has a starting price of $9.89 per metric
ton starting in 2018 and ramps up to $19.93 by 2035.
Transportation and Storage
Valuing natural gas supplies is a critical first step in resource integration. Equally
important is capturing all costs to deliver the natural gas to customers. Daily capacity of
existing transportation resources (described in Chapter 4 – Supply-Side Resources) is
represented by the firm resource duration curves depicted in Figures 5.6 and 5.7.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 94 of 162
Figure 5.6: Existing Firm Transportation Resources – Washington/Idaho

Figure 5.7: Existing Firm Transportation Resources – Oregon

0
50
100
150
200
250
300
350
400
450
500
1 31 61 91 121 151 181 211 241 271 301 331 361
MDth
Day of Year
0
20
40
60
80
100
120
140
160
180
200
1 31 61 91 121 151 181 211 241 271 301 331 361
MDth
Day of Year
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 95 of 162
Current rates for capacity are in Appendix 5.1 – Monthly Price Data by Basin. Forecasting
future pipeline rates can challenging because of the need to estimate the amount and
timing of rate changes. Avista’s estimates and timing of future pipeline rate increases are
based on knowledge obtained from industry discussions and participation in pipeline rate
cases. This IRP assumes that pipelines will file to recover costs at rates equal to increases
in GDP (see Appendix 5.2 – Weighted Average Cost of Capital).

Demand-Side Management
Chapter 3 – Demand-Side Resources describes the methodology used to identify
conservation potential and the interactive process that utilizes avoided cost thresholds for
determining the cost effectiveness of conservation measures on an equivalent basis with
supply-side resources.

Preliminary Results
After incorporating the above data into the SENDOUT® model, Avista generated an
assessment of demand compared to existing resources for several scenarios. Chapter 2
– Demand Forecasts discusses the demand results from these cases, with additional
details in Appendices 2.1 through 2.9.
Figures 5.8 through 5.11 provide graphic summaries of Average Case demand as
compared to existing resources on a peak day. This demand is net of conservation
savings and shows the adequacy of Avista’s resources under normal weather conditions.
For this case, current resources meet demand needs over the planning horizon.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 96 of 162
Figure 5.8: Average Case – Washington/Idaho Existing Resources vs. Peak Day Demand
– February 15th

Figure 5.9: Average Case – Medford / Roseburg Existing Resources vs. Peak Day
Demand – December 20th

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 97 of 162
Figure 5.10: Average Case – Klamath Falls Existing Resources vs. Peak Day Demand –
December 20th

Figure 5.11: Average Case – La Grande Existing Resources vs. Peak Day Demand –
February 15th

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 98 of 162
Figures 5.12 through 5.15 summarize Expected Case peak day demand compared to
existing resources, as well as demand comparisons to the 2014 IRP. This demand is net
of conservation savings. Based on this information, and more specifically where a
resource deficiency is nearly present as shown in Figure 5.9, Avista has time to carefully
monitor, plan and take action on potential resource additions as described in the Ongoing
Activities section of Chapter 8 – Action Plan. Any underutilized resources will be optimized
to mitigate the costs incurred by customers until the resource is required to meet demand.
This management, of both long- and short-term resources, ensures the goal to meet firm
customer demand in a reliable and cost-effective manner as described in Supply Side
Resources – Chapter 4.

Figure 5.12: Expected Case – Washington/Idaho Existing Resources vs. Peak Day
Demand – February 15th

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 99 of 162
Figure 5.13: Expected Case – Medford / Roseburg Existing Resources vs. Peak Day
Demand – December 20th

Figure 5.14: Expected Case – Klamath Falls Existing Resources vs. Peak Day Demand –
December 20th

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 100 of 162
Figure 5.15: Expected Case – La Grande Existing Resources vs. Peak Day Demand –
February 15th

If demand grows faster than expected, the need for new resources will come earlier. Flat
demand risk requires close monitoring for signs of increasing demand and reevaluation
of lead times to acquire preferred incremental resources. Monitoring of flat demand risk
includes a reconciliation of forecasted demand to actual demand on a monthly basis. This
reconciliation helps identify customer growth trends and use-per-customer trends. If they
meaningfully differ compared to forecasted trends, Avista will assess the impacts on
planning from procurement and resource sufficiency standing.
Table 5.3 quantifies the forecasted total demand net of conservation savings and
unserved demand from the above charts.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 101 of 162
Table 5.3: Peak Day Demand – Served and Unserved (MDth/day)

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 102 of 162
New Resource Options
When existing resources are not sufficient to meet expected demand, there are many
important considerations in determining the appropriateness of potential resources.
Interruptible customers’ transportation may be cut, as needed, when existing resources
are not sufficient to meet firm customer demand.

Resource Cost
Resource cost is the primary consideration when evaluating resource options, although
other factors mentioned below also influence resource decisions. Newly constructed
resources are typically more expensive than existing resources, but existing resources
are in shorter supply. Newly constructed resources provided by a third party, such as a
pipeline, may require a significant contractual commitment. Newly constructed resources
are often less expensive per unit, if a larger facility is constructed, because of economies
of scale.

Lead Time Requirements
New resource options can take from one to five or more years to put in service. Open
season processes to determine interest in proposed pipelines, planning and permitting,
environmental review, design, construction, and testing are some of the aspects
contributing to lead time requirements for new facilities. Recalls of released pipeline
capacity typically require advance notice of up to one year. Even DSM programs can
require significant time from program development and rollout to the realization of natural
gas savings.

Peak versus Base Load
Avista’s planning efforts include the ability to serve firm natural gas loads on a peak day,
as well as all other demand periods. Avista’s core loads are considerably higher in the
winter than the summer. Due to the winter-peaking nature of Avista’s demand, resources
that cost-effectively serve the winter without an associated summer commitment may be
preferable. Alternatively, it is possible that the costs of a winter-only resource may exceed
the cost of annual resources after capacity release or optimization opportunities are
considered.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 103 of 162
Resource Usefulness
Available resources must effectively deliver natural gas to the intended region. Given
Avista’s unique service territories, it is often impossible to deliver resources from a
resource option, such as storage, without acquiring additional pipeline transportation.
Pairing resources with transportation increases cost. Other key factors that can contribute
to the usefulness of a resource are viability and reliability. If the potential resource is either
not available currently (e.g., new technology) or not reliable on a peak day (e.g., firm),
they may not be considered as an option for meeting unserved demand.

“Lumpiness” of Resource Options
Newly constructed resource options are often “lumpy.” This means that new resources
may only be available in larger-than-needed quantities and only available every few
years. This lumpiness of resources is driven by the cost dynamics of new construction,
where lower unit costs are available with larger expansions and the economics of
expansion of existing pipelines or the construction of new resources dictate additions
infrequently. The lumpiness of new resources provides a cushion for future growth.
Economies of scale for pipeline construction provide the opportunity to secure resources
to serve future demand increases.

Competition
LDCs, end-users and marketers compete for regional resources. The Northwest has been
efficient in the utilization of existing resources and has an appropriately sized system.
Currently, the region can accommodate the regional demand needs. However, future
needs vary, and regional LDCs may find they are competing with each other and other
parties to secure firm resources for customers.

Risks and Uncertainties
Investigation, identification, and assessment of risks and uncertainties are critical
considerations when evaluating supply resource options. For example, resource costs
are subject to degrees of estimation, partly influenced by the expected timeframe of the
resource need and rigor determining estimates, or estimation difficulties because of the
uniqueness of a resource. Lead times can have varying degrees of certainty ranging from
securing currently available transport (high certainty) to building underground storage
(low certainty).

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 104 of 162
Resource Selection
After identifying supply-side resource options and evaluating them based on the above
considerations, Avista entered the supply-side scenarios (see Table 5.2) and
conservation measures (see Chapter 3 – Demand-Side Resources) into the SENDOUT®
model for it to select the least cost approach to meeting resource deficiencies, if they
exist. SENDOUT® compares demand-side and supply-side resources (see Appendix 5.3
– Supply Side Resource Options for a list of available options) using PVRR analysis to
determine which resource is a least cost/least risk resource.

Demand-Side Resources

Integration by Price
As described in Chapter 3 – Demand-Side Resources, the model runs without future DSM
programs. This preliminary run provides an avoided cost curve for Applied Energy Group
(AEG). AEG then evaluates the cost effectiveness of DSM programs against the initial
avoided cost curve using the appropriate resource cost tests. The therm savings and
associated program costs are incorporated into the SENDOUT® model. After
incorporation, the avoided costs are re-evaluated. This process continues until the
change in avoided cost curve is immaterial.

Avoided Cost
The SENDOUT® model determined avoided-cost figures represent the unit cost to serve
the next unit of demand with a supply-side resource option during a given period. If a
conservation measure’s total resource cost (for Idaho and Oregon), or utility cost (for
Washington), is less than this avoided cost, it will be cost effective to reduce customer
demand and Avista can avoid commodity, storage, transportation and other supply
resource costs.
SENDOUT® calculates marginal cost data by day, month and year for each demand area.
A summary graphical depiction of avoided annual and winter costs for the
Washington/Idaho and Oregon areas is in Figure 5.16. The detailed data is in Appendix
5.4 – Avoided Cost Details. Other than the carbon tax adder embedded in the expected
price curve, avoided costs do not include additional environmental externality adders for
adverse environmental impacts. Appendix 3.2 – Environmental Externalities discusses
this concept more fully and includes specific requirements required in modeling for the
Oregon service territory.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 105 of 162
Figure 5.16: Avoided Cost (Includes Commodity & Transport Cost – 2014 vs. 2016 $/Dth)

Conservation Potential
Using the avoided cost thresholds, AEG selected all potential cost effective DSM
programs. Table 5.4 shows potential DSM savings in each region from the selected
conservation potential for the Expected Case. The conservation potential includes
anticipated annual acquisition and is cumulative.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 106 of 162
Table 5.4: Annual and Average Daily Demand Served by Conservation

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 107 of 162
Conservation Acquisition Goals
The avoided cost established in SENDOUT®, the conservation potential selected, and
the amount of therm savings is the basis for determining conservation acquisition goals
and subsequent DSM program implementation planning. Chapter 3 – Demand-Side
Resources has additional details on this process.

Supply-Side Resources
SENDOUT® considers all options entered into the model, determines when and what
resources are needed, and which options are cost effective. Selected resources represent
the best cost/risk solution, within given constraints, to serve anticipated customer
requirements. Since the Expected Case has no resource additions in the planning
horizon, Avista will continue to review and refine knowledge of resource options and will
act to secure best cost/risk options when necessary or advantageous.

Resource Utilization
Avista’s plans to meet firm customer demand requirements in a cost-effective manner.
This goal encompasses a range of activities from meeting peak day requirements in the
winter to acting as a responsible steward of resources during periods of lower resource
utilization. As the analysis presented in this IRP indicates, Avista has ample resources to
meet highly variable demand under multiple scenarios, including peak weather events.
Avista acquired the majority of its upstream pipeline capacity during the deregulation or
unbundling of the natural gas industry. Pipelines were required to allocate capacity and
costs to their existing customers as they transitioned to transportation only service
providers. The FERC allowed a rate structure for pipelines to recover costs through a
Straight Fixed Variable rate design. This structure is based on a higher reservation charge
to cover pipeline costs whether natural gas is transported or not, and a much smaller
variable charge which is incurred only when natural gas is transported. An additional fuel
charge is assessed to account for the compressors required to move the natural gas to
customers. Avista maintains enough firm capacity to meet peak day requirements under
the Expected Case in this IRP. This requires pipeline capacity contracts at levels in
excess of the average and above minimum load requirements. Given this load profile and
the Straight Fixed Variable rate design, Avista incurs ongoing pipeline costs during non-
peak periods.
Avista chooses to have an active, hands-on management of resources to mitigate
upstream pipeline and commodity costs for customers when the capacity is not utilized
for system load requirements. This management simultaneously deploys multiple long
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 108 of 162
and short-term strategies to meet firm demand requirements in a cost effective manner.
The resource strategies addressed are:

 Pipeline contract terms;
 Pipeline capacity;
 Storage;
 Commodity and transport optimization; and
 Combination of available resources.

Pipeline Contract Terms
Pipeline costs are incurred whether the capacity is utilized or not. Winter demand must
be satisfied and peak days must be met. Ideally, capacity could be contracted from
pipelines only for the time and days it is required. Unfortunately, this is not how pipelines
are contracted or built. Long-term agreements at fixed volumes are the usual
requirements for building or acquiring firm transport. This assures the pipeline of long-
term, reasonable cost recovery.
Avista has negotiated and contracted for several seasonal transportation agreements.
These agreements allow volumes to increase during the demand intensive winter months
and decrease over the lower demand summer period. This is a preferred contracting
strategy because it eliminates costs when demand is low. Avista refers to this as a front
line strategy because it attempts to mitigate costs prior to contracting the resource. Not
all pipelines offer this option. Avista seeks this type of arrangement when available. Avista
currently has some seasonal transportation contracts on TransCanada GTN,
TransCanada BC and TransCanada Alberta. These pipelines match up transport capacity
to move natural gas from Alberta (AECO) to Avista’s service territories. Avista also
contracted for TF2 on NWP. This is a storage specific contract and matches up the
withdrawal capacity at Jackson Prairie with pipeline transport to Avista’s service
territories. TF2 is a firm service and allows for contracting a daily amount of transportation
for a specified number of days rather than a daily amount on an annual basis as is usually
required. For example, one of the TF2 agreements allows Avista to transport 91,200
Dth/day for 31 days. This is a more cost effective strategy for storage transport than
contracting for an annual amount. Through NWP’s tariff, Avista maintains an option to
increase and decrease the number of days this transportation option is available. More
days correspond to increased costs, so balancing storage, transport and demand is
important to ensure an optimal blend of cost and reliability.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 109 of 162
Pipeline Capacity
After contracting for pipeline capacity, its management and utilization determine the
actual costs. The worst-case economic scenario is to do nothing and simply incur the
costs associated with this transport contract over the long-term to meet current and future
peak demand requirements. Avista develops strategies to ensure this does not happen
on a regular basis.

Capacity Release
Through the pipeline unbundling of transportation, the FERC establishes rules and
procedures to ensure a fair market developed to manage pipeline capacity as a
commodity. This evolved into the capacity release market and is governed by FERC
regulations through individual pipelines. The pipelines implement the FERC’s posting
requirements to ensure a transparent and fair market is maintained for the capacity. All
capacity releases are posted on the pipelines Bulletin Boards and, depending on the
terms, may be subject to bidding in an open market. This provides the transparency
sought by the FERC in establishing the release requirements. Avista utilizes the capacity
release market to manage both long-term and short-term transportation capacity.
For capacity under contract that may exceed current demand, Avista seeks other parties
that may need it and arranges for capacity releases to transfer rights, obligations and
costs. This shifts all or a portion of the costs away from Avista’s customers to a third party
until it is needed to meet customer demand.
There are many variables in determining the value of transportation. Certain pipeline
paths are more valuable and this can vary by year, season, month and day. The term,
volume and conditions precedent also contribute to the value recoverable through a
capacity release. For example, a release of winter capacity to a third party may allow for
full cost recovery; while a release for the same period that allows Avista to recall the
capacity for up to 10 days during the winter may not be as valuable to the third party, but
of high value to us. Avista may be willing to offer a discount to retain the recall rights
during high demand periods. This turns a seasonal-for-annual cost into a peaking-only
cost. These are market terms and conditions that are negotiated to determine the value
or discount required by both parties.
Avista has several long-term releases, some extending through 2025 providing full
recovery of all the pipeline costs. These releases maintain Avista’s long-term rights to the
transportation capacity without incurring the costs of waiting until demand increases. As
the end of these release terms near, Avista surveys the market against the IRP to
determine if these contracts should be reclaimed or released, and for what duration.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 110 of 162
Avista has releases to third parties that terminate in 2016. Results of this IRP show that
this capacity is not needed in 2016 as originally anticipated, and Avista is negotiating new
terms and conditions to continue full cost recovery until it is required. Through this
process, Avista retains the rights to vintage capacity without incurring the costs or having
to participate in future pipeline expansions that will cost more than current capacity.
On a shorter term, excess capacity not fully utilized on a seasonal, monthly or daily basis
can also be released. Market conditions often dictate less than full cost recovery for
shorter-term requirements. Mitigating some costs for an unutilized, but required resource
reduces costs to our customers.

Segmentation
Through a process called segmentation, Avista creates new firm pipeline capacity for the
service territory. This doubles some of the capacity volumes at no additional cost to
customers. With increased firm capacity, Avista can continue some long-term releases,
or even reduce some contract levels, if the release market does not provide adequate
recovery.

Storage
As a one-third owner of the Jackson Prairie Storage facility, Avista holds an equal share
of capacity (space available to store gas) and delivery (the amount of natural gas that can
be withdrawn on a daily basis).
Storage allows lower summer-priced gas to be stored and used in the winter during high
demand or peak day events. Similar to transportation, unneeded capacity and delivery
can be optimized by selling into a higher priced market. This allows Avista to manage
storage capacity and delivery to meet growing peak day requirements when needed.
The injection of natural gas into storage during the summer utilizes existing pipeline
transport and helps increase the utilization factor of pipeline agreements. Avista employs
several storage optimization strategies to mitigate costs. Revenue from this activity flows
through the annual PGA/Deferral process.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 111 of 162
Commodity and Transportation Optimization
Another strategy to mitigate transportation costs is to participate in the daily market to
assess if unutilized capacity has value. Avista seeks daily opportunities to purchase gas,
transport it on existing unutilized capacity, and sell it into a higher priced market to capture
the cost of the gas purchased and recover some pipeline charges. The recovery is market
dependent and may or may not recover all pipeline costs, but mitigates pipeline costs to
customers.

Combination of Resources Unutilized resources like supply, transportation, storage and capacity can combine to
create products that capture more value than the individual pieces. Avista has structured
long-term arrangements with other utilities that allow available resources utilization and
provide products that no individual component can satisfy. These products provide more
cost recovery of the fixed charges incurred for the resources while maintaining the rights
to utilize the resource for future customer needs.

Resource Utilization Summary
As determined through the IRP modeling of demand and existing resources, new
resources under the Expected Case are not required over the next 20 years. Avista
manages the existing resources to mitigate the costs incurred by customers until the
resource is required to meet demand. The recovery of costs is often market based with
rules governed by the FERC. Avista is recovering full costs on some resources and partial
costs on others. The management of long- and short-term resources meets firm customer
demand in a reliable and cost-effective manner.

Gate Station Analysis
In past IRP’s, Avista identified a risk associated with the aggregated methodology for
supply and demand forecasting in previous IRPs. The forecasting methodology is
consistent with operational practices which aggregate capacity at individual points for
scheduling/nomination purposes. Typically, the amount of natural gas that can flow from
a contract demand (i.e., receipt/supply quantity) is fixed and the deliverable amount
(i.e., maximum daily delivery obligation or delivery quantity) to gate stations is greater.
(See Figure 5.17) However, aggregation could mask deficiencies at individual gate
stations.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 112 of 162
Figure 5.17: Gate Station Modeling Challenge

To address this concern, a gate-by-gate analysis was developed outside of SENDOUT®.
The analysis involved coordination between Gas Supply, Gas Engineering and intrastate
pipeline personnel. Utilizing historical gate station flow data and demand forecasting
methodologies detailed in the IRP, forecasted peak-day gate station demand was
calculated. This demand was compared to contracted and operational capacities at each
gate station.
If forecasted demand exceeded contracted and/or operational capacities, further analysis
was completed. The additional analysis involved assessing the economic way to address
the gate deficiency. This could involve a gate station expansion, reassigning maximum
daily delivery obligations, targeted DSM programs or distribution system enhancements.
Avista has completed the gate station analysis. The analysis found peak day deficiencies
at seven gate stations. The data set was static so some of the gate station shortages
have already experienced mitigation fixes. The area of La Grande is one such example
where the Ladd Canyon city gate station was completed in December 2015 to help offset
the flow at the La Grande city gate. Also, a high pressure reinforcement project will begin
in 2017 to help ensure a physical limitation at this city gate is not exceeded. As for the
remaining city gate deficiencies, the potential fixes will be reviewed between Avista’s Gas
Supply and Distribution Engineering departments. A current list of distribution planning
capital projects can be found in table 7.1 and city gate station upgrades in table 7.2
(section as described in Chapter 7 – Distribution Planning).

10,000
2000
Contract Demand: 10,000
Supply (Receipt Quantity)
MDDO’s: 11,000
Gate Station (Delivery
Quantity)
3000
4000
2000
1000
1000
4000
2000
Demand: 8,000
Behind the Gate
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 113 of 162
Conclusion
The IRP portfolio analysis summarized in this chapter was performed on the Average
Case and then on the Expected Case demand scenario. Although the results show no
resource deficiencies during the 20-year forecasted term, Avista has chosen to utilize the
Expected Case for peak operational planning activities because this case is the most
likely outcome given experience, industry knowledge and understanding of future natural
gas markets. This case provides reasonable demand growth given current expectations
of natural gas prices over the planning horizon. If realized, this case allows Avista
protection against resource shortages and does not over commit to additional long-term
resources.
Avista recognizes that there are other potential outcomes. The process described in this
chapter applies to the alternate demand and supply resource scenarios covered in
Chapter 6 – Alternate Scenarios, Portfolios and Stochastic Analysis.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 114 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 115 of 162
6: Alternate Scenarios, Portfolios and Stochastic
Analysis

Overview
Avista applied the IRP analysis in Chapter 5 – Integrated Resource Portfolio to alternate
demand and supply resource scenarios to develop a range of alternate portfolios. This
deterministic modeling approach considered different underlying assumptions vetted with
the TAC members to develop a consensus about the number of cases to model.
Avista also performed stochastic modeling for estimating probability distributions of
potential outcomes by allowing for random variation in natural gas prices and weather
based on fluctuations in historical data. This statistical analysis, in conjunction with the
deterministic analysis, enabled statistical quantification of risk from reliability and cost
perspectives related to resource portfolios under varying price and weather conditions.

Alternate Demand Scenarios
As discussed in the Demand Forecasting section, Avista identified alternate scenarios for
detailed analysis to capture a range of possible outcomes over the planning horizon.
Table 6.1 summarizes these scenarios and Chapter 2 – Demand Forecasts and
Appendices 2.6 and 2.7 describes them in detail. The scenarios consider different
demand influencing factors and price elasticity effects for various price influencing factors.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 116 of 162
Table 6.1: 2016 IRP Scenarios
Proposed Scenarios Expected Expected High Growth Low Growth Cold Day 20yr Average
INPUT ASSUMPTIONS Case - Low Prices Case & Low Prices & High Prices Weather Std Case
Customer Growth Rate Reference Case
Cust Growth Rates
Reference Case
Cust Growth Rates High Growth Rate Low Growth Rate Reference Case
Cust Growth Rates
Reference Case
Cust Growth Rates
Use per Customer 3 yr Flat +3 yr Flat +3 yr Flat +3 yr Flat +3 yr Flat +3 yr Flat +
Price Elast.Price Elast.Price Elast. +Price Elast.Price Elast.Price Elast.
CNG/NGV
Demand Side Management Yes Yes Yes Yes Yes Yes
Weather Planning Standard Coldest Day Coldest Day Coldest Day Coldest Day
Alternate Planning
Standard Normal
Prices
Price curve Low Expected Low High Expected Expected
Carbon Legislation ($/Ton) $9.89 - 19.93 $9.89 - 19.93 None $9.89 - 19.93 $9.89 - 19.93 $9.89 - 19.93
First Gas Year Unserved
WA/ID N/A N/A 2033 N/A N/A N/A
Medford N/A N/A 2027 N/A N/A N/A
Roseburg N/A N/A 2027 N/A N/A N/A
Klamath N/A N/A 2034 N/A N/A N/A
La Grande N/A N/A 2031 N/A N/A N/A
RESULTS

Demand profiles over the planning horizon for each of the scenarios shown in Figures 6.1
and 6.2 reflect the two winter peaks modeled for the different service territories (Dec. 20
and Feb. 15).
Figure 6.1 Peak Day (Feb 15) – 2016 IRP Demand Scenarios

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 117 of 162
Figure 6.2 Peak Day (Dec 20) – 2016 IRP Demand Scenarios

As in the Expected Case, Avista used SENDOUT® to model the same resource
integration and optimization process described in this section for each of the six demand
scenarios (see Appendix 2.7 for a complete listing of portfolios considered). This identified
the first year unserved dates for each scenario by service territory shown in Figure 6.3.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 118 of 162
Figure 6.3: First Year Peak Demand Not Met with Existing Resources

As anticipated, the High Growth & Low Price scenario has the most rapid growth and the
earliest first year unserved dates. This scenario includes customer growth rates higher
than the Expected Case, incremental demand driven by emerging markets and no
adjustment for price elasticity. Even with aggressive assumptions, resource shortages do
not occur until late in the planning horizon.

 2033 in Washington/Idaho
 2027 in Medford/Roseburg
 2034 in Klamath Falls
 2031 in La Grande

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 119 of 162
The model chose to solve these unserved demand areas via the following supply
resources:
 Washington/Idaho – Increase contracting on Alberta System, Foothills, and GTN
pipeline by 13,000 Dth/day.
 Medford/Roseburg – Add an upsized compressing station on the Medford Lateral
increasing deliverability by 50,000 Dth/day.
 Klamath Falls – Increase the Operating Pressure on the Klamath Falls Lateral.
 La Grande – Increase contract delivery on Northwest Pipeline.
Steeper demand highlights the flat demand risk discussed earlier. The likelihood of this
scenario occurring is remote due to a yearly recurrence of coldest day on record weather
paired with a much steeper growth of customer population; however, any potential for
accelerated unserved dates warrants close monitoring of demand trends and resource
lead times as described in the Ongoing Activities section of Chapter 8 – Action Plan. The
remaining scenarios do not identify resource deficiencies in the planning horizon.
Due to their importance and connection with the IRP process, additional detailed
information on certain selected scenarios is included in the following appendices:
 Demand and Existing Resources graphs by service territory (High Growth Case
only) – Appendix 6.1
 Peak Day Demand, Served and Unserved table (all cases) – Appendix 6.2

Alternate Supply Resources
Avista identified supply-side resources that could meet resource deficiencies or provide
a least cost solution. There are other options Avista considered in its modeling approach
to solve for High Growth & Low Price unserved conditions and to determine whether the
Expected Case with existing resources is least cost/least risk. Some of the currently
available resources are included in Table 6.2 and potential future resources are included
in Table 6.3:

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 120 of 162
Table 6.2: Available Supply Resources

Table 6.3: Future Supply Resources

For example, contracted city gate deliveries in the form of a structured purchase
transaction could meet peak conditions. However, the market-based price and other
terms are difficult to reliably determine until a formal agreement is negotiated. Exchange
agreements also have market-based terms and are hard to reliably model when the
resource need is later in the planning horizon.
Many of the potential resources are not yet commercially available or well tested,
technically making them speculative. Resources such as coal-bed methane, LNG imports
and natural gas hydrates would fall into this category. Avista will continue to monitor all
resources and assess their appropriateness for inclusion in future IRPs as described in
Chapter 8 – Action Plan.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 121 of 162
One resource which will be closely observed is exported LNG. While Avista considered
LNG exports, it was primarily as a price-influencing factor. However, if the proposed
export LNG terminal in Oregon is approved and a pipeline built to supply that facility, it
potentially could bring new supply through Avista’s service territory. Avista will monitor
(Chapter 8 – Action Plan) this situation through industry publications and daily operations
to consider inclusion of this supply scenario for future IRPs.

Deterministic – Portfolio Evaluation
There is no resource deficiency identified in the planning period and the existing resource
portfolio is adequate to meet forecasted demand. The alternate demand scenarios and
supply scenarios are placed in the model as predicted future conditions that the supply
portfolio will have to satisfy via least cost and least risk strategies. This creates bounds
for analyzing the Expected Case by creating high and low boundaries for customer count,
weather and pricing. Each portfolio runs through SENDOUT® where the supply resources
(Chapter 4 – Supply Side Resources) and conservation resources (Chapter 3 – Demand
Side Management – see tables 3.2, 3.3 and 3.4) are compared and selected on a least
cost basis. Once new resources are determined, a net present value of the revenue
requirement (PVRR) is calculated.
Table 6.4 summarizes the PVRR of the portfolios considered. In addition to the portfolios
in table 6.4, modeling was done to compare the existing resources to alternate resources.
The analysis begins as deterministic based on the Expected Case portfolio. Existing
transportation was tested in small increments to show the overall change to the PVRR.
In the first alternate resource cases, 10,000 Dth/day was taken from NWP and then an
equal amount added to GTN. The second case reversed these amounts where 10,000
additional Dth/day were added to NWP and taken away from GTN. Both NWP and GTN
quantities are flowing into the WA-ID service territory. The third case removes transport
to Oregon on NWP, from Sumas, and is replaced with an enhanced compressor on the
Medford lateral that connects to the GTN mainline. All three alternate resource portfolios
show a higher cost than the Expected Case PVRR. This testing of existing resources in
comparison to alternate resources determines whether a least cost and least risk solution
is present in the existing resource stack. If any of these scenarios showed a lower
deterministic system PVRR, a stochastic analysis would be used to look at the optimal
resource stack and its cost under varying conditions.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 122 of 162
Table 6.4: PVRR by Portfolio

Stochastic Analysis1
The scenario (deterministic) analysis described earlier in this chapter represents specific
what if situations based on predetermined assumptions, including price and weather.
These factors are an integral part of scenario analysis. To understand a particular
portfolio’s response to cost and risk, through price and weather, Avista applied stochastic
analysis to generate a variety of price and weather events.
Deterministic analysis is a valuable tool for selecting an optimal portfolio. The model
selects resources to meet peak weather conditions in each of the 20 years. However, due
to the recurrence of design conditions in each of the 20 years, total system costs over the
planning horizon can be overstated because of annual recurrence of design conditions
and the recurrence of price increases in the forward price curve. As a result, deterministic
analysis does not provide a comprehensive look at future events. Utilizing Monte Carlo
simulation in conjunction with deterministic analysis provides a more complete picture of
portfolio performance under multiple weather and price profiles.
1 SENDOUT® uses Monte Carlo simulation to support stochastic analysis, which is a mathematical technique for evaluating risk and uncertainty. Monte Carlo simulation is a statistical modeling method
used to imitate future possibilities that exist with a real-life system.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 123 of 162
This IRP employs stochastic analysis in two ways. The first tested the weather-planning
standard and the second assessed risk related to costs of our Expected Case (existing
portfolio) under varying price environments. The Monte Carlo simulation in SENDOUT®
can vary index price and weather simultaneously. This simulates the effects each have
on the other.

Weather In order to evaluate weather and its effect on the portfolio, Avista developed 200
simulations (draws) through SENDOUT®’s stochastic capabilities. Unlike deterministic
scenarios or sensitivities, the draws have more variability from month-to-month and year-
to-year. In the model, random monthly total HDD draw values (subject to Monte Carlo
parameters – see Table 6.5) are distributed on a daily basis for a month in history with
similar HDD totals. The resulting draws provide a weather pattern with variability in the
total HDD values, as well as variability in the shape of the weather pattern. This provides
a more robust basis for stress testing the deterministic analysis.

Table 6.5: Example of Monte Carlo Weather Inputs – Spokane

Avista models five weather areas: Spokane, Medford, Roseburg, Klamath Falls and La
Grande. Avista assessed the frequency that the peak day occurs in each area from the
simulation data. The stochastic analysis shows that in over 200, 20-year simulations,
peak day (or more) occurs with enough frequency to maintain the current planning
standard for this IRP. This topic remains a subject of continued analysis. For example,
the Medford weather pattern over the 200 20-year draws (i.e, 4,000 years). HDDs at or
above peak weather (61 HDDs) occur 128 times. This equates to a peak day occurrence
once every 31 years (4,000 simulation years divided by 128 occurrences). The Spokane
area has the least occurrences of peak day (or more) occurrences and La Grande has
the most occurrences. This is primarily due to the frequency in which each region’s peak
day HDD occurs within the historical data, as well as near peak day HDDs. See Figures
6.4 through 6.8 for the number of peak day occurrences by weather area.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 124 of 162
Figure 6.4: Frequency of Peak Day Occurrences – Spokane

Figure 6.5: Frequency of Peak Day Occurrences – Medford

0
1
2
3
1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154 163 172 181 190 199
#
o
f

P
e
a
k

D
a
y

O
c
c
u
r
r
e
n
c
e
s
Draws
0
1
2
3
1 10 19 28 37 46 55 64 73 82 91
10
0
10
9
11
8
12
7
13
6
14
5
15
4
16
3
17
2
18
1
19
0
19
9
#
o
f

P
e
a
k

D
a
y

O
c
c
u
r
r
e
n
c
e
s
Draws
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 125 of 162
Figure 6.6: Frequency of Peak Day Occurrences – Roseburg

Figure 6.7: Frequency of Peak Day Occurrences – Klamath Falls

0
1
2
3
1 10 19 28 37 46 55 64 73 82 91
10
0
10
9
11
8
12
7
13
6
14
5
15
4
16
3
17
2
18
1
19
0
19
9
#
o
f

P
e
a
k

D
a
y

O
c
c
u
r
r
e
n
c
e
s
Draws
0
1
2
3
4
5
1 10 19 28 37 46 55 64 73 82 91
10
0
10
9
11
8
12
7
13
6
14
5
15
4
16
3
17
2
18
1
19
0
19
9
#
o
f

P
e
a
k

D
a
y

O
c
c
u
r
r
e
n
c
e
s
Draws
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 126 of 162
Figure 6.8: Frequency of Peak Day Occurrences – La Grande

Price
While weather is an important driver for the IRP, price is also important. As seen in recent
years, significant price volatility can affect the portfolio. In deterministic modeling, a single
price curve for each scenario is used for analysis. There is risk that the price curve in the
scenario will not reflect actual results.
Avista used Monte Carlo simulation to test the portfolio and quantify the risk to customers
when prices do not materialize as forecast. Avista performed a simulation of 200 draws,
varying prices, to investigate whether the Expected Case total portfolio costs from the
deterministic analysis is within the range of occurrences in the stochastic analysis. Figure
6.9 shows a histogram of the total portfolio cost of all 200 draws, plus the Expected Case
results. This histogram depicts the frequency and the total cost of the portfolio among all
the draws, the mean of the draws, the standard deviation of the total costs, and the total
costs from the Expected Case. The figure confirms that Expected Case total portfolio cost
is within an acceptable range of total portfolio costs based on 200 unique pricing
scenarios.

-
1
2
3
4
5
6
7
8
1 10 19 28 37 46 55 64 73 82 91
10
0
10
9
11
8
12
7
13
6
14
5
15
4
16
3
17
2
18
1
19
0
19
9
#
o
f

P
e
a
k

D
a
y

O
c
c
u
r
r
e
n
c
e
s
Draws
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 127 of 162
Figure 6.9: 2016 IRP Total 20-Year Cost

Performing stochastic analysis on weather and price in the demand analysis provided a
statistical approach to evaluate and confirm the findings in the scenario analysis with
respect to adequacy and reasonableness of the weather-planning standard and the
natural gas price forecast. This analytical perspective provides confidence in the
conclusions and stress tests the robustness of the selected portfolio of resources, thereby
mitigating analytical risks.

Regulatory Requirements
IRP regulatory requirements in Idaho, Oregon and Washington call for several key
components. The completed plan must demonstrate that the IRP:

 Examines a range of demand forecasts.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 128 of 162
 Examines feasible means of meeting demand with both supply-side and demand-
side resources.
 Treats supply-side and demand-side resources equally.
 Describes the long-term plan for meeting expected demand growth.
 Describes the plan for resource acquisitions between planning cycles.
 Takes planning uncertainties into consideration.
 Involves the public in the planning process.

Avista addressed the applicable requirements throughout this document. Appendix 1.2 –
IRP Guideline Compliance Summaries lists the specific requirements and guidelines of
each jurisdiction and describes Avista’s compliance.
The IRP is also required to consider risks and uncertainties throughout the planning and
analytical processes. Avista’s approach in addressing this requirement was to identify
factors that could cause significant deviation from the Expected Case planning
conclusions. This included dynamic demand analytical methods and sensitivity analysis
on demand drivers that impacted demand forecast assumptions. From this, Avista
created 16 demand sensitivities and modeled five demand scenario alternatives, which
incorporated different customer growth, use-per-customer, weather, and price elasticity
assumptions.
Avista analyzed peak day weather planning standard, performing sensitivity on HDDs and
modeling an alternate weather-planning standard using the coldest day in 20 years.
Stochastic analysis using Monte Carlo simulations in SENDOUT® supplemented this
analysis. Avista also used simulations from SENDOUT® to analyze price uncertainty and
the effect on total portfolio cost.
Avista examined risk factors and uncertainties that could affect expectations and
assumptions with respect to DSM programs and supply-side scenarios. From this, Avista
assessed the expected available supply-side resources and potential conservation
savings for evaluation.
The investigation, identification, and assessment of risks and uncertainties in our IRP
process should reasonably mitigate surprise outcomes.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 129 of 162
Conclusion
The Expected Case has the lowest cost and risk when considering alternate supply
resources. This is primarily due to Avista’s geographic location, the longstanding
subscription to these pipeline services as well as the supply basins/storage facilities
available to secure supply. Avista has geographically disparate areas where in many
cases only one pipeline is available to deliver supply. The cost of building or acquiring
new supply resources would likely increase cost while keeping risk at similar levels.
The High Growth and Low Growth Case demand analyses provide a range for evaluating
demand trajectories relative to the Expected Case. Based on this analysis there appears
to be sufficient time to plan for forecasted resource needs. Even under an extreme growth
scenario, the first forecasted deficiency does not occur until 2027. Many things could
happen between now and when the first resource needs occur, so Avista will carefully
monitor (Chapter 8 – Action Plan) demand trends through reconciling and comparing
forecast to actual customer counts and continually update and evaluate all demand-side
and supply-side alternatives.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 130 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 131 of 162
7: Distribution Planning

Overview
Avista’s IRP evaluates the safe, economical and reliable full-path delivery of natural gas
from basin to the customer meter. Securing adequate natural gas supply and ensuring
sufficient pipeline transportation capacity to Avista’s city gates become secondary issues
if distribution system growth behind the city gates increases faster than expected and the
system becomes severely constrained. Important parts of the distribution planning
process include forecasting local demand growth, determining potential distribution
system constraints, analyzing possible solutions and estimating costs for eliminating
constraints.
Analyzing resource needs to this point has focused on ensuring adequate capacity to the
city gates, especially during a peak event. Distribution planning focuses on determining if
there will be adequate pressure during a peak hour. Despite this altered perspective,
distribution planning shares many of the same goals, objectives, risks and solutions as
integrated resource planning.
Avista’s natural gas distribution system consists of approximately 3,300 miles of
distribution main and services pipelines in Idaho, 3,700 miles in Oregon and 5,800 miles
in Washington; as well as numerous regulator stations, service distribution lines,
monitoring and metering devices, and other equipment. Currently, there are no storage
facilities or compression systems within Avista’s distribution system. Distribution network
pipelines and regulating stations operate and maintain system pressure solely from the
pressure provided by the interstate transportation pipelines.

Distribution System Planning
Avista conducts two primary types of evaluations in its distribution system planning
efforts: capacity requirements and integrity assessments.
Capacity requirements include distribution system reinforcements and expansions.
Reinforcements are upgrades to existing infrastructure, or new system additions, which
increase system capacity, reliability and safety. Expansions are new system additions to
accommodate new demand. Collectively, these are distribution enhancements.
Ongoing evaluations of each distribution network in the four primary service territories
identify strategies for addressing local distribution requirements resulting from customer
growth. Customer growth assessments are made based on factors including IRP demand
forecasts, monitoring gate station flows and other system metering, new service requests,
field personnel discussion, and inquiries from major developers.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 132 of 162
Avista regularly conducts integrity assessments of its distribution systems. Ongoing
system evaluation can also indicate distribution-upgrading requirements for system
maintenance needs rather than customer and load growth. In some cases, the timing for
system integrity upgrades coincides with growth-related expansion requirements.
These planning efforts provide a long-term planning and strategy outlook and integrate
into the capital planning and budgeting process, which incorporates planning for other
types of distribution capital expenditures and infrastructure upgrades.
Gas Engineering planning models are also compared with capacity limitations at each
city gate station. Referred to as city gate analysis, the design day hourly demand
generated from planning analyses must not exceed the actual physical limitation of the
city gate station. A capacity deficiency found at a city gate station establishes a potential
rebuild or addition of a new city gate station.

Network Design Fundamentals
Natural gas distribution networks rely on pressure differentials to flow natural gas from
one place to another. When pressures are the same on both ends of a pipe, the natural
gas does not move. As natural gas exits the pipeline network, it causes a pressure drop
due to its movement and friction. As customer demand increases, pressure losses
increase, reducing the pressure differential across the pipeline network. If the pressure
differential is too small, flow stalls and the network could run out of pressure.
It is important to design a distribution network such that intake pressure from gate stations
and/or regulator stations within the network is high enough to maintain an adequate
pressure differential when natural gas leaves the network.
Not all natural gas flows equally throughout a network. Certain points within the network
constrain flow and restrict overall network capacity. Network constraints can occur as
demand requirements evolve. Anticipating these demand requirements, identifying
potential constraints and forming cost-effective solutions with sufficient lead times without
overbuilding infrastructure are the key challenges in network design.

Computer Modeling
Developing and maintaining effective network design is aided by computer modeling for
network demand studies. Demand studies have evolved with technology in the past
decade to become a highly technical and powerful means of analyzing distribution system
performance. Using a pipeline fluid flow formula, a specified parameter for each pipe
element can be simultaneously solved. Many pipeline equations exist, each tailored to a
specific flow behavior. Through years of research, these equations have been refined to
the point where modeling solutions closely resemble actual system behavior.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 133 of 162
Avista conducts network load studies using GL Noble Denton’s Synergi software. This
modeling tool allows users to analyze and interpret solutions graphically.

Determining Peak Demand
Avista’s distribution network is comprised of high pressure (90-500 psig) and intermediate
pressure (5-60 psig) mains. Avista operates its intermediate networks at a relatively low
maximum pressure of 60 psig or less for ease of maintenance and operation, public
safety, reliable service and cost considerations. Since most distribution systems operate
through relatively small diameter pipes, there is essentially no line-pack capability for
managing hourly demand fluctuations. Line pack is the difference between the natural
gas contents of the pipeline under packed (fully pressurized) and unpacked
(depressurized) conditions. Line pack is negligible in Avista’s distribution system due to
the smaller diameter pipes and lower pressures. In transmission and inter-state pipelines,
line-pack contributes to the overall capacity due to the larger diameter pipes and higher
operating pressures.
Core demand typically has a morning peaking period between 6 a.m. and 10 a.m. and
the peak hour demand for these customers can be as much as 50 percent above the
hourly average of daily demand. Because of the importance of responding to hourly
peaking in the distribution system, planning capacity requirements for distribution systems
uses peak hour demand.1

Distribution System Enhancements
Demand studies facilitate modeling multiple demand forecasting scenarios, constraint
identification and corresponding optimum combinations of pipe modification, and
pressure modification solutions to maintain adequate pressures throughout the network.
Distribution system enhancements do not reduce demand nor do they create additional
supply. Enhancements can increase the overall capacity of a distribution pipeline system
while utilizing existing gate station supply points. The two broad categories of distribution
enhancement solutions are pipelines and regulators.

Pipelines
Pipeline solutions consist of looping, upsizing and uprating. Pipeline looping is the most
common method of increasing capacity in an existing distribution system. It involves
constructing new pipe parallel to an existing pipeline that has, or may become, a
constraint point. Constraint points inhibit flow capacities downstream of the constraint
1 This method differs from the approach that Avista uses for IRP peak demand planning, which focuses
on peak day requirements to the city gate.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 134 of 162
creating inadequate pressures during periods of high demand. When the parallel line
connects to the system, this alternative path allows natural gas flow to bypass the original
constraint and bolsters downstream pressures. Looping can also involve connecting
previously unconnected mains. The feasibility of looping a pipeline depends upon the
location where the pipeline will be constructed. Installing gas pipelines through private
easements, residential areas, existing paved surfaces, and steep or rocky terrain can
increase the cost to a point where alternative solutions are more cost effective.
Pipeline upsizing involves replacing existing pipe with a larger size pipe. The increased
pipe capacity relative to surface area results in less friction, and therefore a lower
pressure drop. This option is usually pursued when there is damaged pipe or where pipe
integrity issues exist. If the existing pipe is otherwise in satisfactory condition, looping
augments existing pipe, which remains in use.
Pipeline uprating increases the maximum allowable operating pressure of an existing
pipeline. This enhancement can be a quick and relatively inexpensive method of
increasing capacity in the existing distribution system before constructing more costly
additional facilities. However, safety considerations and pipe regulations may prohibit the
feasibility or lengthen the time before completion of this option. Also, increasing line
pressure may produce leaks and other pipeline damage creating costly repairs. A
thorough review is conducted to ensure pipeline integrity before pressure is increased.

Regulators
Regulators, or regulator stations, reduce pipeline pressure at various stages in the
distribution system. Regulation provides a specified and constant outlet pressure before
natural gas continues its downstream travel to a city’s distribution system, customer’s
property or natural gas appliance. Regulators also ensure that flow requirements are met
at a desired pressure regardless of pressure fluctuations upstream of the regulator.
Regulators are at city gate stations, district regulator stations, farm taps and customer
services.
Compression
Compressor stations present a capacity enhancing option for pipelines with significant
natural gas flow and the ability to operate at higher pressures. For pipelines experiencing
a relatively high and constant flow of natural gas, a large volume compressor installation
along the pipeline boosts downstream pressure.
A second option is the installation of smaller compressors located close together or
strategically placed along a pipeline. Multiple compressors accommodate a large flow
range and use smaller and very reliable compressors. These smaller compressor stations
are well suited for areas where natural gas demand is growing at a relatively slow and
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 135 of 162
steady pace, so that purchasing and installing these less expensive compressors over
time allows a pipeline to serve growing customer demand into the future.
Compressors can be a cost effective option to resolving system constraints; however,
regulatory and environmental approvals to install a compressor station, along with
engineering and construction time can be a significant deterrent. Adding compressor
stations typically involves considerable capital expenditure. Based on Avista’s detailed
knowledge of the distribution system, there are no foreseeable plans to add compressors
to the distribution network.

Conservation Resources
The evaluation of distribution system constraints includes consideration of targeted
conservation resources to reduce or delay distribution system enhancements. The
consumer is still the ultimate decision-maker regarding the purchase of a conservation
measure. Because of this, Avista attempts to influence conservation through the DSM
measures discussed in Chapter 3 – Demand-Side Resources, but does not depend on
estimates of peak day demand reductions from conservation to eliminate near-term
distribution system constraints. Over the longer-term, targeted conservation programs
may provide a cumulative benefit that could offset potential constraint areas and may be
an effective strategy.

Distribution Scenario Decision-Making Process
After achieving a working load study, analyses are performed on every system at design
day conditions to identify areas where potential outages may occur.
Avista’s design HDD for distribution system modeling is determined using the coldest day on record for each given service area. This practice is consistent with the peak day
demand forecast utilized in other sections of Avista’s natural gas IRP.
Utilizing a peak planning standard of the coldest temperature on record may seem aggressive given a temperature experienced rarely, or only once. Given the potential
impacts of an extreme weather event on customers’ personal safety and property damage to customer appliances and Avista’s infrastructure, it is a prudent regionally accepted
planning standard.

These areas of concern are then risk ranked against each other to ensure the highest risk
areas are corrected first. Within a given area, projects/reinforcements are selected using
the following criteria:

 The shortest segment(s) of pipe that improves the deficient part of the distribution
system.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 136 of 162
 The segment of pipe with the most favorable construction conditions, such as
ease of access or rights or traffic issues.
 Minimal to no water, railroad, major highway crossings, etc.
 The segment of pipe that minimizes environmental concerns including minimal to
no wetland involvement, and the minimization of impacts to local communities
and neighborhoods.
 The segment of pipe that provides opportunity to add additional customers.
 Total construction costs including restoration.
Once a project/reinforcement is identified, the design engineer or construction project
coordinator begins a more thorough investigation by surveying the route and filing for permits. This process may uncover additional impacts such as moratoriums on road
excavation, underground hazards, discontent among landowners, etc., resulting in another iteration of the above project/reinforcement selection criteria. Figure 7.1 provides
a schematic representation of the distribution scenario process.

Figure 7.1 Distribution Scenario Process

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 137 of 162
An example of the distribution scenario decision making process is from the Medford high
pressure loop reinforcement where the analysis resulted in multiple paths or pipeline routes. The initial path was based on quantitative factors, specifically the shortest length
and least cost route. However, as field investigations and coordination with local city and county governments began, alternative routes had to be determined to minimize future
conflicts, environmental considerations, and field and community disruptions. The final path was based on several qualitative factors that including:
 Available right-of-way along city streets
 Availability of private easements from property owners
 Restrictions due to City of Medford future planned growth with limited planning
information; and
 Potential to avoid conflict with other utilities including a large electric substation along the initial route.

Planning Results
Table 7.1 summarizes the cost and timing, as of the publication date of this IRP, of major
distribution system enhancements addressing growth-related system constraints, system
integrity issues and the timing of expenditures.
The Distribution Planning Capital Projects criteria includes:
 Prioritized need for system reliability (necessary to maintain reliable service)
 Scale of project (large in magnitude and will require significant engineering
and design support); and
 Budget approval (will require approval for capital funding)
These projects are preliminary estimates of timing and costs of major reinforcement
solutions. The scope and needs of distribution system enhancement projects generally
evolve with new information requiring ongoing reassessment. Actual solutions may differ
due to differences in actual growth patterns and/or construction conditions that differ from
the initial assessment and timing of planned completion may change based on the
aforementioned ongoing reassessment of information.
The following discussion provides information about key near-term projects.
La Grande High Pressure Reinforcement: This project will reinforce the La Grande and
Elgin high pressure distribution system and is fed from the Ladd Canyon Gate Station
which will displace flow and remove the capacity constraint at the La Grande Gate station.
Currently, the distribution system cannot maintain adequate pressures at Elgin during
cold winter conditions. Approximately 16,900 feet of high pressure (HP) steel gas main
will be installed beginning in 2017. This reinforcement will ensure that the physical
limitation at the La Grande city gate is not exceeded.
North Spokane Reinforcement: This project will reinforce the area north of Spokane
along U.S. Highway 2. This mixed-use area experiences low pressure during winter at
unpredictable times given demand profiles of the diverse customer base. Completion of
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 138 of 162
this reinforcement will improve pressures in the U.S. Highway 2 North Kaiser area.
Approximately 10,800 feet of HP steel gas main will be installed in a newly established
easement along U.S. Highway 2. Phase 2 includes approximately 12,400 feet of HP steel
gas main will be installed along the electric transmission easement to improve pressures
on the western end of North Spokane. This project also includes two district regulator
stations. Timing is yet to be determined on Phase 2.

Coeur d’Alene High Pressure Reinforcement: This project will reinforce the Coeur
d’Alene distribution system as well as greatly improve the Hayden Lake distribution
system, which currently cannot maintain adequate pressure during cold winter conditions.
Approximately 17,200 feet of HP steel gas main and two district regulator stations will be
designed in 2016 with construction in 2017 and 2018.

Table 7.1 Distribution Planning Capital Projects
La Grande High Pressure
Reinforcement
$3,500,000
North Spokane Reinforcement $2,000,000
Coeur d’Alene High Pressure
Reinforcement
$250,000 $4,000,000 $4,000,000
Schweitzer Mountain Rd High Pressure
Reinforcement
$1,500,000
*Details of project described in IRP as of August 2016

Table 7.2 shows city gate stations identified as over utilized or under capacity. Estimated
cost, year and the plan to remediate the capacity concern are shown.
These projects are preliminary estimates of timing and costs of city gate station upgrades.
The scope and needs of each project generally evolve with new information requiring
ongoing reassessment. Actual solutions may differ due to differences in actual growth
patterns and/or construction conditions that differ from the initial assessment.

In addition to improving the Coeur d’Alene distribution system, the Coeur d’Alene High
Pressure Reinforcement mentioned above will redirect the flow of gas away from the
capacity constrained city gate stations: CDA East, Post Falls, and CDA West.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 139 of 162
To maintain minimum design pressures to serve Elgin on a design day, two projects were
required: rebuild the Union City Gate Station in La Grande to increase the physical capacity (completed in December 2015), and the La Grande High Pressure
Reinforcement to minimize pressure drop across the distribution system (scheduled for completion in 2017).
The remaining city gate station projects in Table 7.2 have relatively small capacity
constraints, and thus will be periodically reevaluated to determine if upgrades need to be accelerated or deferred. Under currently planning considerations, these projects will be
tentatively scheduled for 2019 or later.

Table 7.2 City Gate Station Upgrades
CDA (East),
ID CDA East #221 Coeur d’Alene High
Pressure Reinforcement $10M 2016-
17
La Grande, OR La Grande #815 Union HP Connector $3M 2017
Athol, ID Athol #219 TBD - 2019+
Bonners Ferry, ID Bonners Ferry #208 TBD - 2019+
Colton, WA Colton #316 TBD - 2019+
Genesee, ID Genesee #320 TBD - 2019+
Klamath
Falls, OR Klamath Falls #2703 TBD - 2019+
Mica, WA Mica #15 TBD - 2019+
Pullman, WA Pullman #350 TBD - 2019+
Sprague, WA Sprague #117 TBD - 2019+
Sutherlin, OR Sutherlin #2626 TBD - 2019+
*Details of project described in IRP as of August 2016

CONCLUSION
Avista’s goal is to maintain its natural gas distribution systems reliably and cost effectively
to deliver natural gas to every customer. This goal relies on modeling to increase the
capacity and reliability of the distribution system by identifying specific areas that may
require changes. The ability to meet the goal of reliable and cost effective natural gas
delivery is enhanced through localized distribution planning, which enables coordinated
targeting of distribution projects responsive to customer growth patterns.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 140 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 141 of 162
8: Action Plan

The purpose of an action plan is to position Avista to provide the best cost/risk resource
portfolio and to support and improve IRP planning. The Action Plan identifies needed
supply and demand side resources and highlights key analytical needs in the near term.
It also highlights essential ongoing planning initiatives and natural gas industry trends
Avista will monitor as a part of its planning processes.

2015-2016 Action Plan Review

Action Item
Avista will continue to optimize underutilized resources to recover value for customers
and reduce their costs until resources are required to meet changing demand needs.

Results Avista optimizes underutilized resources to recover value for customers and reduce their
costs until these resources are required to meet demand needs. A new storage
optimization program has been developed to help recover costs at our Jackson Prairie
facility. This program uses the storage field’s intrinsic value to sell natural gas when prices
are higher, to sell on a cash to forward basis, or other market opportunities, all while
maintaining deliverability on a peak day. Avista sells into the daily market for transport
optimization. An example of this optimization is based on the cost difference, or spread,
between the AECO and Malin basins and is almost always economic. Avista will sell into
this market based on the remaining unused transportation in our portfolio.

Action Item
Avista will comply with Commission findings to try to increase the cost effectiveness of
DSM measures by reducing administration and audit costs, analyzing non-natural gas
benefits and increasing measure lives. Avista will monitor natural gas prices as a signpost
for increasing avoided costs. If avoided costs increase, Avista will reevaluate DSM
programs for cost effectiveness and submit to resume natural gas DSM programs.

Results
Avista continues to build on its history of collaboration with all stakeholders in delivering
meaningful cost-effective conservation measures as a way to reduce their energy bills
and promote a cleaner environment. The company considered several approaches to
improve the amount of cost-effective natural gas DSM measures offered to our customers
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 142 of 162
since the 2014 IRP. During 2015, the Avista DSM group began to look at the current
composition and components of natural gas avoided costs and compare them with other
regional and national utilities. The research and proposed additions to Avista’s avoided
cost were presented to Avista’s DSM Advisory Group for feedback on August 19 and 20,
2015 to ensure these were appropriate changes and to seek advice about other avoided
cost component analyses the company should perform. The company also changed how
non-incentive utility costs (NIUC) were being distributed to the overall DSM portfolio from
the ratio of BTUs to ratio of benefits. This helped balance the cost effectiveness between
electric and natural gas measures and programs. After the reevaluation of Avista’s
avoided cost methodology, change in distribution of NIUC, and with an Idaho Commission
ruling that allows companies to emphasize the UTC when seeking prudence for their DSM
programs, Avista filed for and was approved to reinstate its Idaho Natural Gas DSM
programs as of January 1, 2016.

Action Item
Complete the gate station analysis to assess resource deficiencies masked by
aggregated IRP analysis. Any identified deficiencies and potential solutions will be
discussed with Commission Staff. Avista will continue to coordinate analytic efforts
between Gas Supply, Gas Engineering and the intrastate pipelines to perform gate station
analysis and develop least cost solutions should deficiencies exist.

Results
Avista has completed the gate station analysis and communicated the results to TAC
members during the third TAC meeting in the Distribution section. The data set was a
static set from 2014 so some of the gate station shortages have already received some
mitigation fixes. The area of La Grande is one example, where the Ladd Canyon gate
station was completed in December 2015 to help offset the flow at the La Grande city
gate. Also, a high pressure reinforcement project will begin in 2017 to help ensure a
physical limitation at this city gate is not exceeded. As for the remaining city gate
deficiencies, the potential fixes will be reviewed between Avista’s Gas Supply and
Distribution Engineering departments. This review and potential fixes will be addressed
in regular meetings between Avista and all three commissions as addressed in the
ongoing activities section below.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 143 of 162
Action Item
As part of its next IRP process, Avista must convene workshops with Staff and
stakeholders to explore how best to model major resource acquisitions and major capital
investments.

Results
Avista reviewed current resources in the second Technical Advisory Committee (TAC)
meeting and distribution projects in the third TAC meeting. A shortage in the High Growth
and Low Price case was our only case where any demand was unserved. Avista
presented to the TAC members a list of resources modeled as well as the chosen
resources to best solve the unserved demand. In future IRP processes, Avista will review
its assumptions with the TAC members via open dialogue as well as how best to model
these potential resources.

Action Item
For the next IRP, Avista must work with Staff and stakeholders to resolve forecasting
methodology concerns, and seek to identify the most reliable methodology so that future
resource needs may be clearly identified.

Results
Avista provided a Sendout overview following the third TAC meeting on March 30, 2016
to interested TAC members. This overview helped provide a level set on the types of
inputs within the model and a general understanding on how the model works. Avista
also worked directly with Oregon Staff to produce a more accurate forecasting
methodology diagram. This diagram can be viewed in figure 2.5. In addition to this review,
Avista added analysis around modeling alternative resources to test the least cost and
least risk portion of the Expected Case Portfolio. As described in Chapter 6, Avista
modeled three alternative resources and compared PVRR to determine whether a new
resource stack may be appropriate. These actions have resolved forecasting
methodology concerns.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 144 of 162
Action Item
In its next IRP, Avista must include a clear presentation of how Avista decides which
distribution system projects to include in the IRP, and a clear description of the included
projects, along with a justification for recommending or proceeding with the projects.

Results
Avista has provided a description of distribution system projects and the most current
knowledge and analysis supporting the timeframe for completing each project. Also
included is an example to help describe the qualitative versus quantitative analysis for
each distribution project and the new decision process flow (Figure 7.1 – Distribution
Scenario Process).

Action Item
As part of its next IRP process, Avista must convene discussions with Staff and
stakeholders to discuss potential impacts associated with: (1) new regulations to reduce
methane emissions; and (2) potential increases in natural gas prices stemming from
increased demand for natural gas for generation under Section 111 (d) of the Clean Air
Act.

Results
During the second TAC meeting the company reviewed regulations affecting the natural
gas industry at a state or federal level along with an overview of each rule. Avista
presented its carbon pricing and methodology in the third TAC meeting. Avista will
continue to monitor impacts associated with potential methane emissions and the effects
these regulations may have on Avista’s jurisdictions.

Action Item
In Order No. 13-159, the Oregon Commission documented several demand side actions. These actions included filing specific DSM targets with achievable savings and
costs by measure and program for the next two to four years, noting any exceptions by measure, and participate in NEEA’s natural gas market transformation program and
include the achievable savings in the 2016 IRP.
Results
Order No. 13-159 directed Avista to continue DSM programs in Oregon and to achieve
at least 225,000 therms in 2013 and 250,000 therms in 2014. In addition, the company
needed to provide the following results by April 30, 2015:
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 145 of 162
 DSM program savings and cost effectiveness;
 Actions to reduce delivery costs, including administration and audit;
 Activities taken to increase the amount of cost effective efficiency measures;
 Analysis of non-natural gas benefits of existing and proposed DSM measures;
and
 Analysis of measure lives for all DSM measures.
In addition Avista was directed to do the following:
 Develop a potential mechanism for funding a low-income energy efficiency
program and report the mechanism to Staff.
 Determine the possibility of a regional elasticity study through the NWGA or the
AGA.
 Evaluate potential demand from NGV/CNG vehicles and other new uses of
natural gas.

As described in the 2014 IRP and 2014 IRP update, Avista met the actions in the
following manner:
 Avista continued the DSM program in Oregon in 2013 and 2014 and achieved 217,177 therms in 2013 and 192,955 therms in 2014.
 The Commission approved the Company’s DSM targets and exceptions on September 22, 2015 in Commission Order No. 15-288 in LC 61. Subsequently,
Avista agreed to transition its DSM program to the Energy Trust of Oregon (ETO). In a settlement approved on February 29, 2016 via Order No. 16-076.
Avista will transition to the ETO in 2016 with a final transition on January 1, 2017.
 Avista took steps to increase the cost effectiveness of the DSM program. Specifically, measure lives were extended, certain tariff changes were
implemented to reduce administration costs, audit costs were separated from other program costs, a new software program is being implemented for
calculating savings, and a separate low-income energy efficiency program was created.
 Avista worked with Staff and other stakeholders to develop a low-income energy
efficiency program and submitted a report to Staff outlining a proposed
mechanism on October 30, 2013. The company filed tariffs to implement the
Avista Oregon Low-Income Energy Efficiency (AOLIEE) Program on January 8,
2014 and the tariffs were approved and the AOLIEE Program started on March 1,
2014.
 Price elasticity predicts that energy consumers reduce consumption as prices
rise, but the amount of a response is debatable. Avista has reviewed research on price elasticity for natural gas. The analysis shows a wide range of results from
statistically significant to statistically insignificant and even positive in some cases. Avista contacted the AGA and they are willing to facilitate a process if a
regional price elasticity study moves forward. Avista is assessing the costs and
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 146 of 162
benefits of such an undertaking. A regional natural gas price elasticity study will
commence if enough interest develops in the project.
 In our assessment of potential demand impact due to NGV/CNG vehicles,
modeling results show a direct sensitivity to NGV/CNG vehicles. This results in
an increase of 9 MDTh on a February 20th peak day compared to the reference
case. In the Exported LNG case, price elasticity sensitivity shows a decrease in
usage in direct response to higher pricing. The analysis timeframe is over the
IRP's 20-year horizon with no shortages on a peak day in either case.
 Avista is participating in NEEA’s natural gas market transformation initiative. Details about the status of the initiative are in Chapter 3 – Demand Side
Resources.

2017-2018 Action Plan
Avista’s 2017-2018 Action Plan outlines activities for study, development and preparation
for the 2018 IRP.

New Activities for the 2018 IRP

 The price of natural gas has dropped significantly since the 2014 IRP. This is
primarily due to the amount of economically extractable natural gas in shale
formations, more efficient drilling techniques, and warmer than normal weather.
Wells have been drilled, but left uncompleted due to the poor market economics.
This is depressing natural gas prices and forcing many oil and natural gas
companies into bankruptcy. Due to historically low prices Avista will research
market opportunities including procuring a derivative based contract, 10-year
forward strip, and natural gas reserves.
 Avista’s 2018 IRP will contain a dynamic DSM program structure in its analytics.
In prior IRP’s, it was a deterministic method based on Expected Case
assumptions. In the 2018 IRP, each portfolio will have the ability to select
conservation to meet unserved customer demand. Avista will explore methods to
enable a dynamic analytical process for the evaluation of conservation potential
within individual portfolios.
 Monitor actual demand for accelerated growth to address resource deficiencies
arising from exposure to “flat demand” risk. This will include providing Commission
Staff with IRP demand forecast-to-actual variance analysis on customer growth
and use-per-customer at least bi-annually.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 147 of 162
Ongoing Activities

 Continue to monitor supply resource trends including the availability and price
of natural gas to the region, LNG exports, methanol plants, supply and market
dynamics and pipeline and storage infrastructure availability.
 Monitor availability of resource options and assess new resource lead-time
requirements relative to resource need to preserve flexibility.
 Meet regularly with Commission Staff to provide information on market
activities and significant changes in assumptions and/or status of Avista
activities related to the IRP or natural gas procurement practices.
 Appropriate management of existing resources including optimizing
underutilized resources to help reduce costs to customers.

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 148 of 162

Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 149 of 162
9: Glossary of Terms and Acronyms

Achievable Potential
Represents a realistic assessment of expected energy savings, recognizing and
accounting for economic and other constraints that preclude full installation of every
identified conservation measure.
AGA
American Gas Association
Annual Measures
Conservation measures that achieve generally uniform year-round energy savings
independent of weather temperature changes. Annual measures are also often called
base load measures.
Average Case
Represents Avista’s demand forecast for normal planning purposes. This case uses a 20
year rolling average NOAA weather for the five major areas (Spokane, WA., Medford, OR. Klamath Falls, OR, Roseburg, OR. La Grande, OR.).
Avista
The regulated Operating Division of Avista Corp.; separated into north (Washington and
Idaho) and south (Oregon) regions. Avista Utilities generates, transmits and distributes
electricity, in addition to the transmission and distribution of natural gas.
Backhaul
A transaction where gas is transported the opposite direction of normal flow on a
unidirectional pipeline.
Base Load
As applied to natural gas, a given demand for natural gas that remains fairly constant
over a period of time, usually not temperature sensitive.
Base Load Measures
Conservation measures that achieve generally uniform year-round energy savings
independent of weather temperature changes. Base load measures are also often called
annual measures.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 150 of 162
Basis Differential
The difference in price between any two natural gas pricing points or time periods. One
of the more common references to basis differential is the pricing difference between
Henry Hub and any other pricing point in the continent.
British Thermal Unit (BTU)
The amount of heat required to raise the temperature of one pound of pure water one
degree Fahrenheit under stated conditions of pressure and temperature; a therm (see
below) of natural gas has an energy value of 100,000 BTUs and is approximately
equivalent to 100 cubic feet of natural gas.
Capacity
The sum amount of natural gas transportation contracts or storage available in Avista’s
current portfolio.
CD
Contract Demand
C&I
Commercial and Industrial
City Gate (also known as gate station or pipeline delivery point)
The point at which natural gas deliveries transfer from the interstate pipelines to Avista’s
distribution system.
CNG
Compressed Natural Gas
Compression
Increasing the pressure of natural gas in a pipeline by means of a mechanically-driven
compressor station to increase flow capacity.
Conservation Measures
Installations of appliances, products or facility upgrades that result in energy savings.
Contract Demand (CD)
The maximum daily, monthly, seasonal or annual quantities of natural gas, which the
supplier agrees to furnish, or the pipeline agrees to transport, and for which the buyer or
shipper agrees to pay a demand charge.
Core Load
Firm delivery requirements of Avista, which are comprised of residential, commercial and
firm industrial customers.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 151 of 162
Cost Effectiveness
The determination of whether the present value of the therm savings for any given
conservation measure is greater than the cost to achieve the savings.
CPA
Conservation Potential Assessment
CPI
Consumer Price Index, as calculated and published by the U.S. Department of Labor,
Bureau of Labor Statistics.
Cubic Foot (cf)
A measure of natural gas required to fill a volume of one cubic foot under stated conditions
of temperature, pressure and water vapor; one cubic foot of natural gas has the energy
value of approximately 1,000 BTUs and 100 cubic feet of natural gas equates to one
therm (see below).
Curtailment
A restriction or interruption of natural gas supplies or deliveries; may be caused by
production shortages, pipeline capacity or operational constraints or a combination of
operational factors.
Dekatherm (Dth)
Unit of measurement for natural gas; a dekatherm is 10 therms, which is one thousand
cubic feet (volume) or one million BTUs (energy).
Demand-Side Management (DSM)
The activity pursued by an energy utility to influence its customers to reduce their energy
consumption or change their patterns of energy use away from peak consumption
periods.
Demand-Side Resources
Energy resources obtained through assisting customers to reduce their "demand" or use
of natural gas. Also represents the aggregate energy savings attained from installation of
conservation measures.
DSM
Demand-Side Management
Dth
Unit of measurement for natural gas; a dekatherm is 10 therms, which is one thousand
cubic feet (volume) or one million BTUs (energy).
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 152 of 162
EIA
Energy Information Administration
Expected Case
The most likely scenario for peak day planning purposes. This case uses a 20 year rolling average NOAA weather for the five major areas (Spokane, WA., Medford, OR. Klamath
Falls, OR, Roseburg, OR. La Grande, OR.). Combined with this 20 year rolling average
weather is the coldest day on record.
External Energy Efficiency Board
Also known as the "Triple-E" board, this non-binding external oversight group was
established in 1999 to provide Avista with input on DSM issues.
Externalities
Costs and benefits borne by a third party not reflected in the price paid for goods or
services.
Federal Energy Regulatory Commission (FERC)
The government agency charged with the regulation and oversight of interstate natural
gas pipelines, wholesale electric rates and hydroelectric licensing; the FERC regulates
the interstate pipelines with which Avista does business and determines rates charged in
interstate transactions.
FERC
Federal Energy Regulatory Commission
Firm Service
Service offered to customers under schedules or contracts that anticipate no
interruptions; the highest quality of service offered to customers.
Force Majeure
An unexpected event or occurrence not within the control of the parties to a contract,
which alters the application of the terms of a contract; sometimes referred to as "an act
of God;" examples include severe weather, war, strikes, pipeline failure and other similar
events.
Forward Haul
A transaction where gas is transported the normal direction of normal flow on a
unidirectional pipeline.
Forward Market
An over-the-counter marketplace that sets the price of a financial instrument or physical
asset for future delivery.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 153 of 162
Forward Price
The future price for a quantity of natural gas to be delivered at a specified time.
Gas Transmission Northwest (GTN)
A subsidiary of TransCanada Pipeline which owns and operates a natural gas pipeline
that runs from the Canada/USA border to the Oregon/California border. One of the six
natural gas pipelines Avista transacts with directly.
Gross Domestic Product (GDP)
The monetary value of all the finished goods and services produced within a country's
borders in a specific time period.
Geographic Information System (GIS)
A system of computer software, hardware and spatially referenced data that allows
information to be modeled and analyzed geographically.
GHG
Greenhouse Gas
Global Insight, Inc.
A national economic forecasting company.
GTN
Gas Transmission Northwest
Heating Degree Day (HDD)
A measure of the coldness of the weather experienced, based on the extent to which the
daily average temperature falls below 65 degrees Fahrenheit; a daily average
temperature represents the sum of the high and low readings divided by two.
Henry Hub
The physical location in Louisiana that is widely recognized as the most important natural
gas pricing point in the U.S., as well as the trading hub for the New York Mercantile
Exchange (NYMEX).
HP
High Pressure
Injection
The process of putting natural gas into a storage facility; also called liquefaction when the
storage facility is a liquefied natural gas plant.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 154 of 162
Integrity Management Plan
A federally regulated program that requires companies to evaluate the integrity of their
natural gas pipelines based on population density. The program requires companies to
identify high consequence areas, assess the risk of a pipeline failure in the identified areas
and provide appropriate mitigation measures when necessary.
Interruptible Service
A service of lower priority than firm service offered to customers under schedules or
contracts that anticipate and permit interruptions on short notice. The interruption
happens when the demand of all firm customers exceeds the capability of the system to
continue deliveries to all of those customers.
IPUC
Idaho Public Utilities Commission
IRP
Integrated Resource Plan; the document that explains Avista’s plans and preparations to
maintain sufficient resources to meet customers’ natural gas needs at a reasonable price.
Jackson Prairie
An underground natural gas storage project jointly owned by Avista Corp., Puget Sound
Energy and NWP. The project is a naturally occurring aquifer near Chehalis, Wash., which
is located about 1,800 feet beneath the surface and capped with a thick layer of dense
shale.
Liquefaction
Any process converting natural gas from the gaseous to the liquid state. For natural gas,
this process is accomplished through lowering the temperature of the natural gas (see
LNG).
Liquefied Natural Gas (LNG)
Natural gas liquefied by reducing its temperature to minus 260 degrees Fahrenheit at
atmospheric pressure.
Linear Programming
A mathematical method of solving problems by means of linear functions where the
multiple variables involved are subject to constraints; this method is utilized in the
SENDOUT® Gas Model.
Load Duration Curve
An array of daily send outs observed, sorted from highest send out day to lowest to
demonstrate peak requirements and the number of days it persists.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 155 of 162
Load Factor
The average load of a customer, a group of customers or an entire system, divided by the
maximum load; can be calculated over any time period.
Local Distribution Company (LDC)
A utility that purchases natural gas for resale to end-use customers and/or delivers
customer's natural gas or electricity to end users' facilities.
Looping
The construction of a second pipeline parallel to an existing pipeline over the whole or
any part of its length, thus increasing the capacity of that section of the system.
MCF
A unit of volume equal to a thousand cubic feet.
MDDO
Maximum Daily Delivery Obligation
MDQ
Maximum Daily Quantity
MMbtu
A unit of heat equal to one million British thermal units (BTUs) or 10 therms. Used
interchangeably with Dth.
National Energy Board
The Canadian equivalent to the Federal Energy Regulatory Commission (FERC).
National Oceanic Atmospheric Administration (NOAA)
Publishes the latest weather data; the 30-year weather study included in this IRP is based
on this information.
Natural Gas
A naturally occurring mixture of hydrocarbon and non-hydrocarbon gases found in porous
geologic formations beneath the earth's surface, often in association with petroleum; the
principal constituent is methane and it is lighter than air.
Natural Gas Reserves
Proven volumes of natural gas found in geological formations. These formations include
both drilled wells, wells with a remaining volume and have already been drilled, and
undrilled wells.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 156 of 162
New York Mercantile Exchange (NYMEX)
An organization that facilitates the trading of several commodities, including natural gas.
NGV
Natural Gas Vehicles
NOAA
National Oceanic and Atmospheric Administration
Nominal
Discounting method that includes inflation.
Nomination
The scheduling of daily natural gas requirements.
Non-Coincidental Peak Demand
The demand forecast for a 24-hour period for multiple regions that includes at least one
peak day and one non-peak day.
Non-Firm Open Market Supplies
Natural gas purchased via short-term purchase arrangements. May supplement firm
contracts during times of high demand or to displace other volumes when cost-effective.
Also referred to as spot market supplies.
Northwest Pipeline Corporation (NWP)
A principal interstate pipeline serving the Pacific Northwest and one of six natural gas
pipelines Avista transacts with directly. NWP is a subsidiary of The Williams Companies,
headquartered in Salt Lake City, Utah.
NOVA Gas Transmission (NOVA)
See TransCanada Alberta System
Northwest Power and Conservation Council (NPCC)
A regional energy planning and analysis organization headquartered in Portland, Ore.
NPCC
Northwest Power and Conservation Council
NWP
Williams-Northwest Pipeline
NYMEX
New York Mercantile Exchange
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 157 of 162
OPUC
Oregon Public Utility Commission
Peak Day
The greatest total natural gas demand forecasted in a 24-hour period used as a basis for
planning peak capacity requirements.
Peak Day Curtailment
Curtailment imposed on a day-to-day basis during periods of extremely cold weather
when demands for natural gas exceed the maximum daily delivery capability of a pipeline
system.
Peaking Capacity
The capability of facilities or equipment normally used to supply incremental natural gas
under extreme demand conditions (i.e. peaks); generally available for a limited number
of days at this maximum rate.
Peaking Factor
A ratio of the peak hourly flow and the total daily flow at the city-gate stations used to
convert daily loads to hourly loads.
Prescriptive Measures
Avista's DSM tariffs require the application of a formula to determine customer incentives
for natural gas-efficiency projects. For commonly encountered efficiency applications that
are relatively uniform in their characteristics, the utility has the option to define a
standardized incentive based upon the typical application of the efficiency measure. This
standardized incentive takes the place of a customized calculation for each individual
customer. This streamlining reduces both the utility and customer administrative costs of
program participation and enhances the marketability of the program.
Psig
Pounds per square inch gauge  a measure of the pressure at which natural gas is
delivered.
PVRR
Present Value Revenue Requirement
Rate Base
The investment value established by a regulatory authority upon which a utility is
permitted to earn a specified rate of return; generally this represents the amount of
property used and useful in service to the public.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 158 of 162
Real
Discounting method that excludes inflation.
Resource Stack
Sources of natural gas infrastructure or supply available to serve Avista’s customers.
Seasonal Capacity
Natural gas transportation capacity designed to service in the winter months.
Sendout
The amount of natural gas consumed on any given day.
SENDOUT®
Natural gas planning system from Ventyx; a linear programming model used to solve gas
supply and transportation optimization questions.
Service Area
Territory in which a utility system is required or has the right to provide natural gas service
to ultimate customers.
Shoulder Months
Months leading into or out of the winter heating season.
Spot Market Gas
Natural gas purchased under short-term agreements as available on the open market;
prices are set by market pressure of supply and demand.
Storage
The utilization of facilities for storing natural gas which has been transferred from its
original location for the purposes of serving peak loads, load balancing and the
optimization of basis differentials; the facilities are usually natural geological reservoirs
such as depleted oil or natural gas fields or water-bearing sands sealed on the top by an
impermeable cap rock; the facilities may be man-made or natural caverns. LNG storage
facilities generally utilize above ground insulated tanks.
Summer
Months including April, May, June, July, August, September and October where heating
demand is typically much lower than Winter.
TAC
Technical Advisory Committee
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 159 of 162
Tariff
A published volume of regulated rate schedules, plus general terms and conditions under
which a product or service will be supplied.
TF-I
NWP's rate schedule under which Avista moves natural gas supplies on a firm basis.
TF-2
NWP's rate schedule under which Avista moves natural gas supplies out of storage
projects on a firm basis.
Technical Advisory Committee (TAC)
Industry, customer and regulatory representatives that advise Avista during the IRP
planning process.
Technical Potential
An estimate of all energy savings that could theoretically be accomplished if every
customer who could potentially install a conservation measure did so without
consideration of market barriers such as cost and customer awareness.
Therm
A unit of heating value used with natural gas that is equivalent to 100,000 British thermal
units (BTU); also approximately equivalent to 100 cubic feet of natural gas.
Town Code
A town code is an unincorporated area within a county and a municipality within a county
served by Avista natural gas retail services.
TransCanada Alberta System
Previously known as NOVA Gas Transmission; a natural gas gathering and transmission
corporation in Alberta that delivers natural gas into the TransCanada BC System pipeline
at the Alberta/British Columbia border; one of six natural gas pipelines Avista transacts
with directly.
TransCanada BC System
Previously known as Alberta Natural Gas; a natural gas transmission corporation of
British Columbia that delivers natural gas between the TransCanada-Alberta System and
GTN pipelines that runs from the Alberta/British Columbia border to the United States
border; one of six natural gas pipelines Avista transacts with directly.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 160 of 162
Transportation Gas
Natural gas purchased either directly from the producer or through a broker and is used
for either system supply or for specific end-use customers, depending on the
transportation arrangements; NWP and GTN transportation may be firm or interruptible.
TRC
Total Resource Cost
Triple E
External Energy Efficiency Board
Tuscarora Gas Transmission Company
Tuscarora is a subsidiary of Sierra Pacific Resources and TransCanada; this natural gas
pipeline runs from the Oregon/California border to Reno, Nev.; one of the six natural gas
pipelines Avista transacts with directly.
Vaporization
Any process in which natural gas is converted from the liquid to the gaseous state.
WCSB
Western Canadian Sedimentary Basin
Weighted Average Cost of Gas (WACOG)
The price paid for a volume of natural gas and associated transportation based on the
prices of individual volumes of natural gas that make up the total quantity supplied over
an established time period.
Weather Normalization
The estimation of the average annual temperature in a typical or "normal" year based on
examination of historical weather data; the normal year temperature is used to forecast
utility sales revenue under a procedure called sales normalization.
Weather Sensitive Measures
Conservation measures whose energy savings are influenced by weather temperature
changes. Weather sensitive measures are also often referred to as winter measures.
Winter
Months including November, December, January, February and March where heating
demand is at its highest point.
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 161 of 162
Winter Measures
Conservation measures whose energy savings are influenced by weather temperature
changes. Winter measures are also often referred to as weather sensitive measures.
Withdrawal
The process of removing natural gas from a storage facility, making it available for delivery
into the connected pipelines; vaporization is necessary to make withdrawals from an LNG
plant.
WUTC
Washington Utilities and Transportation Commission
Exhibit No. 7
Case No. AVU-G-17-01 J. Morehouse, Avista
Schedule 1, Page 162 of 162