20201125Exhibit 1 - Avista 2018-2019 Electric Impact Evaluations.pdf

Exhibit No. 1: Avista 2018 and 2019 Electric Impact Evaluations
Exhibit No. 1:
Avista 2018 and 2019 Electric Impact Evaluations

PY 2018 Idaho Electric
Impact Evaluation
Report
June 1, 2019
Prepared for:
Avista Corporation
1411 East Mission Avenue
Spokane, WA 99202

Prepared by:
Jeff Cropp
Mitt Jones
Christie Amero
Rachel Fernandez
Jon Lee

i
Table of Contents
Portfolio Executive Summary ................................................................................................................ 1
Evaluation Methodology and Activities ................................................................................................. 1
Summary of Impact Evaluation Results ................................................................................................. 1
Conclusions and Recommendations ...................................................................................................... 2
Nonresidential Impact Evaluation ......................................................................................................... 4
Program Summary ................................................................................................................................. 4
Program Participation Summary ............................................................................................................ 4
Evaluation Goals and Objectives ............................................................................................................ 6
Nonresidential Impact Evaluation Methodology ................................................................................... 6
Nonresidential Impact Evaluation Results ............................................................................................. 8
Nonresidential Conclusions and Recommendations ........................................................................... 10
Residential Impact Evaluation ............................................................................................................. 12
Program Summary ............................................................................................................................... 12
Program Participation Summary .......................................................................................................... 12
Evaluation Goals and Objectives .......................................................................................................... 13
Residential Impact Evaluation Methodology ....................................................................................... 13
Residential Impact Evaluation Results ................................................................................................. 14
Residential Conclusions and Recommendations ................................................................................. 16
Low Income Impact Evaluation ........................................................................................................... 18
Program Summary ............................................................................................................................... 18
Program Participation Summary .......................................................................................................... 18
Evaluation Goals and Objectives .......................................................................................................... 18
Low Income Impact Evaluation Methodology ..................................................................................... 18
Low Income Impact Evaluation Results ............................................................................................... 19
Low Income Conclusions and Recommendations ................................................................................ 19
Fuel Efficiency Impact Evaluation ........................................................................................................ 20
Program Summary ............................................................................................................................... 20
Program Participation Summary .......................................................................................................... 20
Evaluation Goals and Objectives .......................................................................................................... 21
Fuel Efficiency Impact Evaluation Methodology .................................................................................. 22

ii
Fuel Efficiency Impact Evaluation Results ............................................................................................ 22
Fuel Efficiency Conclusions and Recommendations ............................................................................ 24

Tables
Table 1. Electric Program Evaluation Activities (PY 2018) ............................................................................ 1
Table 2. Reported and Verified Electric Savings (PY 2018) ........................................................................... 2
Table 3. Nonresidential Prescriptive Electric Savings (PY 2018) ................................................................... 5
Table 4. Nonresidential Prescriptive Participation Goals by Equipment Rebated ....................................... 5
Table 5. Nonresidential Prescriptive Participation by Project (PY 2018) ...................................................... 6
Table 6. Nonresidential Site Specific Electric Savings (PY 2018) .................................................................. 6
Table 7. Idaho Nonresidential Prescriptive Electric Evaluation Sample ....................................................... 8
Table 8. Idaho Nonresidential Site Specific Electric Evaluation Sample ....................................................... 8
Table 9. Nonresidential Prescriptive Electric Impact Findings ..................................................................... 9
Table 10. Nonresidential Prescriptive Evaluation Summary of Discrepancies ............................................. 9
Table 11. Nonresidential Site Specific Electric Impact Findings (PY 2018) ................................................. 10
Table 12. Nonresidential Site Specific Evaluation Summary of Discrepancies ........................................... 10
Table 13. Residential Prescriptive Reported Electric Savings (PY 2018) ..................................................... 12
Table 14. Residential Prescriptive Participation (PY 2018) ......................................................................... 13
Table 15. Multifamily Direct Install Reported Electric Savings ................................................................... 13
Table 16. Residential Prescriptive Electric Impact Document Review ....................................................... 14
Table 17. Residential Prescriptive Database Review Electric Impact Findings ........................................... 15
Table 18. Residential Prescriptive Electric Impact Document Review Realization Rates ........................... 15
Table 19. Residential Prescriptive Electric Impact Findings ....................................................................... 16
Table 20. Low Income Reported Savings (PY 2018) .................................................................................... 18
Table 21. Low Income Participation (PY 2018) ........................................................................................... 18
Table 22. Low Income Electric Impact Findings .......................................................................................... 19
Table 23. Nonresidential Site Specific Fuel Efficiency Electric Savings (PY 2018) ....................................... 20
Table 24. Residential Prescriptive Fuel Efficiency Reported Electric Savings (PY 2018) ............................. 21
Table 25. Residential Prescriptive Fuel Efficiency Reported Participation (PY 2018) ................................. 21
Table 26. Low Income Fuel Efficiency Reported Electric Savings (PY 2018) ............................................... 21
Table 27. Low Income Fuel Efficiency Participation (PY 2018) ................................................................... 21
Table 28. Nonresidential Fuel Efficiency Electric Impact Findings ............................................................. 23
Table 29. Residential Prescriptive Fuel Efficiency Electric Impact Findings ............................................... 23

iii
Table 30. Residential Prescriptive Fuel Efficiency Electric Impact Document Review Realization Rate ..... 23
Table 31. Low Income Fuel Efficiency Program Electric Impact Findings ................................................... 23

Figure
Figure 1. Residential Impact Process .......................................................................................................... 14

1
Portfolio Executive Summary
For several decades, Avista Corporation has been administering demand-side management programs to
reduce electricity and natural gas energy use for its portfolio of customers. Most of these programs have
been implemented in-house, but a few have external implementers. Avista contracted with Cadmus to
complete process and impact evaluations of its PY 2018 and PY 2019 electric demand-side management
programs in Idaho. This report presents our electric impact evaluation findings for PY 2018. Cadmus did
not apply net-to-gross adjustments to savings values, except in cases where deemed energy savings
values already incorporate net-to-gross as a function of the market baseline.
Evaluation Methodology and Activities
Cadmus conducted the Idaho portfolio evaluation using a variety of methods and activities, shown in
Table 1.
Table 1. Electric Program Evaluation Activities (PY 2018)
Sector Program Document/Database
Review
Verification/Metering
Site Visits
Nonresidential Prescriptive (Multiple) ü ü
Site Specific ü ü
Residential
Simple Steps, Smart Savings™ ü --
HVAC ü --
Shell ü --
ENERGY STAR® Homes ü --
Multifamily Direct Install ü --
Low Income Low Income ü --
Fuel Efficiency
Site Specific (Nonresidential) ü --
Prescriptive (Residential) ü --
Low Income ü --

Summary of Impact Evaluation Results
Overall, the Idaho electric portfolio achieved a 98% realization rate and acquired 29,805,007 kWh in
annual verified savings (Table 2). Cadmus collected the Avista-reported savings through database
extracts from Avista’s Customer Care and Billing (residential) and InforCRM (nonresidential) databases
and from data provided by third-party implementers to determine the verified savings that represent
our findings. In the second year of the two-year evaluation cycle (PY 2019), Cadmus will conduct utility
billing regression analyses to evaluate the most accurate energy savings for most residential programs.

2
Table 2. Reported and Verified Electric Savings (PY 2018)
Sector Reported Savings (kWh) Verified Savings (kWh) Realization Rate
Nonresidential 22,832,307 22,630,556 99%
Residential 5,400,520 5,108,673 95%
Low Income 228,498 252,699 111%
Fuel Efficiency 1,824,345 1,813,079 99%
Total 30,285,671 29,805,007 98%

Conclusions and Recommendations
During the PY 2018 evaluation, Cadmus identified several areas for improvement, outlined below by
sector.
Nonresidential Conclusions and Recommendations
While some individual project results varied, the overall nonresidential sector performed strongly in
PY 2018. Most of the projects Cadmus sampled for evaluation were well-documented and matched
what we found during site visit verifications.
Cadmus has two recommendations for improving the nonresidential sector energy savings:
• Ensure that the final reported savings calculations reflect the most up-to-date project details,
including post-installation verification photos, equipment submittals, and invoices. During two
project verifications, Cadmus found different installed equipment sizes, quantities, or
performance ratings than used in the reported savings calculations.
• Ensure that power metered data and pressure and airflow trend data collected for compressed
air projects are analyzed on a day-type approach (instead of taking the overall averages for the
metered period) to improve the accuracy of the energy-savings calculations. The day-type
analysis method is recommended by the Department of Energy’s Advanced Manufacturing
Office and Compressed Air Challenge® and is used in the AIRMaster+ free online software tool.1
This method provides a more granular estimation of the baseline and installed system flow
rates, performance, and energy use.
Residential Conclusions and Recommendations
During the evaluation, Avista confirmed that unit energy savings (UES) values used to calculate reported
savings for numerous residential measures had not been updated to match the 2018 Avista technical
reference manual (TRM) UES values. This was especially pronounced in the residential HVAC program,
where reported savings under-represented savings for heat pump measures. Under the direction of
Avista, Cadmus adjusted reported savings for these measures to match the 2018 TRM UES values.

1 Office of Energy Efficiency and Renewable Energy. April 10, 2014. “AIRMaster+ Motor Driven Systems.”
https://www.energy.gov/eere/amo/articles/airmaster

3
Based on reported savings, the ENERGY STAR Homes program achieved 253% of goal, but reported
participation and verified savings both showed that the program achieved approximately 72% of goal,
which indicates that reported savings values are well over the current TRM UES values.
Cadmus offers three recommendations regarding Avista’s residential electric programs:
• Ensure that reported savings for all prescriptive measures are calculated using the current TRM
or current 2017 Regional Technical Forum2 (RTF) UES values.
• Continue to encourage the adoption of efficient lighting through the Simple Steps, Smart Savings
program. The Northwest Energy Efficiency Alliance Residential Building Stock Assessment II
shows that roughly 40% of installed lamps in single family homes in Washington and Idaho are
incandescents or halogens.
• The Multifamily Direct Install (MFDI) program has proven to be an efficient, effective mechanism
for installing high-efficiency lighting and aerators in multifamily units. The Northwest Energy
Efficiency Alliance Residential Building Stock Assessment II “Multifamily Buildings Report”
estimated that 44% of lighting in multifamily units use incandescent or halogen technology.
Cadmus recommends to focus on replacing high-use, low-efficiency lamps where practical, to
maximize program cost-effectiveness while keeping savings high.
Fuel Efficiency Recommendations
Cadmus recommends that Avista update reported residential savings for fuel efficiency measures to use
current TRM UES values, particularly for measures where the differences are especially notable, such as
conversions to natural gas water heaters and natural gas wall furnaces.

2 Regional Technical Forum. 2017. Standard Protocols. https://rtf.nwcouncil.org/standard-protocols

4
Nonresidential Impact Evaluation
Through its nonresidential portfolio of programs, Avista promotes the purchase of high-efficiency
equipment for commercial and industrial utility customers. Avista provides rebates to partially offset the
difference in cost between high-efficiency equipment and standard equipment.
Program Summary
Avista completed and incented 685 nonresidential electric measures in Idaho in PY 2018 and reported
total electric energy savings of 22,832,307 kWh. Through the nonresidential sector, Avista offers
incentives for high-efficiency equipment and controls through three program paths: Prescriptive, Site
Specific, and Fuel Efficiency.
The Prescriptive program path is selected for smaller, straightforward equipment installations that
generally have similar operating characteristics (such as lighting, simple HVAC systems, food service
equipment, and variable frequency drives).
The Site Specific program path is reserved for more unique projects that require custom savings
calculations and technical assistance from Avista’s account executives (such as compressed air, process
equipment and controls, and comprehensive lighting retrofits).
Fuel Efficiency measures are part of the Site Specific program path, but involve a combination of electric
savings and natural gas penalties. These measures typically involve replacing electric space heating or
water heating systems with natural gas equipment. Please refer to the Fuel Efficiency Impact Evaluation
section for Cadmus’ evaluation methodology and findings for nonresidential Fuel Efficiency measures.
Program Participation Summary
This section summarizes nonresidential sector participation and progress toward PY 2018 goals through
the Prescriptive and Site Specific program paths.
Nonresidential Prescriptive Program Path
Table 3 shows electric energy savings goals assigned to Avista’s nonresidential Prescriptive program path
for PY 2018 as well as reported savings and a comparison between reported savings and goals.

5
Table 3. Nonresidential Prescriptive Electric Savings (PY 2018)
Program Name Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Interior Lighting 6,635,450 8,038,814 121%
Exterior Lighting 1,405,118 4,243,826 302%
Shell Measure 629 929 148%
Green Motors 45,180 42,466 94%
Motor Control (Variable Frequency
Drives [VFD]) 75,595 112,931 149%
Fleet Heat 8,000 0 0%
Food Service Equipment 94,730 8,527 9%
AirGuardian 18,000 0 0%
Energy Smart Grocera 724,348 3,402 0%
Total 9,007,050 12,450,896 138%
a The Energy Smart Grocer savings goal includes Site Specific Energy Smart Grocer measures. The Site Specific portion
constitutes approximately 10% of the overall goal.

Table 4 shows participation goals by rebated equipment quantity, as provided by Avista. The PY 2018
nonresidential tracking database extract listed individual projects, but did not include rebated
equipment quantity. For reference, Table 5 provides participation by unique application numbers.
Table 4. Nonresidential Prescriptive Participation Goals by Equipment Rebated
Program Type Planned Participation
Interior Lighting 26,904
Exterior Lighting 4,302
Shell Measure 325
Green Motors 18
Motor Control (VFD) 55
Fleet Heat 1
Food Service Equipment 18
AirGuardian 3
Energy Smart Grocera 2,097
a The Energy Smart Grocer goal includes Site Specific Energy Smart Grocer participants.

6
Table 5. Nonresidential Prescriptive Participation by Project (PY 2018)
Program Type Participation Reporteda
Interior Lighting 315
Exterior Lighting 273
Shell Measure 3
Green Motors 11
Motor Control (VFD) 3
Fleet Heat 0
Food Service Equipment 4
AirGuardian 0
Energy Smart Grocer 1
Total 610
a A participant is defined as a unique application number.

Nonresidential Site Specific Program Path
Table 6 shows electric savings goals assigned to the Site Specific program path in Avista’s nonresidential
sector for PY 2018, as well as reported savings. Note that the table does not include reported electric
savings for the Fuel Efficiency sector, such as those associated with the Multifamily Market
Transformation program.
Table 6. Nonresidential Site Specific Electric Savings (PY 2018)
Program Path Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Site Specific 4,000,000 10,381,411 260%

Evaluation Goals and Objectives
For the PY 2018 quarterly, semiannual, and annual reports, Cadmus conducted nonresidential impact
activities to determine verified savings for most programs.
Nonresidential Impact Evaluation Methodology
To evaluate impact evaluation savings for the PY 2018 nonresidential sector, Cadmus performed several
activities in two waves:
• Selected an evaluation sample and requested project documentation from Avista
• Performed project documentation review
• Prepared on-site M&V plans
• Performed site visits and collected on-site data (such as trend data, photos, and
operating schedules)
• Used site visit findings to calculate verified savings by measure
• Applied realization rates to total reported savings population to determine overall verified savings

7
The program context, along with Cadmus’ sample design, document review, and on-site verification
activities, is described in more detail below.
Program Context
As the first step of our evaluation activities, we gained an understanding of the programs and measures
being evaluated. Specifically, Cadmus explored documents and data records:
• Avista’s annual business plans, which detail processes and energy savings justifications
• Project documents from external sources (customers, program consultants, and implementation
contractors)
Based on the initial review, Cadmus outlined the distribution of program contributions to the overall
portfolio of programs. In addition, the review allowed us to understand the sources for UES for each
measure offered through the programs, along with the sources for energy-savings algorithms and the
internal quality assurance and quality control processes for large nonresidential sector projects.
Following this review, Cadmus designed the sample strategy for the impact evaluation activities, as
discussed in the following section.
Sample Design
We based the first evaluation sample on program data from January 2018 to April 2018, and we based
the second evaluation sample on program data from May 2018 through December 2018. As a guideline,
Cadmus used the proposed, overall PY 2018 and PY 2019 nonresidential sample sizes by subprogram in
the M&V plan, seeking to complete approximately one-quarter of the sample during the first wave and
another one-quarter during the second wave.
For each activity wave, we broke down submitted program applications by path and measure (such as
Site Specific shell measure, Prescriptive lighting, or Prescriptive motor controls), allowing us to select the
highest-savings applications in each category with certainty. For applications with reported savings
greater than 1% of total savings by category, Cadmus assigned random numbers and sampled randomly.
We removed applications with less than 1% of total savings by category from the sample consideration,
except where another application at the same location or facility was previously selected (and where we
could assess both applications with one site visit, which is a cost-effective verification strategy even if
the second application represents minimal claimed savings).
Cadmus sampled randomly selected sites across both Washington and Idaho since Avista’s programs are
implemented similarly in both states. We pooled the results from the randomly selected sites to
calculate a realization rate by stratum that we applied to projects in both states. We applied verified
savings for sites selected with certainty only to the state in which they had been implemented.
Table 7 summarizes the Idaho nonresidential Prescriptive program path evaluation sample. Across both
states, Cadmus sampled 40 Prescriptive applications at 34 unique sites. Of the sampled applications, we
selected 21 for certainty review based on the scale of savings, measure type, or location, and we

8
selected the remaining 19 applications randomly. There was no participation in the AirGuardian and
Fleet Heat programs in Idaho in PY 2018.
Table 7. Idaho Nonresidential Prescriptive Electric Evaluation Sample
Program Type Applications Sampled Sampled Savings (kWh) Percentage of Reported Savings
Interior Lighting 6 2,311,797 29%
Exterior Lighting 2 110,360 3%
Shell Measure 1 198 21%
Green Motors 4 18,678 44%
Motor Control (VFD) 2 104,755 93%
Fleet Heat 0 0 N/A
Food Service Equipment 0 0 0%
AirGuardian 0 0 N/A
Energy Smart Grocer 1 3,402 100%
Nonresidential Prescriptive 16 2,549,190 20%

Table 8 summarizes the Idaho nonresidential Site Specific program path evaluation sample. Across both
states, Cadmus sampled 18 Site Specific applications at 15 unique sites. Of the sampled applications, we
selected 12 for certainty review based on the scale of savings, measure type, or location, and we
selected the remaining six applications randomly.
Table 8. Idaho Nonresidential Site Specific Electric Evaluation Sample
Program Path Applications Sampled Sampled Savings (kWh) Percentage of Reported Savings
Site Specific 7 7,648,853 74%

Document Review
We requested and reviewed project documentation for each sampled application and prepared M&V plans
to guide our site visits. Project documentation typically included incentive applications, calculation tools
(usually based on the 2017 RTF), invoices, equipment specification sheets, and post-inspection reports.
On-Site Verification
Cadmus performed site visits at 46 unique nonresidential locations to assess electric savings for
58 unique Prescriptive and Site Specific measures (not including Fuel Efficiency measures). Site visits
involved verifying the installed equipment type, make and model numbers, operating schedules, and
setpoints, as applicable. We did not consider it necessary to conduct power metering or light logging for
PY 2018 site visits. Cadmus collected two weeks of trend data for two of the Site Specific industrial
process measures at one industrial site. We used the project documentation review and on-site findings
to adjust the reported savings calculations where necessary.
Nonresidential Impact Evaluation Results
This section summarizes the nonresidential sector Prescriptive and Site Specific program paths’ electric
impact evaluation results for PY 2018.

9
Nonresidential Prescriptive Programs
Table 9 shows reported and verified electric energy savings for Avista’s Nonresidential sector
Prescriptive program path and the realization rates between verified and reported savings for PY 2018.
The overall nonresidential sector Prescriptive program path electric realization rate was 100%.
Table 9. Nonresidential Prescriptive Electric Impact Findings
Program Type Reported Savings (kWh) Verified Savings (kWh) Realization Rate
Interior Lighting 8,038,814 8,012,238 100%
Exterior Lighting 4,243,826 4,243,826 100%
Shell Measure 929 929 100%
Green Motors 42,466 42,870 101%
Motor Control (VFD) 112,931 113,171 100%
Fleet Heat 0 0 100%
Food Service Equipment 8,527 8,527 100%
AirGuardian 0 0 100%
Energy Smart Grocer 3,402 3,402 100%
Nonresidential Prescriptive 12,450,896 12,424,964 100%

Of the evaluated applications, Cadmus identified discrepancies for seven based on the site visit and
project documentation review. Table 10 summarizes the reasons for discrepancies between reported
and verified savings.
Table 10. Nonresidential Prescriptive Evaluation Summary of Discrepancies
Program Type Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
Interior Lighting
2 ê Cadmus reduced the lighting fixture in-service rate for two projects to
account for incented fixtures that were on the site but in storage.
1 ê Savings in project documentation were slightly lower than reported
savings in program database.
1 é
Cadmus accounted for additional savings from delamping for reported 4-
lamp LED fixtures that only had two lamps. The participant had removed
two lamps per fixture due to brightness.
Green Motor
Rewind 2 é Reported savings for two projects referenced the 2017 RTF. Cadmus
applied deemed motor savings from the 2018 TRM workbook.
Motor Control
(VFD) 1 ê
Cadmus reduced the reported quantity of 2.5 horsepower return air fans
with VFDs from three to one and added two 3 horsepower return air fans
with VFDs.

Nonresidential Site Specific Program
Table 11 shows reported and verified electric energy savings for Avista’s PY 2018 nonresidential sector
Site Specific program path, as well as a comparison between verified and reported savings for PY 2018.
The overall Site Specific program path electric realization rate was 98%. Note that the table does not
include reported and verified electric savings for measures in the Fuel Efficiency path.

10
Table 11. Nonresidential Site Specific Electric Impact Findings (PY 2018)
Program Path Reported Savings (kWh) Verified Savings (kWh) Realization Rate
Site Specific 10,381,411 10,205,592 98%

Of the evaluated applications, Cadmus identified discrepancies in five based on the site visit and project
documentation review. Table 12 summarizes the reasons for discrepancies between reported and
verified savings.
Table 12. Nonresidential Site Specific Evaluation Summary of Discrepancies
Project
Type
Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
Exterior
Lighting
1 ê The site installed a higher quantity of exterior LED fixtures. The reported savings
in database did not match the implementer’s submitted calculation workbook.
1 ê The site installed fewer LED pole lighting fixtures and more LED wall pack fixtures
than reported.
Industrial
Process 1 é
The site converted two pressure roll vacuum units from double zone to single
zone units, eliminating the need for one of the four 500-horsepower vacuum
pumps. Reported savings calculations assumed the pump motor to be 100%
efficient. Cadmus adjusted the savings calculations to incorporate losses for a
conservative, high-efficiency 500-horsepower motor (95.8%).
Compressed
Air 1 ê
The site replaced two fixed-speed air compressors with two 350-horsepower
variable speed rotary-screw air compressors. Project documentation included
post-installation power and airflow metered data. Cadmus updated the reported
savings calculations by breaking out pre- and post-period airflow and baseline
system performance on a weekday basis, rather than an overall metered period
basis. Although the difference in average overall airflow was minimal between the
verified and reported methodology, there were days (such as Wednesdays and
Fridays) that differed by 175 CFM to 230 CFM from reported. This difference had a
significant impact on the performance of the baseline air compressors.
Interior
Lighting 1 é
Cadmus included the calculated cooling load electric energy savings in the interim
verified savings. These savings were calculated in the project documentation but
not included in the reported savings.

Nonresidential Conclusions and Recommendations
The nonresidential sector achieved total verified electric energy savings of 22,631 MWh in PY 2018 with
a combined realization rate of 99%. The nonresidential sector also exceeded the combined Prescriptive
and Site Specific program paths’ electric savings goal of 13,007 MWh by 74%.
While some individual project results varied, the overall nonresidential sector performed strongly in
PY 2018. Most of the projects Cadmus sampled for evaluation were well-documented and matched
what we found during site visit verification.
Cadmus has two recommendations for improving the nonresidential sector energy savings:
• Ensure that the final reported savings calculations reflect the most up-to-date project details,
including post-installation verification photos, equipment submittals, and invoices. During two

11
project verifications, Cadmus found different installed equipment sizes, quantities, or
performance ratings than used in the reported savings calculations.
• Ensure that power metered data and pressure and airflow trend data collected for compressed
air projects are analyzed on a day-type approach (instead of taking the overall averages for the
metered period) to improve the accuracy of the energy-savings calculations. The day-type
analysis method is recommended by the U.S. Department of Energy’s Advanced Manufacturing
Office and Compressed Air Challenge® and is used in the AIRMaster+ free online software tool.
This method provides a more granular estimation of the baseline and installed system flow
rates, performance, and energy use.

12
Residential Impact Evaluation
Cadmus designed the residential sector impact evaluation to verify reported program participation and
energy savings. We used data collected and reported in the tracking database, online application forms,
Avista TRM and RTF savings review, and applicable updated deemed savings values.
Program Summary
Avista completed and incented 250,234 residential electric measures or units in Idaho in PY 2018 and
reported total electric energy savings of 5,400,520 kWh, not including participation and savings from
Fuel Efficiency measures, which are included below in the Fuel Efficiency Impact Evaluation section. The
residential programs comprise two primary paths—Prescriptive and MFDI. The Prescriptive path
includes Simple Steps, Smart Savings, which encourages consumers to purchase and install high-quality
LEDs, light fixtures, and energy-efficient showerheads; the Residential HVAC program, which incents
high-efficiency heating and cooling equipment; the Residential Shell program, which provides rebates to
encourage customers to install high-efficiency windows and storm windows; and the ENERGY STAR
Homes program, which offers 15% to 25% energy savings relative to state energy code. Through the
MFDI program, Avista provides free direct-install measures to multifamily residences (of five units or
more) and common areas.
Program Participation Summary
This section summarizes residential sector participation and progress toward PY 2018 goals for the
residential Prescriptive and residential MFDI programs.
Residential Prescriptive Programs
Table 13 shows savings goals assigned to Avista’s residential sector Prescriptive programs for PY 2018, as
well as reported savings and the goal portion achieved in PY 2018.
Table 13. Residential Prescriptive Reported Electric Savings (PY 2018)
Program Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Simple Steps, Smart Savings 1,326,717 3,693,056 278%
HVAC 730,543 765,230 105%
Shell 60,854 95,748 157%
ENERGY STAR Homes 46,144 116,567 253%
Residential Prescriptive Total 2,164,258 4,670,601 216%

Table 14 summarizes participation goals and reported participation in Avista’s residential sector
Prescriptive programs for PY 2018, along with the percentage of goal achieved.

13
Table 14. Residential Prescriptive Participation (PY 2018)
Program Participation Goals Participation Reported Percentage of Goal
Simple Steps, Smart Savingsa 88,465 240,437 272%
HVACb 518 462 89%
Shellc 5,016 7,979 159%
ENERGY STAR Homesb 14 26 186%
Residential Prescriptive Total 94,013 248,904 265%
a Participation is defined as the number of purchased units.
b Participation is defined as the number of rebates.
c Participation is defined as square feet of installed windows or storm windows.

Multifamily Direct Install Program
Table 15 shows reported savings and participation for the MFDI program in PY 2018. Avista launched
this program as a pilot in PY 2018 and did not set annual program goals, then transitioned this from a
pilot to an ongoing study in September 2018.
Table 15. Multifamily Direct Install Reported Electric Savings
Program Path Savings Reported (kWh) Participation Reported
Multifamily Direct Install 729,920 1,330

Evaluation Goals and Objectives
For the PY 2018 quarterly, semiannual, and annual reports, Cadmus verified savings for most programs
through a combination of database review and document review, which are described below.
Residential Impact Evaluation Methodology
To determine the residential sector verified savings for PY 2018, Cadmus employed two impact
evaluation methods for most residential programs:3
• Database review
• Document review
Similar to previous practice, Cadmus calculated adjusted savings based on results of the database review
and applied realization rates from our document reviews. Verified savings represented adjusted savings
multiplied by the document review realization rates, as shown in Figure 1.

3 With approval from Avista, Cadmus ceased performing a third impact activity—verification surveys—in the
third quarter of PY 2018 to eliminate redundancy between verification surveys and document review.

14
Figure 1. Residential Impact Process

Database Review
For the impact evaluation database review, Cadmus used UES values, as provided in the TRM, to
calculate savings for measures reported in the measure tracking database. This impact activity may help
identify incorrect UES values used to calculate reported savings. Savings calculated during the database
review are defined as adjusted savings.
Document Review
For the document review, Cadmus compared information from rebate forms and other supporting
documents to measure tracking data for a random sample of projects. This impact activity may identify
installed measures that did not meet eligibility requirements, quantities that did not match the measure
tracking database, and other discrepancies. Following our review of all projects, we calculated a
realization rate for the document review by dividing savings calculated for the sample (using the revised
information) by reported savings for the sample. We then multiplied this realization rate by adjusted
savings for the entire program to determine verified savings.
Cadmus conducted document reviews for the programs shown in Table 16, drawing roughly equal
samples from participants in each quarter.
Table 16. Residential Prescriptive Electric Impact Document Review
Program Completed for PY 2018
HVAC 34
Shell 34

Residential Impact Evaluation Results
The following sections summarize findings and provide verified savings for both of Cadmus’ impact
evaluation methodologies. The database review resulted in the largest number of adjustments to
reported savings.

15
Database Review
Table 17 shows database review findings, with adjusted savings being higher than reported savings for
some programs and lower for others. Adjusted savings differed from reported savings because reported
UES values differed from TRM values for several measures. Avista determined that the reported savings
for these measures used values from an older customer database that did not align with those in the
current TRM. For measures with reported savings based on measure-specific parameters, Cadmus could
not confirm the reported savings calculations, which depended on inputs that were not included in the
tracking data (such as air infiltration and duct sealing).
Table 17. Residential Prescriptive Database Review Electric Impact Findings
Program Reported Electric Savings (kWh) Adjusted Electric Savings (kWh) Percentage Change
Simple Steps, Smart Savings 3,693,056 3,454,438 -6%
HVAC 765,230 786,170 3%
Shell 95,748 100,535 5%
ENERGY STAR Homes 116,567 83,738 -28%
Residential Prescriptive Total 4,670,601 4,424,881 -5%

Document Review
Table 18 summarizes document review findings to date. The HVAC program had a 96% electric
realization rate and the Shell program had an 85% electric realization rate.
Table 18. Residential Prescriptive Electric Impact Document Review Realization Rates
Program Document Audit
Count
Sample Reported
Savings (kWh)
Sample Verified
Savings (kWh)
Document Audit
Realization Rate
HVAC 34 32,997 31,691 96%
Shell 34 49,224 41,915 85%

Cadmus identified several discrepancies during the document review through PY 2018:
• For two window measures, documentation showed a square footage for installed windows that
differed from that reported. In both cases, the documented square footage was lower than that
reported and resulted in lower verified savings based on the corrected area.
• For four window measures reported at sites with electric heating, project documents identified
heating fuels other than electricity. Cadmus added natural gas savings and removed electricity
savings at two sites identified as using natural gas heating. Documentation for the other two
sites identified the heating fuel as liquid propane for one site and wood pellets for the other, so
Cadmus removed electricity savings for these sites.
• One heat pump water heater measure had a tank capacity of 80 gallons per the documentation.
To qualify for the measure, however, the heat pump water heater had to have a tank size below
55 gallons, so Cadmus removed savings for this measure.

16
Table 19 shows verified savings, which apply the realization rates shown in Table 18 to the adjusted
savings calculated based on the database review. The verified savings represent Cadmus’ best estimate
of savings without conducting a billing analysis.
Table 19. Residential Prescriptive Electric Impact Findings
Program Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Verified Electric
Savings (kWh)a
Realization
Rates
Simple Steps, Smart Savings 3,693,056 3,454,352 3,454,352 94%
HVACb 765,230 786,170 755,054 99%
Shell 95,748 100,535 85,608 89%
ENERGY STAR Homes 116,567 83,738 83,738 72%
Residential Prescriptive Total 4,670,601 4,424,796 4,378,753 94%
a Verified savings represents adjusted savings only for Simple Steps, Smart Savings and ENERGY STAR Homes.
b Includes heat pump water heater installations not included in Table 17 of the PY 2018 Idaho Annual Conservation Report.
Table 19 of that report includes heat pump water heater savings.

Residential Conclusions and Recommendations
Verified electricity savings show a realization rate of 94% on realized savings of 4,378,753 kWh for the
residential Prescriptive programs, which is 202% of the savings goal for the year, due largely to program
participation that was 265% of goal. Reported savings for the MFDI program added 729,920 kWh, for
total acquired savings of 5,108,673 kWh.
Lighting measures account for a high percentage of residential sector program path savings: Simple
Steps, Smart Savings provided 68% of residential savings, mostly through lighting measures, and MFDI
provided 14% of savings, also mostly through lighting measures. The HVAC program accounted for 15%
of savings, with Shell and ENERGY STAR Homes accounting for a combined 3% of residential sector
savings.
During the evaluation, Avista confirmed that the UES values used to calculate reported savings for
numerous residential sector measures had not been updated to match 2018 TRM UES values. This was
especially pronounced in the Residential HVAC program, where reported savings under-represented
savings for heat pump measures. Under the direction of Avista, Cadmus adjusted reported savings for
these measures to match the 2018 TRM UES values.
Based on reported savings, the ENERGY STAR Homes program achieved 253% of goal, but reported
participation and verified savings both showed that the program achieved approximately 72% of goal,
which indicates that reported savings values are well over the 2018 TRM UES values.
Cadmus offers three recommendations regarding Avista’s residential sector electric programs:
• Ensure that reported savings for all Prescriptive measures are calculated using current TRM or
RTF UES values.
• Continue to encourage the adoption of efficient lighting through the Simple Steps, Smart Savings
program. The Northwest Energy Efficiency Alliance Residential Building Stock Assessment II

17
shows that roughly 40% of installed lamps in single family homes in Washington and Idaho are
based either on incandescent or halogen technology.
• The MFDI program has proven to be an efficient, effective mechanism for installing high-
efficiency lighting and aerators in multifamily units. The Northwest Energy Efficiency Alliance
Residential Building Stock Assessment II “Multifamily Buildings Report” estimated that 44% of
lighting in multifamily units use incandescent or halogen technology. Cadmus recommends to
focus on replacing high-use, low-efficiency lamps where practical, to maximize program cost-
effectiveness while keeping savings high.

18
Low Income Impact Evaluation
Cadmus designed the Low Income programs’ impact evaluation to verify reported program participation
and energy savings. We used data collected and reported in the tracking database and conducted a TRM
savings review.
Program Summary
Avista leverages the infrastructure of a single Community Action Partnership agency to deliver energy
effiicency programs for the company’s low-income residential customers in the Idaho service territory.
The program is designed to serve Avista residential customers in Idaho whose income falls between 175
percent and 250 percent of federal poverty level. For PY 2018, the program achieved 228,498 kWh of
reported electric savings in Idaho, not including savings for the Low Income Fuel Efficiency measures,
which are reported separately in the Fuel Efficiency Impact Evaluation section.
Program Participation Summary
Table 20 shows Avista savings goals for the Low Income sector for PY 2018, as well as reported savings
and goal portions achieved in PY 2018.
Table 20. Low Income Reported Savings (PY 2018)
Program Savings Goals (kWh) Reported Savings (kWh)a Percentage of Goal
Low Income 159,162 228,498 144%
a Reported savings do not include Low Income Fuel Efficiency savings, shown in the Fuel Efficiency Impact Evaluation section.

Table 21 summarizes participation goals for the Low Income programs, along with participation reported
and achieved in PY 2018.
Table 21. Low Income Participation (PY 2018)
Program Participation Goalsa Participation Reporteda Percentage of Goal
Low Income 16,419 63,436 386%
a Participation numbers do not include Low Income Fuel Efficiency participation, shown in the Fuel Efficiency Impact
Evaluation section, or recipients of LED bulbs at giveaway events. Participation is defined as the number of installed units or
square feet of installed insulation or windows.

Evaluation Goals and Objectives
For quarterly and semiannual reports in PY 2018 and PY 2019, Cadmus will determine verified savings
for the Low Income programs through a database review (described above in the Database Review
section). This approach will provide a strong estimate of achieved savings until Cadmus can perform
billing analysis at the end of the two-year evaluation cycle.
Low Income Impact Evaluation Methodology
Cadmus’ impact evaluation for the Low Income programs’ measures consisted of database review
(described above in the Database Review section). We used UES values provided in the TRM to calculate

19
savings for measures reported in the measure tracking database. Cadmus labeled savings calculated
during the database review as adjusted savings.
Low Income Impact Evaluation Results
Table 22 shows reported and adjusted electric savings for Low Income conservation measures. The table
does not include savings for Low Income programs Fuel Efficiency path measures (shown in the Low
Income Fuel Efficiency Impact Findings section below).
Table 22. Low Income Electric Impact Findings
Program Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Verified Electric
Savings (kWh) Realization Rate
Low Income 228,498 252,699 252,699 111%

Low Income Conclusions and Recommendations
With a realization rate of 111% for electricity savings, the Low Income programs achieved savings of
252,699 kWh in PY 2018, or about 159% of goal. Reported program participation reached 386% of the
expected value. Cadmus recommends that Avista adjust its participation goal to better align with PY
2018 findings, and also adjust savings per participant, as the participation goal was exceeded by a much
larger margin than the total savings goal.
Roughly one third of verified Low Income program savings resulted from LED bulbs given out at events.
Cadmus understands that Avista relies on a Community Action Partnership to deliver Low Income
savings in Idaho. Cadmus’ PY 2019 evaluation activities will include a process review of the Low Income
programs, which may help identify opportunities to improve program performance.

20
Fuel Efficiency Impact Evaluation
Cadmus designed the Fuel Efficiency sector impact evaluation to verify reported program participation
and energy savings. We used data collected and reported in the tracking database and details from
online application forms, as well as reviewed TRM and RTF savings and applicable updated deemed
savings values.
Program Summary
Fuel Efficiency measures replace electric space heating or water heating systems with equipment using
natural gas. These measures are offered within the Nonresidential Site Specific path, Residential
Prescriptive programs, and Low Income programs. Across these programs, the Fuel Efficiency measures
reported participation of 190 in PY 2018 and electric energy savings of 1,824,345 kWh.
Fuel Efficiency measures provide positive electricity savings and negative natural gas savings, reflecting
negative avoided costs. Cadmus incorporated these negative avoided costs in the electric cost-
effectiveness calculations. We report the negative natural gas savings in the PY 2018 Idaho Natural Gas
Impact Evaluation Report.
Program Participation Summary
This section summarizes Fuel Efficiency sector participation and progress toward PY 2018 goals for the
Nonresidential Site Specific path, Residential Prescriptive programs, and Low Income programs.
Nonresidential Site Specific Path
The Nonresidential sector Site Specific program path includes Fuel Efficiency measures that replace
electric space heating or water heating systems with natural gas equipment. Fuel Efficiency measures
provide positive electricity savings and negative natural gas savings, reflecting negative avoided costs.
Two types of measures are considered Fuel Efficiency in the PY 2018 Idaho Nonresidential sector
database:
• Site Specific HVAC combined
• Site Specific multifamily
Table 23 shows electric savings goals and reported electric savings for the Nonresidential sector Fuel
Efficiency measures. There were only five participants in PY 2018. Avista confirmed that it did not set
participation goals for Site Specific Fuel Efficiency measures outside the Multifamily Market
Transformation program.
Table 23. Nonresidential Site Specific Fuel Efficiency Electric Savings (PY 2018)
Fuel Efficiency Measure Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Nonresidential Site Specific N/A 65,061 N/A
Multifamily Market Transformation 299,574 207,408 69%

21
Residential Prescriptive Programs
Table 24 shows Avista PY 2018 savings goals for Residential Prescriptive Fuel Efficiency measures as well
as reported savings and percentage of goal through PY 2018.
Table 24. Residential Prescriptive Fuel Efficiency Reported Electric Savings (PY 2018)
Fuel Efficiency Measure Savings Goals (kWh) Reported Savings (kWh) Percentage of Goal
Residential Prescriptive Fuel Efficiency 2,727,600 1,449,994 53%

Table 25 shows the Avista PY 2018 participation goal and reported participation for Residential
Prescriptive Fuel Efficiency measures, as well as the participation percentage of goal through PY 2018.
Table 25. Residential Prescriptive Fuel Efficiency Reported Participation (PY 2018)
Fuel Efficiency Measure Participation Goalsa Participation Reporteda Percentage of Goal
Residential Prescriptive Fuel Efficiency 271 170 63%
a Participation is defined as the number of rebates.

Low Income Programs
Table 26 shows Avista PY 2018 savings goals for Low Income Fuel Efficiency measures, as well as
reported savings and percentage of goal through PY 2018.
Table 26. Low Income Fuel Efficiency Reported Electric Savings (PY 2018)
Fuel Efficiency Measure Savings Goals (kWh) Reported Savings (kWh) Percentage of Goal
Low Income Fuel Efficiency 344,850 101,882 30%

Table 27 summarizes participation goals for Low Income Fuel Efficiency measures, as well as
participation reported and achieved through PY 2018.
Table 27. Low Income Fuel Efficiency Participation (PY 2018)
Fuel Efficiency Measure Participation Goalsa Participation Reporteda Percentage of Goal
Low Income Fuel Efficiency 46 15 33%
a Participation is defined as the number of rebates.

Evaluation Goals and Objectives
For quarterly and semiannual reports in PY 2018 and PY 2019, Cadmus will determine verified savings
for Nonresidential Site Specific and Residential Prescriptive Fuel Efficiency measures through database
review (described above in the Database Review section) and document review (described above in the
Document Review section). For Low Income Fuel Efficiency measures, Cadmus will determine adjusted
savings through database review. These approaches will provide strong estimates of achieved savings
until Cadmus can perform billing analysis at the end of the two-year evaluation cycle.

22
Fuel Efficiency Impact Evaluation Methodology
The impact methodology for Fuel Efficiency measures is outlined below for the Nonresidential Site
Specific path, Residential Prescriptive programs, and Low Income programs.
Nonresidential Site Specific Fuel Efficiency Impact Methodology
Cadmus followed the same impact evaluation methodology for Fuel Efficiency measures as outlined in
the Nonresidential Impact Evaluation Methodology section. We selected six Multifamily Market
Transformation program projects for our evaluation of the Nonresidential sector Fuel Efficiency
measures, all of which were in Washington. Of the sampled applications, we selected five for certainty
review based on the scale of savings, measure type, or location, and selected the remaining application
randomly.
Cadmus performed site visits at five unique Nonresidential locations to assess electric savings for the six
unique Multifamily Market Transformation program measures. Site visits involved verifying installed
equipment type, make and model numbers, operating schedules, and set points, as applicable.
Residential Prescriptive Fuel Efficiency Impact Methodology
For our impact evaluation of Residential Prescriptive Fuel Efficiency measures, we followed the
methodology described in the For the PY 2018 quarterly, semiannual, and annual reports, Cadmus
verified savings for most programs through a combination of database review and document review,
which are described below.
Residential Impact Evaluation Methodology section and conducted database review and document
review. We completed document reviews for 34 Fuel Efficiency participants in PY 2018.
Low Income Fuel Efficiency Impact Methodology
For our impact evaluation of Low Income Fuel Efficiency measures, we focused on a database review
(described above in the Database Review section). We used unit savings values provided in the TRM to
calculate savings for measures reported in the measure tracking database. Savings calculated during the
database review are adjusted savings. For Low Income programs’ measures in general (including Low
Income Fuel Efficiency measures), these savings are also considered verified savings.
Fuel Efficiency Impact Evaluation Results
The following sections summarize findings for the Nonresidential Site Specific path, Residential
Prescriptive programs, and Low Income programs Fuel Efficiency measures. All Fuel Efficiency measures
provide positive electricity savings and negative natural gas savings because these measures replace
electric space heating or water heating systems with equipment that uses natural gas. Negative savings,
reflecting negative avoided costs, are incorporated in the electric cost-effectiveness calculations. We
also report these negative savings in the PY 2018 Idaho Natural Gas Impact Evaluation Report.

23
Nonresidential Fuel Efficiency Impact Findings
Table 28 shows reported and verified electric energy savings for Avista’s Nonresidential sector Fuel
Efficiency measures—along with realization rates—through PY 2018.
Table 28. Nonresidential Fuel Efficiency Electric Impact Findings
Fuel Efficiency Measure Reported Savings (kWh) Verified Savings (kWh) Realization Rate
Nonresidential Site Specific 65,061 65,061 100%
Multifamily Market Transformation 207,408 202,324 98%
Total 272,469 267,385 98%

Of the evaluated applications, Cadmus identified discrepancies in the randomly-sampled Multifamily
Market Transformation program measure based on the evaluation site visit and project documentation
review. The site installed more efficient furnaces than reported, which resulted in lower natural gas
energy consumption of the installed units versus baseline efficiency units, meaning that less electricity
was offset for this measure than reported.
Residential Prescriptive Fuel Efficiency Impact Findings
Table 29 shows reported, adjusted, and verified electric energy savings for the Residential Prescriptive
Fuel Efficiency measures. Database review yielded higher savings than reported because of
discrepancies in the UES values used.
Table 29. Residential Prescriptive Fuel Efficiency Electric Impact Findings
Fuel Efficiency Measure Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Verified Electric
Savings (kWh)
Realization
Rate
Residential Prescriptive Fuel Efficiency 1,449,994 1,508,360 1,442,640 100%

In reviewing documentation for 34 Residential Fuel Efficiency measures, Cadmus found an issue with
two measures. Both were natural gas furnaces installed at sites where the furnace replaced an oil-fired
heating system. We eliminated the electricity savings for the natural gas furnaces, because the replaced
system did not heat using electricity. These adjustments led to a document review realization rate of
96%, as shown in Table 30.
Table 30. Residential Prescriptive Fuel Efficiency Electric Impact Document Review Realization Rate
Fuel Efficiency
Measure
2018-2019 Target
Document Audit
Count
Document Audit
Count Achieved
to Date
Sample
Reported
Savings (kWh)
Sample Verified
Savings (kWh)
Document Audit
Realization Rate
Residential Prescriptive
Fuel Efficiency 68 34 343,579 328,609 96%

Low Income Fuel Efficiency Impact Findings
Table 31 shows reported and adjusted electric energy savings for Low Income Fuel Efficiency measures.

24
Table 31. Low Income Fuel Efficiency Program Electric Impact Findings
Fuel Efficiency Measure Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Verified Electric
Savings (kWh) Realization Rate
Low Income Fuel Efficiency 101,882 103,054 103,054 101%

Fuel Efficiency Conclusions and Recommendations
Nonresidential Site Specific and Multifamily Market Transformation Fuel Efficiency measures achieved
verified savings of 267,385 kWh, yielding a 98% realization rate. The Multifamily Market Transformation
Fuel Efficiency measures achieved only 69% of the electric energy savings goal of 299,574 kWh.
Residential Prescriptive Fuel Efficiency measures achieved verified savings of 1,442,640 kWh, yielding a
99% realization rate and achieving 53% of savings goal. Cadmus recommends that Avista update
reported savings to use current TRM UES values, particularly for measures where the differences are
especially notable, such as conversions to natural gas water heaters and conversions to natural gas wall
furnaces.
For Low Income Fuel Efficiency measures, verified savings fell short of Avista’s savings goals, achieving
30% of the savings target and 33% of the participation target.

PY 2019 Idaho Electric
Impact Evaluation Report
May 29, 2020
Prepared for:
Avista Corporation
1411 East Mission Avenue
Spokane, WA 99202

i
Table of Contents
Portfolio Executive Summary ................................................................................................................ 1
Evaluation Methodology and Activities ................................................................................................. 1
Summary of Impact Evaluation Results ................................................................................................. 1
Conclusions and Recommendations ...................................................................................................... 2
Nonresidential Impact Evaluation ......................................................................................................... 5
Program Summary ................................................................................................................................. 5
Program Participation Summary ............................................................................................................ 5
Nonresidential Impact Evaluation Methodology ................................................................................... 7
Nonresidential Impact Evaluation Results ............................................................................................. 9
Nonresidential Conclusions and Recommendations ........................................................................... 13
Residential Impact Evaluation ............................................................................................................. 15
Program Summary ............................................................................................................................... 15
Program Participation Summary .......................................................................................................... 15
Residential Impact Evaluation Methodology ....................................................................................... 16
Residential Impact Evaluation Results ................................................................................................. 21
Residential Conclusions and Recommendations ................................................................................. 24
Low-Income Impact Evaluation ........................................................................................................... 25
Program Summary ............................................................................................................................... 25
Program Participation Summary .......................................................................................................... 25
Low-Income Impact Evaluation Methodology ..................................................................................... 25
Low-Income Impact Evaluation Results ............................................................................................... 26
Low-Income Conclusions and Recommendations ............................................................................... 26
Fuel Efficiency Impact Evaluation ........................................................................................................ 27
Program Summary ............................................................................................................................... 27
Program Participation Summary .......................................................................................................... 27
Fuel Efficiency Impact Evaluation Methodology .................................................................................. 28
Fuel Efficiency Impact Evaluation Results ............................................................................................ 28
Fuel Efficiency Conclusions and Recommendations ............................................................................ 31

ii
Tables
Table 1. Electric Program Evaluation Activities ............................................................................................ 1
Table 2. Reported and Evaluated Electric Savings ........................................................................................ 2
Table 3. Nonresidential Prescriptive Electric Savings ................................................................................... 6
Table 4. Nonresidential Prescriptive Participation Goals by Equipment Rebated ....................................... 6
Table 5. Nonresidential Prescriptive Participation by Project ...................................................................... 7
Table 6. Nonresidential Site Specific Electric Savings ................................................................................... 7
Table 7. Idaho Nonresidential Prescriptive Electric Evaluation Sample ....................................................... 9
Table 8. Idaho Nonresidential Site Specific Electric Evaluation Sample ....................................................... 9
Table 9. Nonresidential Prescriptive Electric Impact Findings ................................................................... 10
Table 10. Nonresidential Prescriptive Evaluation Summary of Discrepancies ........................................... 10
Table 11. Nonresidential Site Specific Electric Impact Findings ................................................................. 12
Table 12. Nonresidential Site Specific Evaluation Summary of Discrepancies ........................................... 12
Table 13. Residential Reported Electric Savings ......................................................................................... 16
Table 14. Residential Participation ............................................................................................................. 16
Table 15. Residential Electric Impact Document Review ........................................................................... 18
Table 16. Residential Database Review Electric Impact Findings ............................................................... 21
Table 17. Residential Electric Impact Document Review Realization Rates ............................................... 21
Table 18. Residential Programs Billing Analysis Results ............................................................................. 22
Table 19. Residential Electric Impact Findings ........................................................................................... 23
Table 20. Low-Income Reported Savings ................................................................................................... 25
Table 21. Low-Income Participationa .......................................................................................................... 25
Table 22. Low-Income Electric Impact Findings ......................................................................................... 26
Table 23. Avista Portfolio Fuel Efficiency Reported Electric Savings .......................................................... 27
Table 24. Avista Portfolio Fuel Efficiency Reported Participation .............................................................. 28
Table 25. Nonresidential Fuel Efficiency Electric Impact Findings ............................................................. 29
Table 26. Residential Fuel Efficiency Analysis Results ................................................................................ 29
Table 27. Residential Fuel Efficiency Electric Impact Findings ................................................................... 30
Table 28. Low-Income Fuel Efficiency Program Electric Impact Findings ................................................... 30

Figure
Figure 1. Residential Impact Process .......................................................................................................... 17

1
Portfolio Executive Summary
For several decades, Avista Corporation has administered demand-side management programs to
reduce the electricity and natural gas energy use of its portfolio of customers. Avista contracted with
Cadmus to complete process and impact evaluations of its program year (PY) 2018 and PY 2019 electric
demand-side management programs in Idaho. This report presents Cadmus’ electric impact evaluation
findings for PY 2019. Cadmus did not apply net-to-gross adjustments to savings values, except in cases
where deemed energy savings values already incorporated net-to-gross as a function of the market
baseline.
Evaluation Methodology and Activities
Cadmus conducted the Idaho portfolio evaluation using a variety of methods and activities, shown in
Table 1.
Table 1. Electric Program Evaluation Activities
Sector Program Document/Database
Review
Verification/Metering
Site Visits Billing Analysis
Nonresidential Prescriptive (Multiple) ü ü --
Site Specific ü ü ü
Residential
Simple Steps, Smart Savings™ ü -- --
HVAC ü -- ü
Shell ü -- ü
ENERGY STAR® Homes ü -- --
Multifamily Direct Install ü -- ü
Multifamily Direct Install
Supplemental Lighting ü -- --
Low-Income Low-Income ü -- ü
Fuel Efficiency
Site Specific (Nonresidential) ü -- --
Residential ü -- ü
Low-Income ü -- ü

Summary of Impact Evaluation Results
Overall, the Idaho electric portfolio achieved a 97% realization rate and acquired 25,230,990 kWh in
annual evaluated savings (Table 2). Cadmus collected Avista’s reported savings through database
extracts from its Customer Care and Billing (residential) and InforCRM and iEnergy (nonresidential)
databases and from data provided by third-party implementers to determine evaluated savings.
Although some individual project results varied, both the Residential and Nonresidential sector
performed strongly in PY 2018 and PY 2019.

2
Table 2. Reported and Evaluated Electric Savings
Sector Reported Savings (kWh) Evaluated Savings (kWh) Realization Rate
Nonresidential 17,826,108 16,443,270 92%
Residential 6,426,003 7,035,960 109%
Low-Income 234,102 232,126 99%
Fuel Efficiency 1,521,494 1,519,634 102%
Total 26,007,707 25,230,990 97%
Note: totals may not sum due to rounding.
Conclusions and Recommendations
During the PY 2019 evaluation, Cadmus identified several areas for improvement, outlined below by
sector.
Nonresidential Conclusions and Recommendations
The Nonresidential sector achieved total evaluated electric energy savings of 16,443 MWh in PY 2019,
with a realization rate of 92%. The Nonresidential sector achieved 78% of the combined Prescriptive and
Site Specific program paths’ electric savings goal of 21,215 MWh.
While some individual project results varied, the overall Nonresidential sector performed strongly in
PY 2019. Most of the projects Cadmus sampled for evaluation were well documented and matched what
the team found during site visit verification.
Cadmus encountered some challenges evaluating the PY 2019 Nonresidential program due to midyear
changes Avista made to its application tracking database system. The new iEnergy database stores and
reports data in different formats and different aggregation levels than the previous system.
As the transition occurred midyear and some applications were entered into both systems, Avista and
Cadmus staff had to manually combine and recategorize data from the new database to match up with
the format used for the old database. Cadmus identified several issues with exports from the new
database as well as underlying errors with the way the new system calculated some savings. Avista has
corrected the issues Cadmus identified, and the new iEnergy database has the potential to facilitate
more accurate savings estimates, more detailed project tracking, and more thorough evaluations in the
future.
Cadmus offers the following recommendations for improving the Nonresidential sector’s energy savings:
• Ensure that the final reported savings calculations reflect the most up-to-date project details,
including post-installation verification photos, equipment submittals, and invoices. During two
project verifications, Cadmus found the installed equipment sizes, quantities, or performance
ratings differed from those used in the reported savings calculations.
• Review hours of use (HOU) estimates for interior and exterior lighting projects when reviewing
submissions and conducting installation verification (IV). Applications claiming 8,760 hours
should be particularly scrutinized. Before any new equipment installations, confirm the presence
or absence of lighting controls and record how they were configured. Cadmus found a small

3
percentage of Prescriptive and Site Specific projects where lighting HOU and controls varied
from the submitted details.
• Ensure lighting projects are correctly categorized as interior or exterior. Cadmus evaluated two
Prescriptive lighting projects with fixtures listed under the wrong measure category.
• Review measurement and verification plans for Site Specific projects carefully and early in the
process to ensure an appropriate measurement basis. Also, work with site contacts to establish
trend logs for relevant building management system or industrial control system data points
during the baseline period.
• Continue to pursue improvements with Avista IV reports. Cadmus staff found that the level of
detail in IV reports varied. Cadmus recommends that all IV reports include basic information and
explicitly state the quantity and type of equipment found. For lighting projects, this would
include confirmed fixture types, quantities, installation locations, controls, and estimated HOU.
For most other equipment, this would include nameplates, model numbers, and quantities.
Residential Conclusions and Recommendations
Evaluated electricity savings show a realization rate of 109% on evaluated savings of 7,036 MWh for the
Residential programs, which is 156% of the savings goal for the year. The high percentage of achieved
savings relative to the goal results largely from high program participation (134% of goal) and the strong
overall realization rate for the Residential sector (109%).
Lighting measures accounted for 73% of the total Residential sector savings. The following list shows the
percentage of Residential evaluated savings provided by each program:
• The Simple Steps, Smart Savings program provided 55% of Residential evaluated savings, mostly
through lighting measures.
• The Multifamily Direct Install and Multifamily Direct Install Supplemental Lighting programs
provided 23% of evaluated savings, again mostly through lighting measures.
• The Residential HVAC program accounted for 19% of evaluated savings.
• The Shell and ENERGY STAR Homes programs accounted for a combined 3% of Residential
evaluated savings.
Realization rates varied by program, from 84% for the Shell program to 202% for the HVAC program,
which resulted in a strong overall realization rate of 109% for PY 2019. Cadmus identified few
discrepancies through the document review, finding that the great majority of projects were well
documented and met program requirements.
Cadmus offers three recommendations regarding Avista’s Residential electric programs:
• Based on billing analysis conducted for this evaluation, adjust Avista’s Technical Reference
Manual (TRM) to provide higher savings values for variable-speed motors installed with the
G Natural Gas Furnace measure and lower savings for replacement windows in electrically
heated homes. The billing analysis showed savings for the variable-speed motor measure nearly
four times the Avista TRM value on average. This was seemingly due to a shift away from

4
secondary electric heating (such as portable heaters or wall heaters) in some homes after
replacing a gas furnace with a high-efficiency model. For replacement windows in electrically
heated homes, the billing analysis estimated unit savings of 72% the 2019 TRM value.
• The MFDI program has proven to be an efficient, effective mechanism for installing high-
efficiency lighting and aerators in multifamily units. Continue to focus on replacing high-use,
low-efficiency lamps where practical, to maximize program cost-effectiveness while maintaining
high savings.
• Ensure that reported savings for all measures are calculated using current TRM or Regional
Technical Forum unit energy savings (UES) values, and that the TRM provides values for all
measures. Cadmus did not find large-scale problems with the PY 2019 measure tracking data,
but the team did note numerous measure-tracking records that reported zero savings, despite
the record showing the measure was completed and that a rebate was issued. In addition, some
instances of PY 2019 measures used UES values from the 2018 TRM, and reported values for
some measures (most notably, smart thermostats) did not match TRM values.
Fuel Efficiency Recommendations
Nonresidential Site Specific and Multifamily Market Transformation Fuel Efficiency measures achieved
evaluated savings of 300,230 kWh, yielding a 100% realization rate. The Multifamily Market
Transformation Fuel Efficiency measures exceeded the electric energy savings goal of 234,960 kWh by
28%. Cadmus does not recommend any changes to the Nonresidential Site Specific and Multifamily
Market Transformation Fuel Efficiency programs.
Residential Fuel Efficiency measures achieved evaluated savings of 1,181,596 kWh, yielding a 102%
realization rate and achieving 118% of savings goal. Cadmus recommends that Avista update TRM values
to match measure-level UES values calculated by the billing analysis. Cadmus also recommends that
Avista ensure all measures are represented in the TRM.
For Low-Income Fuel Efficiency measures, evaluated savings were 37,808 kWh, with a realization rate of
100%, but fell short of Avista’s savings goals, achieving 37% of the savings target and 43% of the
participation target. Billing analysis indicated that program electric savings are likely higher, based on
the billing analysis realization rate of 144% for Low-Income Fuel Efficiency measures as a whole. Based
on this finding, Cadmus recommends increasing the Avista TRM’s UES values.

5
Nonresidential Impact Evaluation
Through its Nonresidential portfolio of programs, Avista promotes the purchase of high-efficiency
equipment for commercial and industrial utility customers. Avista provides rebates to partially offset the
difference in cost between high-efficiency equipment and standard equipment. Cadmus conducted
Nonresidential impact evaluation activities to determine program year (PY) 2019 evaluated savings for
most programs; the team conducted measurement and verification of Prescriptive and Site Specific
projects across the full PY 2019 sample.
Program Summary
Avista completed and rebated 542 nonresidential electric measures in Idaho in PY 2019 and reported
total electric energy savings of 17,826,108 kWh. Through the Nonresidential sector, Avista offers
incentives for high-efficiency equipment and controls through three program paths: Prescriptive, Site
Specific, and Multifamily Market Transformation.
The Prescriptive program path applies to smaller, straightforward equipment installations that generally
have similar operating characteristics (such as lighting, simple HVAC systems, food service equipment,
and variable-frequency drives). The Site Specific program path applies to more unique projects that
require custom savings calculations and technical assistance from Avista’s account executives (such as
compressed air, process equipment and controls, and comprehensive lighting retrofits).
Multifamily Market Transformation, a Site Specific program, prompts building owners and developers to
consider natural gas as the fuel of choice when constructing new multifamily housing. These measures,
represented by a combination of electric savings and natural gas penalties, typically involve replacing
electric space-heating or water-heating systems with natural gas equipment. See the Fuel Efficiency
Impact Evaluation section for the evaluation methodology and the results discussion for Nonresidential
Fuel Efficiency measures.
Program Participation Summary
This section summarizes Nonresidential sector participation and progress toward PY 2019 goals through
the Prescriptive and Site Specific program paths.
Nonresidential Prescriptive Program Path
Table 3 shows electric energy savings goals assigned to Avista’s Nonresidential Prescriptive program
path for PY 2019, as well as reported savings and a comparison between reported savings and goals.

6
Table 3. Nonresidential Prescriptive Electric Savings
Program Name Savings Goals (kWh) Savings Reported
(kWh) Percentage of Goal
Interior Lighting 7,414,179 4,669,357 63%
Exterior Lighting 4,299,232 3,192,110 74%
Shell Measure 1,109 8,871 800%
Green Motors 49,098 38,828 79%
Motor Control (Variable Frequency Drives [VFD]) 75,595 0 0%
Fleet Heat 8,000 0 0%
Food Service Equipment 32,429 9,506 29%
AirGuardian 18,000 136,244 757%
Energy Smart Grocera 317,248 0 0%
Total 12,214,890 8,054,916 66%
a The Energy Smart Grocer savings goal includes Site Specific Energy Smart Grocer measures. The Site Specific portion
constitutes approximately 10% of the overall goal.

Table 4 shows participation goals by rebated equipment quantity, as provided by Avista. The PY 2019
nonresidential tracking database extract listed individual projects but did not include rebated equipment
quantity. For reference, Table 5 provides participation by unique application numbers.
Table 4. Nonresidential Prescriptive Participation Goals by Equipment Rebated
Program Type Planned Participation
Interior Lighting 121,200
Exterior Lighting 9,850
Shell Measure 435
Green Motors 17
Motor Control (VFD) 55
Fleet Heat 1
Food Service Equipment 6
AirGuardian 3
Energy Smart Grocera 814
a The Energy Smart Grocer goal includes Site Specific Energy Smart Grocer participants.

7
Table 5. Nonresidential Prescriptive Participation by Project
Program Type Participation Reporteda
Interior Lighting 249
Exterior Lighting 260
Shell Measure 6
Green Motors 12
Motor Control (VFD) 0
Fleet Heat 0
Food Service Equipment 3
AirGuardian 1
Energy Smart Grocer 0
Totalb 492
a Participant is defined as a unique application number.
b Total unique applications. One application may contain measures from multiple programs.

Nonresidential Site Specific Program Path
Table 6 shows electric savings goals assigned to the Site Specific program path in Avista’s Nonresidential
sector for PY 2019, as well as reported savings. The table does not include reported electric savings for
the Fuel Efficiency sector, such as those associated with the Multifamily Market Transformation
program.
Table 6. Nonresidential Site Specific Electric Savings
Program Path Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Site Specific 9,000,000 9,771,192 109%

Nonresidential Impact Evaluation Methodology
As the first step in evaluating PY 2019 savings for the Nonresidential sector, Cadmus explored the
following documents and data records to gain an understanding of the programs and measures slated
for evaluation:
• Avista’s annual business plans, detailing processes and energy savings justifications
• Project documents from external sources (such as customers, program consultants, or
implementation contractors)
Based on the initial review, Cadmus checked the distribution of program contributions with the overall
program portfolio. The review provided insight into the sources for unit energy savings (UES) claimed for
each measure offered in the programs, along with sources for energy-savings algorithms, internal
quality assurance, and quality control processes for large Nonresidential sector projects.

8
Following this review, Cadmus designed a sample strategy for impact evaluation activities. Cadmus
performed the following evaluation activities in two waves:
• Selected evaluation sample and requested project documentation from Avista
• Reviewed project documentation
• Prepared on-site measurement and verification plans
• Performed site visits and collected on-site data (such as trend data, photos, and
operating schedules)
• Used site visit findings to calculate evaluated savings by measure
• Applied realization rates to the total reported savings population to determine overall program
year evaluated savings
Sample Design
Cadmus created two sample waves for PY 2019. Sample 1 included program data from January 2019
through June 2019, and sample 2 included program data from July 2019 through December 2019. As a
guideline, Cadmus used the proposed, overall PY 2019 nonresidential sample sizes by subprogram in the
measurement and verification plan, seeking to complete approximately half of the sample in each wave.
For each activity wave, Cadmus organized submitted program applications by path and measure (such as
the Site Specific Shell Measure, Prescriptive Lighting, or Prescriptive Motor Controls), allowing the team
to select the highest-savings applications in each category with certainty. For non-certainty applications,
the team assigned random numbers and developed a random sample. In some cases, Cadmus sampled
another application at the same location or a facility that was previously selected (and where the team
could assess both applications with one site visit). This was a cost-effective verification strategy even if
the second application represented minimal claimed savings.
As Avista implements its programs similarly in both states, Cadmus sampled randomly selected sites
across Washington and Idaho. The team pooled results from the randomly selected sites to calculate a
realization rate by stratum and applied that realization rate to projects in both states. Cadmus applied
evaluated savings for sites selected with certainty only to the state in which they had been
implemented.
Table 7 summarizes the Idaho Nonresidential Prescriptive program evaluation sample. In Idaho, Cadmus
sampled 18 Prescriptive applications at 14 unique sites. Of the sampled applications, the team selected
three for certainty review based on the scale of savings, measure type, or location. Cadmus then
selected the remaining 15 applications randomly. No customers participated in the Fleet Heat, Motor
Control, and Energy Smart Grocer programs in Idaho in PY 2019.

9
Table 7. Idaho Nonresidential Prescriptive Electric Evaluation Sample
Program Type Applications Sampled Sampled Savings (kWh) Percentage of Reported Savings
Interior Lighting 7 576,688 12%
Exterior Lighting 5 26,001 1%
Shell Measure 1 3,920 44%
Green Motors 4 19,706 51%
Food Service Equipment 2 4,393 46%
AirGuardian 1 136,244 100%
Nonresidential Prescriptive 18 766,951 10%
Note: Totals may not sum due to rounding. Two applications contained both interior and exterior lighting measures.

Table 8 summarizes the Idaho Nonresidential Site Specific program path evaluation sample. In Idaho,
Cadmus sampled five Site Specific applications at two unique sites. Of the sampled applications, the
team selected four for certainty review based on the scale of savings, measure type, or location. Cadmus
selected the remaining application randomly.
Table 8. Idaho Nonresidential Site Specific Electric Evaluation Sample
Program Path Applications Sampled Sampled Savings (kWh) Percentage of Reported Savings
Site Specific 5 7,737,047 79%

Document Review
Cadmus requested and reviewed project documentation for each sampled application and prepared
measurement and verification plans to guide its site visits. Typically, project documentation included
incentive applications, calculation tools (usually based on the 2017 Regional Technical Forum [RTF]),1
invoices, equipment specification sheets, and installation verification (IV) reports.
On-Site Verification
Cadmus performed site visits at 16 unique nonresidential locations to assess electric savings for
25 unique Prescriptive and Site Specific measures (not including Fuel Efficiency measures). Site visits
involved verifying the installed equipment type, make and model numbers, operating schedules, and
setpoints, as applicable. Cadmus used the project documentation review and on-site findings to adjust
reported savings calculations where necessary. The team did not consider it necessary to conduct power
metering or light logging for PY 2019 site visits and used trend data provided by the participant to
evaluate Site Specific industrial process measures.
Nonresidential Impact Evaluation Results
This section summarizes the Nonresidential Prescriptive and Site Specific program paths’ electric impact
evaluation results for PY 2019.

1 Regional Technical Forum. 2017. “Standard Protocols.” https://rtf.nwcouncil.org/standard-protocols

10
Nonresidential Prescriptive Programs
Table 9 shows reported and evaluated electric energy savings for Avista’s Nonresidential Prescriptive
program and the realization rates between evaluated and reported savings for PY 2019. The overall
Nonresidential Prescriptive program path electric realization rate was 100%.
Table 9. Nonresidential Prescriptive Electric Impact Findings
Program Type Reported Savings (kWh) Evaluated Savings (kWh) Realization Rate
Interior Lighting 4,669,357 4,518,758 97%
Exterior Lighting 3,192,110 3,303,660 103%
Shell Measure 8,871 10,400 117%
Green Motors 38,828 38,828 100%
Food Service Equipment 9,506 9,506 100%
AirGuardian 136,244 136,244 100%
Nonresidential Prescriptive 8,054,916 8,017,396 100%

Of the evaluated applications, Cadmus identified discrepancies for 18 based on the site visit and project
documentation review. Table 10 summarizes the reasons for discrepancies between reported and
evaluated savings.
Table 10. Nonresidential Prescriptive Evaluation Summary of Discrepancies
Project Type Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
Attic
Insulation 1 ↓
• Avista reported incorrect savings values for one attic insulation project due
to an error in its new database software. Cadmus reviewed all Prescriptive
Shell measures to confirm that only one project was affected by the bug.
Cadmus treated the affected project as a certainty project and evaluated
savings using the typical savings calculator methodology.
Refrigerated
Cases 1 ↓
• Cadmus was only able to verify installation of 15 of the 17 refrigerator doors
claimed on the application of one refrigerated cases measure and reduced
the savings proportionally.

11
Project Type Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
Interior
Lighting
5 ↓
• Cadmus reduced the fixture counts for one project as the evaluated
installed quantity on the site was lower than the quantity reported on
the application.
• Cadmus reduced the hours of use (HOU) for one project that reported 24/7
operations after determining that occupancy controls and schedule controls
were in place to reduce the lighting runtime prior to and after the project.
• The Avista database categorized two projects as interior lighting that only
had exterior fixtures. These savings were subtracted from interior lighting
and added to exterior lighting.
• Cadmus reduced fixture counts and increased HOU at one site where the
building underwent a remodel shortly after completing the project and no
longer matched the conditions reported at the time the application
was submitted.
4 ↑
• Cadmus determined that the store hours at one site were higher than
reported on the application. The team also determined that new occupancy
controls were added that were not reported on the application, further
decreasing installed HOU relative to baseline HOU.
• Cadmus found that the installed fixtures for one project had a lower
wattage than reported on the application.
Exterior
Lighting
3 ↓
• Cadmus reduced fixture counts and increased HOU at one site where the
building underwent a remodel shortly after completing the project and no
longer matched the conditions reported at the time the application
was submitted.
• Cadmus calculated savings for an outdoor display sign using the actual
quantity and wattage of the lamps inside the sign. The Avista calculator
used an estimated watts-per-square-foot method for exterior sign lighting
based on assumed typical values. The team found the assumed baseline
watts per square foot to be unreasonably high for the type of lighting
typically installed in outdoor signs.
4 ↑
• Cadmus updated the savings calculations to use the actual verified fixture
wattage instead of the assumed typical value for three projects.
• Cadmus determined that two exterior lighting measures were incorrectly
categorized as interior lighting measures in the Avista database and
transferred those savings to exterior lighting.

Throughout the evaluation, Cadmus found that the level of detail in IV reports varied. Most IV reports
the team reviewed only stated that the reviewer “found the installation to match the application
submitted,” including for a portion of projects where the inspections found discrepancies between the
installation and the application. Some IV reports did not contain any text at all and only provided
unlabeled photos. Cadmus evaluated a lighting project where the IV report only contained one
photograph of each fixture type and no information about quantities.
Nonresidential Site Specific Program
Table 11 shows reported and evaluated electric energy savings for Avista’s PY 2019 Nonresidential Site
Specific program path, as well as a comparison between evaluated and reported savings for PY 2019.

12
The overall Site Specific program path electric realization rate was 86%. The table does not include
reported and evaluated electric savings for measures in the Fuel Efficiency path.
Table 11. Nonresidential Site Specific Electric Impact Findings
Program Path Reported Savings (kWh) Evaluated Savings (kWh) Realization Rate
Site Specific 9,771,192 8,425,874 86%

Of the evaluated applications, Cadmus identified discrepancies in three based on the site visit and
project documentation review. Table 12 summarizes the reasons for discrepancies between reported
and evaluated savings.
Table 12. Nonresidential Site Specific Evaluation Summary of Discrepancies
Project Type Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
Green Motor
Rewind 1 ↑
• The reported savings reference for 2017 RTF. Cadmus applied deemed
motor savings from the 2018 Technical Reference Manual (TRM)
workbook.
Refrigerator Door
Gaskets 1 ↓
• The reported savings for one refrigerator door gasket project
corresponded to 17 doors. Cadmus only received documentation for
and verified installation of 15 doors at this site.
Interior Lighting
14 ↓
• Cadmus reduced the fixture counts for three projects as the verified
installed quantity on the site was lower than the quantity reported on
the application.
• Cadmus reduced the HOU for four projects that reported 24/7
operations after determining that occupancy controls and schedule
controls were in place to reduce the lighting runtime prior to and after
the project.
• The Avista database categorized two projects as interior lighting that
only had exterior fixtures. These savings were subtracted from interior
lighting and added to exterior lighting.
• Cadmus reduced the lighting hours from 100% on to 75% on one
project, based on interviews with on-site staff. Cadmus also found a
lower installed fixture quantity than that reported in the application.
• Cadmus could not replicate the reported savings on one project based
on reported fixture types and quantities. However, the team retained
the reported quantities as they could not visit all spaces at the site for
verification.
• Cadmus determined that 13 W fixtures were installed in place of the
9 W fixtures reported on the application.
2 ↑
• Cadmus determined that the store hours at one site were higher than
reported on the application. The team also determined that new
occupancy controls were added which were not reported on the
application, further decreasing installed HOU relative to baseline HOU.
• Cadmus found that the installed fixtures for one project had a lower
wattage than reported on the application.
Exterior Lighting 3 ↓
• Cadmus reduced exterior lighting HOU from 8,760 to 4,288 for one
project after determining that all exterior fixtures at the site were
controlled by photocells.

13
Project Type Number of
Occurrences
Savings
Impact Reason(s) for Discrepancy
• Cadmus reduced fixture counts and increased HOU at one site where
the building underwent a remodel shortly after completing the project
and no longer matched the conditions reported at the time the
application was submitted.
• Cadmus calculated savings for an outdoor display sign using the actual
quantity and wattage of the lamps inside the sign. The Avista calculator
used an estimated watts-per-square-foot method for exterior sign
lighting based on assumed typical values.
4 ↑
• Cadmus updated the savings calculations to use the actual verified
fixture wattage instead of the assumed typical value for two projects.
• Cadmus determined that two exterior lighting measures were
incorrectly categorized as interior lighting measures in the Avista
database and transferred those savings to exterior lighting.
Motor Control
(VFD) 2 ↓ • Cadmus determined that two return air fans with VFDs and reported as
3.0 horsepower were actually 2.5 horsepower.
Shell Measure 1 ↓
• Cadmus determined there was no space cooling and space was heated
with natural gas. As a result, the team removed electric savings from
ceiling/wall insulation.
2 ↓
• Avista reported incorrect savings values for two Shell insulation projects
due to an error in their new database software. Cadmus reviewed all
Prescriptive Shell measures to confirm that only two projects were
affected by the bug. Cadmus treated the two affected projects as
certainty projects and evaluated savings using the typical savings
calculator methodology.
Industrial Process 2 ↑
• Cadmus recalculated savings for two motor replacement and VFD
installation projects in a paper mill based on trend data from the post-
installation period. The team found that the average kilowatt
consumption of some installed motors was lower than predicted.
Industrial Motor
Controls 1 ê
• Cadmus determined that the baseline power consumption estimation
for a motor replacement project included unrelated equipment from
the same power distribution bus. Cadmus revised the analysis using
additional trend data and updated assumptions to ensure the baseline
and post-installation calculations were consistent. The team found the
estimated power consumption in both periods to be lower than
reported in the original analysis, but significantly lower in the baseline,
resulting in reduced savings.

Nonresidential Conclusions and Recommendations
The Nonresidential sector achieved total evaluated electric energy savings of 16,443 MWh in PY 2019,
with a combined realization rate of 92%. The Nonresidential sector achieved 78% of the combined
Prescriptive and Site Specific program paths’ electric goal of 21,215 MWh.
While some individual project results varied, the overall Nonresidential sector performed strongly in
PY 2019. Most of the projects Cadmus sampled for evaluation were well documented and matched what
the team found during site visit verification.

14
Cadmus encountered some challenges evaluating the PY 2019 Nonresidential program due to midyear
changes Avista made to their application tracking database system. The new iEnergy database stores
and reports data in different formats and different aggregation levels than the previous system.
As the transition occurred midyear and some applications were entered into both systems, Avista and
Cadmus staff had to manually combine and recategorize data from the new database to match up with
the format used for the old database. Cadmus identified several issues with exports from the new
database as well as underlying errors with the way some savings were calculated by the new system.
Avista has corrected the issues Cadmus identified, and the new iEnergy database has the potential to
facilitate more accurate savings estimates, more detailed project tracking, and more thorough
evaluations in the future.
Cadmus offers the following recommendations for improving the Nonresidential sector’s energy savings:
• Ensure that the final reported savings calculations reflect the most up-to-date project details,
including post-installation verification photos, equipment submittals, and invoices. During two
project verifications, Cadmus found different installed equipment sizes, quantities, or
performance ratings than used in the reported savings calculations.
• Review HOU estimates for interior and exterior lighting projects when reviewing submissions
and conducting IV. Applications claiming 8,760 hours should be particularly scrutinized. Before
any new equipment installations, confirm the presence or absence of lighting controls and
record how they were configured. Cadmus found several Prescriptive and Site Specific projects
where lighting HOU and controls varied from submitted details.
• Ensure the correct categorization of lighting projects as interior or exterior. Cadmus evaluated
two Prescriptive lighting projects with fixtures listed under the wrong measure category.
• Review measurement and verification plans for Site Specific projects carefully early in the
process to ensure an appropriate measurement basis, and work with site contacts to establish
trend logs for relevant building management system or industrial control system data points
during the baseline period.
• Provide more thorough documentation with Avista IV reports. Cadmus staff found that the level
of detail in IV reports varied. Cadmus recommends that all IV reports include basic information,
explicitly stating the quantity and type of equipment found. For lighting projects this would
include confirmed fixture types, quantities, installation locations, controls, and estimated HOU.
For most other equipment, this would include nameplates, model numbers, and quantities.

15
Residential Impact Evaluation
Cadmus designed the Residential sector impact evaluation to verify reported program participation and
energy savings. The team used data collected and reported in the tracking database, online application
forms, Avista TRM and RTF savings review, and analysis of participant electricity consumption data to
evaluate savings. This approach provided the strongest estimate of achieved savings practical for each
program, given its delivery method, magnitude of savings, number of participants, and availability of
billing data.
Program Summary
Avista completed and rebated 349,056 residential electric measures or units in Idaho in PY 2019 and
reported total electric energy savings of 6,426,003 kWh, not including participation and savings from
Fuel Efficiency measures, which are included in the Fuel Efficiency Impact Evaluation section.
Participation is defined as installed pieces of equipment (such as a furnace or showerhead) for some
measures and square feet of surface for others (such as wall insulation and windows replacement).
The Residential path includes several programs:
• Simple Steps, Smart Savings, which encourages consumers to purchase and install high-quality
LEDs, light fixtures, and energy-efficient showerheads
• Residential HVAC, which offers incentives for high-efficiency heating and cooling equipment
• Residential Shell, which provides rebates to encourage customers to install high-efficiency
windows and storm windows
• ENERGY STAR Homes, which offers 15% to 25% of energy savings relative to state energy codes
• Multifamily Direct Install (MFDI), which provides free direct-install measures to multifamily
residences (five units or more) and common areas
• MFDI Supplemental Lighting, which revisited multifamily properties served by the MFDI program
to install additional common area lighting.
Program Participation Summary
This section summarizes Residential sector participation and progress toward PY 2019 goals.
Residential Programs
Table 13 shows savings goals assigned to Avista’s Residential sector programs for PY 2019, as well as
reported savings and the goal portion achieved in PY 2019. All programs except ENERGY STAR Homes
and Residential HVAC exceeded savings goals based on reported savings, leading to an overall
achievement of 142% of the savings goal for Residential programs.

16
Table 13. Residential Reported Electric Savings
Program Savings Goals (kWh) Savings Reported (kWh) Percentage of Goal
Simple Steps, Smart Savings 2,495,393 3,879,137 155%
HVAC 674,367 659,957 98%
Shell 139,065 190,390 137%
ENERGY STAR Homes 86,190 66,262 77%
Multifamily Direct Install 957,450 1,289,539 135%
Multifamily Direct Install Supplemental Lighting 168,000 340,719 203%
Residential Total 4,520,465 6,426,003 142%
Note: totals may not sum due to rounding.

Table 14 summarizes participation goals and reported participation in Avista’s Residential sector
programs for PY 2019, along with the percentage of goal achieved.
Table 14. Residential Participation
Program Participation Goals Participation Reported Percentage of Goal
Simple Steps, Smart Savingsa 190,126 317,124 167%
HVACb 462 750 162%
Shellc 67,184 27,404 41%
ENERGY STAR Homesb 26 18 69%
Multifamily Direct Installd 1,473 3,057 208%
Multifamily Direct Install Supplemental Lightinge 750 703 94%
Residential Total 260,021 349,056 134%
a Participation is defined as the number of purchased units.
b Participation is defined as the number of rebates.
c Participation is defined as square feet of installed windows or storm windows.
d Participation is defined as the number of living units and common areas served.
e Participation is defined as the number of installed units.

Residential Impact Evaluation Methodology
To determine the Residential sector’s evaluated savings for PY 2019, Cadmus employed a combination
of three impact evaluation methods:2
• Database review
• Document review
• Billing analysis
First, Cadmus calculated adjusted savings for each program based on results of a database review. For
the HVAC, Shell, and Fuel Efficiency programs, Cadmus also applied realization rates for the document

2 With approval from Avista, Cadmus ceased performing a fourth impact activity—verification surveys—in Q3
PY 2018; this eliminated redundancy between verification surveys and document reviews.

17
reviews. For these programs, the team calculated prescriptive evaluated savings by multiplying adjusted
savings by the document review realization rate, as shown in Figure 1.
Figure 1. Residential Impact Process

To provide the most rigorous evaluation method where practical, Cadmus analyzed consumption data
for all available participants of the HVAC, Shell, Fuel Efficiency, and MFDI programs. As described in
more detail in the Billing Analysis section, the team applied billing analysis results to determine
evaluated savings only for measures where savings could be isolated (that is, where a sufficient number
of participants could be identified who installed only that measure) and where confidence and precision
met specific targets. Program-level realization rates for the HVAC, Shell, and Fuel Efficiency programs
incorporate billing analysis results for some measures.
Database Review
For the impact evaluation database review, Cadmus used UES values, as provided in the TRM, to
calculate savings for measures reported in the measure tracking database. This impact activity may help
identify incorrect UES values used to calculate reported savings. Savings calculated during the database
review are defined as adjusted savings.
Document Review
For the document review, Cadmus compared information from rebate forms and other supporting
documents to measure tracking data for a random sample of projects. This impact activity may identify
installed measures that did not meet eligibility requirements, quantities that did not match the measure
tracking database, and other discrepancies. Following a review of all projects, Cadmus calculated a
realization rate for the document review by dividing savings calculated for the sample (using the revised
information) by reported savings for the sample. The team then multiplied this realization rate by
adjusted savings for the entire program to determine prescriptive evaluated savings for PY 2019.
Cadmus conducted document reviews for the programs shown in Table 15, drawing roughly equal
samples from participants in each quarter.

18
Table 15. Residential Electric Impact Document Review
Program Completed through Q2 PY 2019
HVAC 51
Shell 51

Billing Analysis
For the Residential sector, Cadmus conducted billing analysis using available electricity and natural gas
consumption data from Avista for the HVAC, Shell, Fuel Efficiency, and MFDI programs. Evaluating
Simple Steps, Smart Savings program savings through billing analysis was not practical because
participants of the midstream retail program were largely unknown. The ENERGY STAR Homes program
had too few participants to produce meaningful billing analysis results.
HVAC, Shell, and Fuel Efficiency Savings Estimates
With the HVAC, Shell, and Fuel Efficiency programs, Cadmus eliminated the effects of multiple energy
efficiency measures by only including participants in the analysis who installed one measure. With these
programs, the goal was to provide average unit savings values at the measure level to ensure the most
accurate values possible were used for evaluated savings and cost-effectiveness.
Cadmus used the unit savings value provided by the billing analysis for a given measure when results for
that measure met two requirements: the number of sites in the participant group was at least five, and
the relative precision achieved was no greater than ±40% at the 90% confidence level. If results
calculated using only Idaho participants met these requirements, the team used those results. If results
based only on Idaho participants failed to meet the requirements, Cadmus used combined results for
Idaho and Washington if those results passed. If no billing analysis results passed for a given measure,
Cadmus applied the results of database review and document review to determine evaluated savings.
Data Sources
To conduct the consumption analysis, Cadmus used program measure tracking data provided by Avista,
monthly electric and gas consumption data provided by Avista, and weather data (which included actual
average daily temperatures for 13 weather stations in Idaho and Washington from the National Oceanic
and Atmospheric Administration) for the billing analysis period. The team used zip codes to match daily
heating and cooling degree days to respective monthly bill read dates. Additionally, Cadmus used typical
meteorological year (TMY3) 15-year normal weather values from 1991–2005, obtained from National
Oceanic and Atmospheric Administration for the same weather stations, in assessing energy use under
normal weather conditions.
Participant and Comparison Group Designation
Cadmus gathered data for a participant (treatment) group comprising all HVAC, Shell, and Fuel Efficiency
program participants with measures installed in 2018. This allowed for enough pre- and post-
consumption data to analyze the various measures’ effects.

19
To isolate the impact of exogenous factors (such as energy rate changes, economic condition changes,
and non-programmatic effects) on energy use, Cadmus utilized a quasi-experimental3 design that
involved selection of a comparison group, composed of participants with installation dates in late
PY 2019. Through this approach, the team compared the treatment group’s pre- and post-change
energy use (assumed to capture the program treatment) to the comparison group’s change in energy
use (reflecting what would have happened absent the program). To ensure similarity between
treatment and control groups, the team chose to use future participants as the comparison group
because they would have similar qualifications and could be assumed to have not participated in energy
efficiency programs prior to program treatment.
Data Screening
Starting with all HVAC, Shell, and Fuel Efficiency participants and the comparison group, Cadmus cleaned
the data and screened for several criteria to identify final analysis samples. Data cleaning included
performing account-level reviews of the pre- and post-period monthly consumption of all individual
participants to identify anomalies (such as periods of unoccupied units) that could bias the results.
Cadmus conducted the consumption analysis using participants who had not moved since participating
and who had at least 10 months of pre- and post-period billing data.
Cadmus applied several screens to remove anomalies, incomplete records, and outlier accounts. The
following are examples of accounts excluded from the analyses:
• Accounts missing records, prohibiting the team from merging participant program tracking data
with consumption data.
• Accounts with low annual use in the pre- or post-period, such as less than 1,240 kWh annually.
• Customers with incorrect signs on Princeton Scorekeeping Method (PRISM) parameter
estimates.
• Accounts with other extreme values, including vacancies in billing data (outliers), non-program-
related heating or cooling system changes (such as added or removed heating or cooling loads),
baseload equipment changes, or changes in occupancy. This included screening for accounts
with large gaps in interval data, such as having zero consumption across multiple months.
Analysis
To estimate measure-level impacts, Cadmus employed a pre- and post-installation savings analysis using
household-level PRISM models that accounted for differences in pre- and post-installation weather
conditions. The team estimated the heating and cooling PRISM model using variable 45°F to 85°F
heating and cooling bases in both the pre- and post-period for each customer.

3 A quasi-experimental design is when treatment and control groups are not randomized prior to treatment. In
this case, the comparison group was created after the treatment had occurred and participants self-selected
the treatment.

20
Multifamily Direct Install
With the MFDI program, isolating individual measures was not possible, because most living units
received a range of LED light bulbs as well as water saving measures such as aerators and showerheads.
To provide an accurate estimate of the energy savings for the program a whole, Cadmus performed a
complex and rigorous evaluation involving matching tracking data with billing data at the account level.
Cadmus estimated weather-normalized facility level usage. There were two main components of usage
that were combined to develop the pre- and post-facility level usage estimates: unit-level usages and
common area usages.
Cadmus referenced the same data sources for MFDI consumption analysis as those identified for HVAC,
Shell, and Fuel Efficiency analyses (see Data Sources section above) as well as the participant and
comparison group approach to isolate the impact of exogenous factors (see Participant and Comparison
Group Designation). Additionally, Cadmus cleaned the data to remove anomalies, incomplete records,
and outlier accounts (see Data Screening).
Analysis
To estimate program impacts, Cadmus employed a pre- and post-installation savings analysis using
household-level PRISM models that account for differences in pre- and post- installation
weather conditions.
Cadmus estimated the heating and cooling PRISM model using variable 45°F to 85°F heating and cooling
bases in both the pre- and post-period for each MFDI unit and common area account. Because some
units in a facility could not be matched to billing data or did not pass the screening process, the team
found it necessary to extrapolate the available weather normalized pre- and post-period unit level
PRISM usages to the facility level for all units. For each facility, the number of units in the facility was
known. To obtain the final unit level component, Cadmus calculated the average pre-period usage, post-
period usage, and savings per unit. Cadmus then multiplied those per-unit values by the number of units
in the facility to obtain the total unit component facility usages, savings, and ex ante estimates. If the
facility also had a common area component, Cadmus added usage for that area to the facility level unit
usage component to develop the final total facility usage.
Cadmus then applied weighting to calculate the final program savings estimate. A facility with 100 units
has more weight than a facility with 10 units. The final savings estimates and ex ante estimates were
weighted by the number of units.
The MFDI Idaho participant group showed a reliable relative precision estimate of ±17% at the 90%
confidence level for the 34 facilities included in the analysis and savings of 5.8%. The comparison group
had only four facilities and showed a reduction in usage of 1.2%; however, with very high relative
precision estimate of ±303%. This large confidence band around the comparison group shows that the
change in usage was not significantly different than zero (that is, a savings increase was within the error
bound), so a comparison group adjustment was not applied.

21
Residential Impact Evaluation Results
The following sections summarize findings and provide evaluated savings for both of Cadmus’ impact
evaluation methodologies. The database review resulted in the largest number of adjustments to
reported savings.
Database Review
Table 16 shows database review findings, with adjusted savings being higher than reported savings for
some programs and lower for others. Adjusted savings differed from reported savings because reported
UES values differed from TRM values for several measures. The larger adjusted savings for the HVAC,
ENERGY STAR Homes, and Shell programs resulted partly from some instances where the tracking data
reported zero energy savings, despite the records showing the projects were complete and rebates were
paid. For the MFDI program, adjustments included applying RTF UES values for multifamily direct-install
aerators that were lower than the older values used by the implementer. The discrepancy with MFDI
Supplemental Lighting resulted mostly from the omission of heating interactive effects for measures in
common areas indicated as heated.
Table 16. Residential Database Review Electric Impact Findings
Program Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Percentage
Change
Simple Steps, Smart Savings 3,879,137 3,879,137 0%
HVAC 659,957 684,425 4%
Shell 190,390 216,838 14%
ENERGY STAR Homes 66,262 69,615 5%
Multifamily Direct Install 1,289,539 1,258,897 -2%
Multifamily Direct Install Supplemental Lighting 340,719 332,718 -2%
Residential Total 6,426,003 6,441,629 0%
Note: totals may not sum due to rounding.

Document Review
Table 17 summarizes document review findings. The HVAC program had a 100% electric realization rate,
and the Shell program had an 90% electric realization rate.
Table 17. Residential Electric Impact Document Review Realization Rates
Program Document Audit
Count
Sample Reported
Savings (kWh)
Sample Evaluated
Savings (kWh)
Document Review Realization
Rate
HVAC 51 50,106 48,800 97%
Shell 51 73,925 64,268 87%

Cadmus’ document review (through Q2 PY 2019) identified the following discrepancies:
• For four window measures, documentation showed a square footage for installed windows that
differed from the reported window area. In three cases, the documented window area was
lower than the reported area and resulted in lower evaluated savings. In one case, the

22
documented window area was more than that reported and resulted in higher evaluated savings
based on the corrected area.
• For four window measures reported for sites with electric heating, project documents identified
heating fuels other than electricity. Cadmus added natural gas savings and removed electricity
savings at two sites identified as using natural gas heating. Documentation for the other two
sites identified liquid propane as the heating fuel for one site and wood pellets as the fuel for
the other; consequently, Cadmus removed electricity savings for these sites.
• One PY 2018 heat pump water-heater measure had a tank capacity of 80 gallons, per the
documentation. However, conditions for the rebate required a tank size below 55 gallons in
PY 2018; consequently, Cadmus removed savings for this measure.
Billing Analysis
Table 18 shows measure-level billing analysis results, used when calculating evaluated electric energy
savings. The participant count and relative precision for each measure easily met requirements
established to ensure meaningful results, which required a participant count of at least five and a
relative precision no greater than ±40% at the 90% confidence level.
Table 18. Residential Programs Billing Analysis Results
Measure
2019 Avista
TRM UES
(kWh)
na
Pre-Installation
Weather
Normalized
Usage (kWh)
Annual UES
(kWh)
Realization
Rate
Relative
Precision at
90%
Confidence
Participant
State
E Variable
Speed Motorb 414.00 195 12,251 1,528.63 369% 30% Idaho
E Storm
Window with
Electric Heat
10.30 15,876 17,543 11.00 107% 26% Idaho and
Washington
E Window
Replc from
Single Pane W
Electric Heatb
15.25 15,876 17,543 11.00 72% 26% Idaho and
Washington
a To provide unit savings values that align with TRM units, this table presents participant count in sq. ft. of
window surface for storm widow and replacement window measures.
b Results shown represent combined analysis of storm window and window-replacement measures, to
maximize relative precision. Separate results for each measure appeared similar.

Billing analysis results showed surprisingly high savings for the E Variable Speed Motor measure, with a
realization rate of 369% relative to the 2019 Avista TRM UES value of 414 kWh. These participants
generally also replaced an existing gas furnace with a high-efficiency model (via the G Natural Gas
Furnace measure). The high electric energy savings appears to have resulted at least partly from a shift
in some homes away from secondary electric heating, such as portable electric heaters or electric wall
heaters, after installing the new gas furnace. Specifically, 66 of 159 participants in Idaho increased
natural gas usage after installing the high-efficiency furnace with variable speed fan motor, and they
sharply reduced electricity consumption. This pattern was not strong enough to suggest that the primary

23
heating system had changed from some other fuel to natural gas, but it did suggest that the high-
efficiency furnace prompted participants to move away from secondary heating with electricity.
Billing analysis provided relatively low electric energy savings for replacement windows relative to the
2019 TRM value of 15.25 kWh per square foot of window area, resulting in a realization rate of 72%. To
provide participant counts high enough to support statistically significant estimates, Cadmus combined
participants for the storm window and replacement window measures. Because billing analysis results
for only Idaho failed to meet the ±40% precision requirement, Cadmus based evaluated Idaho savings on
the combined results for Idaho and Washington participants. Note that in PY 2019, only one Idaho
project reported savings through the storm window measure, claiming savings for 150 square feet of
installed storm window, the realization rate for that measure has little impact on program savings.
Billing analysis for the MFDI program showed strong electric energy savings for the program as a whole.
As noted previously in
Residential Impact Evaluation Methodology, isolating the impact of individual measures was not
possible for MFDI because most living units received a range of LED light bulbs as well as water saving
measures such as aerators and showerheads. To provide an accurate estimate of the energy savings for
the program as a whole, Cadmus performed a complex and rigorous evaluation involving 1,549 living
units in 34 apartment buildings and complexes. The analysis yielded a realization rate of 96% for electric
energy savings in Idaho, with a relative precision of ±17% at a 90% confidence level. The billing analysis
did not evaluate savings from the MFDI Supplemental Lighting program.
Evaluated Savings
To calculate evaluated savings, Cadmus used unit savings values determined through billing analysis for
the measures shown in Table 18. For the remaining measures, Cadmus applied the results of database
review and, where applicable, document review to evaluate savings for each measure. The analysis then
rolled up measure-level evaluated savings to calculate evaluated savings and a realization rate for each
program. Table 19 shows the resulting evaluated savings and realization rates.
Table 19. Residential Electric Impact Findings
Program Reported Electric
Savings (kWh)
Evaluated Electric
Savings (kWh)a Realization Rates
Simple Steps, Smart Savings 3,879,137 3,879,137 100%
HVAC 659,957 1,335,085 202%
Shell 190,390 160,507 84%
ENERGY STAR Homes 66,262 69,615 105%
Multifamily Direct Install 1,289,539 1,258,897 98%
Multifamily Direct Install Supplemental Lighting 340,719 332,718 98%
Residential Total 6,426,003 7,035,960 109%
Note: totals may not sum due to rounding.

24
Residential Conclusions and Recommendations
Evaluated electricity savings show a realization rate of 109% on evaluated savings of 7,036 MWh for the
Residential programs, which is 156% of the savings goal for the year. The high percentage of achieved
savings relative to the goal results from program participation that was 134% of goal and the strong
overall realization rate for the Residential sector.
Lighting measures accounted for 73% of the total Residential sector savings. The following shows the
percentage of residential evaluated savings provided by each program:
• The Simple Steps, Smart Savings program provided 55% of Residential evaluated savings, mostly
through lighting measures.
• The MFDI and MFDI Supplemental Lighting programs provided 23% of evaluated savings, again
mostly through lighting measures.
• The Residential HVAC program accounted for 19% of evaluated savings.
• The Shell and ENERGY STAR Homes programs accounted for a combined 3% of residential
evaluated savings.
Realization rates varied by program from 84% for the Shell program to 202% for the HVAC program,
resulting in a strong overall realization rate of 109% for PY 2019. Cadmus identified few discrepancies
through document review, which found that the great majority of projects were well documented and
met program requirements.
Cadmus offers three recommendations regarding Avista’s Residential electric programs:
• Based on billing analysis conducted for this evaluation, adjust the Avista TRM to provide higher
savings values for variable-speed motors installed with the G Natural Gas Furnace measure and
lower savings for replacement windows in electrically heated homes. The billing analysis showed
savings for the variable-speed motor measure nearly four times the Avista TRM value on
average, seemingly due to a shift away from secondary electric heating (such as portable
heaters or wall heaters) in some homes after replacing a gas furnace with a high-efficiency
model. For replacement windows in electrically heated homes, the billing analysis estimated
unit savings of 72% the 2019 TRM value.
• The MFDI program has proven to be an efficient, effective mechanism for installing high-
efficiency lighting and aerators in multifamily units. Continue to focus on replacing high-use,
low-efficiency lamps where practical, to maximize program cost-effectiveness while maintaining
high savings.
• Ensure that reported savings for all measures are calculated using current TRM or RTF UES
values, and that the TRM provides values for all measures. Cadmus noted no large-scale
problems with the PY 2019 measure tracking data but did note numerous measure-tracking
records that reported zero savings, despite appearing to have been completed and a rebate
having been issued. In addition, some instances of PY 2019 measures used UES values from the
2018 TRM, and reported values for some measures (most notably, smart thermostats) did not
match TRM values.

25
Low-Income Impact Evaluation
Cadmus designed the Low-Income program impact evaluation to verify reported program participation
and energy savings. Evaluation methods included database review and billing analysis.
Program Summary
Avista leverages the infrastructure of a single Community Action Partnership agency to deliver energy
effiicency programs for the company’s low-income residential customers in the Idaho service territory.
The program is designed to serve Avista’s residential customers in Idaho whose income falls between
175% and 250% of federal poverty level. For PY 2019, the program achieved 234,102 kWh of reported
electric savings in Idaho, not including savings for the Low-Income Fuel Efficiency measures, which are
reported separately in the Fuel Efficiency Impact Evaluation section.
Program Participation Summary
Table 20 shows Avista savings goals for the Low-Income sector for PY 2019, as well as reported savings
and goal portions achieved in PY 2019.
Table 20. Low-Income Reported Savings
Program Savings Goals (kWh) Reported Savings (kWh)a Percentage of Goal
Low-Income 148,972 234,102 157%
a Reported savings do not include Low-Income Fuel Efficiency savings, shown in the Fuel Efficiency Impact Evaluation section.

Table 21 summarizes participation goals for the Low-Income programs, along with participation
reported and achieved in PY 2019.
Table 21. Low-Income Participationa
Program Participation Goals Participation Reported Percentage of Goal
Low-Income 24,834 39,758 160%
a Participation numbers do not include Low-Income Fuel Efficiency participation, shown in the Fuel Efficiency Impact
Evaluation section. Participation is defined as the number of installed units or square feet of installed insulation or windows.
Low-Income Impact Evaluation Methodology
Cadmus evaluated Low-Income program measures by conducting a database review (described in the
Database Review section) and billing analysis. The team used UES values provided in the TRM to
calculate savings for measures reported in the measure tracking database. Cadmus labeled savings
calculated during the database review as adjusted savings.
For many measures reported in the tracking database, notes indicated that savings were capped at 20%
of consumption. When duplicating savings calculations using TRM values, Cadmus used the newly
calculated value if it was less than the capped value, but used the capped value where the TRM value
indicated greater savings.

26
Cadmus conducted billing analysis for the Low-Income program using all electricity consumption data
available from Avista for PY 2018 and PY 2019 program participants. Because of the relatively small
number of Low-Income program participants, Cadmus was unable to isolate measure-level savings for
the program (which are necessary for cost effectiveness calculations). In addition, realization rates for
Idaho participants showed enough variation that billing analysis results did not meet the required
confidence and precision threshold, either for Idaho participants or for Idaho and Washington
participants combined.
Low-Income Impact Evaluation Results
Table 22 shows reported and adjusted electric savings for Low-Income conservation measures. The table
does not include savings for Low-Income programs Fuel Efficiency path measures (shown in the Low-
Income Fuel Efficiency Impact Findings section below).
Table 22. Low-Income Electric Impact Findings
Program Reported Electric Savings
(kWh)
Adjusted Electric Savings
(kWh)
Evaluated Electric
Savings (kWh)
Realization
Rate
Low-Income 234,102 232,126 232,126 99%

During the database and TRM review, Cadmus noted a number of errors or challenges with the measure
tracking data. For example, although the 2019 Avista TRM moved to providing a savings value per
square foot of living space for air sealing, many instances of air sealing in the tracking data used the
previous TRM value and did not include the area of the home. Some instances of some measures
reported low or high electric savings values. The errors largely offset one another at the program level,
as shown by the program’s 99% realization rate.
Low-Income Conclusions and Recommendations
With a realization rate of 99% for electricity savings, the Low-Income program achieved savings of
232,126 kWh in PY 2019, or 156% of goal. Reported program participation reached 160% of the
expected value. Roughly 26% of evaluated Low-Income program savings resulted from LED bulbs given
out at events.
For many instances of measures in the Low-Income tracking data, notes indicated that savings were
capped at 20% of consumption. The tracking data did not include adequate information to determine
when savings values are appropriately capped. Cadmus recommends providing annual consumption for
each measure in the tracking data, if practical, so that the evaluation can verify savings were capped at
20% of consumption for applicable measures.

27
Fuel Efficiency Impact Evaluation
Cadmus designed the Fuel Efficiency sector impact evaluation to verify reported program participation
and energy savings. Evaluation methods included a database review, document review, and billing
analysis.
Program Summary
Fuel Efficiency measures replace electric space heating or water heating systems with equipment using
natural gas. These measures are offered within the Nonresidential Site Specific path (which includes
HVAC Combined, refrigerator case doors, industrial process, and Multifamily Market Transformation
measures), Residential programs, and Low-Income programs. Across these programs, Avista reported
electric energy savings of 1,494,614 kWh for 161 Fuel Efficiency measures.
Fuel Efficiency measures provide positive electricity savings and negative natural gas savings, reflecting
negative avoided costs. Cadmus incorporated these negative avoided costs in the electric cost-
effectiveness calculations and reported the negative natural gas consumption impacts in the PY 2019
Idaho Natural Gas Impact Evaluation Report.
Program Participation Summary
This section summarizes Fuel Efficiency sector participation and progress toward PY 2019 goals for the
Nonresidential Site Specific path, Residential programs, and Low-Income programs.
Table 23 shows savings goals, reported savings, and percentage of goal for Nonresidential Site Specific,
Multifamily Market Transformation, Residential, and Low-Income Fuel Efficiency measures. Avista did
not set savings goals for the Site Specific Fuel Efficiency measures outside of the Multifamily Market
Transformation program.
Table 23. Avista Portfolio Fuel Efficiency Reported Electric Savings
Program Savings Goals (kWh) Reported Savings (kWh) Percentage of Goal
Multifamily Market Transformation 234,960 300,230 128%
Residential Fuel Efficiency 1,002,795 1,156,576 115%
Low-Income Fuel Efficiency 101,640 37,808 37%

Table 24 shows Avista’s PY 2019 participation goals and reported participation for Multifamily Market
Transformation, Residential, and Low-Income Fuel Efficiency measures. Avista did not set participation
goals for Site Specific Fuel Efficiency measures outside of the Multifamily Market Transformation
program. There were four Multifamily Market Transformation program participants and no
Nonresidential Site Specific participants in PY 2019.

28
Table 24. Avista Portfolio Fuel Efficiency Reported Participation
Fuel Efficiency Measure Participation Goalsa Participation Reporteda Percentage of Goal
Multifamily Market Transformation 40 4 10%
Residential Fuel Efficiency 141 143 101%
Low-Income Fuel Efficiency 30 13 43%
a Participation is defined as the number of rebates.
Fuel Efficiency Impact Evaluation Methodology
The impact methodology for Fuel Efficiency measures is outlined below for the Nonresidential Site
Specific path, Residential programs, and Low-Income programs.
Nonresidential Site Specific Fuel Efficiency Impact Methodology
Cadmus followed the same impact evaluation methodology for Fuel Efficiency measures as outlined in
the Nonresidential Impact Evaluation Methodology section. The team randomly sampled one
Multifamily Market Transformation program project in Washington for the evaluation of the
Nonresidential Fuel Efficiency measures. Cadmus did not evaluate the single Nonresidential Site Specific
Combined HVAC application in the Idaho Fuel Efficiency program; however, the team evaluated two
applications with the same measure category in the electric and gas Site Specific programs and found
realization rates of 100% on those two projects. Verification site visits involved verifying installed
equipment type, make and model numbers, operating schedules, and set points.
Residential Fuel Efficiency Impact Methodology
Cadmus applied billing analysis results to evaluate electric consumption impacts for all Residential Fuel
Efficiency measures using the methodology described previously in Billing Analysis. Cadmus also
completed database review of all PY 2019 reported Residential Fuel Efficiency savings as well as
document reviews for 50 Fuel Efficiency participants from Q1 PY 2018 through Q2 PY 2019.
Low-Income Fuel Efficiency Impact Methodology
To evaluate electric consumption impacts for the Low-Income Fuel Efficiency measures, Cadmus
conducted a database review (described above in the Database Review section) and billing analysis. The
relatively low number of participants for the Low-Income program made it impractical for the billing
analysis to isolate consumption impacts for specific measures. Using unit savings values provided in the
TRM, Cadmus calculated electric consumption impacts for measures reported in the measure-tracking
database. For Low-Income program measures in general (including Low-Income Fuel Efficiency
measures), the evaluation relied on results from the database review to determine evaluated electric
energy consumption impacts.
Fuel Efficiency Impact Evaluation Results
The following sections summarize findings for the Nonresidential Site Specific path, Residential
programs, and Low-Income programs Fuel Efficiency measures. All Fuel Efficiency measures provide
positive electricity savings and negative natural gas savings because these measures replace electric
space heating or water heating systems with equipment that uses natural gas. Negative savings,

29
reflecting negative avoided costs, are incorporated in the electric cost-effectiveness calculations. The
team also report these negative savings in the PY 2019 Idaho Natural Gas Impact Evaluation Report.
Nonresidential Fuel Efficiency Impact Findings
Table 25 shows reported and evaluated electric energy savings for Avista’s Nonresidential Fuel Efficiency
measures, along with realization rates, through PY 2019.
Table 25. Nonresidential Fuel Efficiency Electric Impact Findings
Fuel Efficiency Measure Reported Savings (kWh) Evaluated Savings (kWh) Realization Rate
Multifamily Market Transformation 300,230 300,230 100%
Total 300,230 300,230 100%

Cadmus did not identify any discrepancies affecting electric savings in the randomly sampled Multifamily
Market Transformation program application based on the evaluation site visit and project
documentation review.
Residential Fuel Efficiency Impact Findings
Table 26 shows measure-level billing analysis results used when calculating PY 2019 electric
consumption impacts. The participant count and relative precision for each measure easily met
requirements established to ensure meaningful results, which required a participant count of at least
five and relative precision no greater than ±40% at the 90% confidence level. The billing analysis found
the electric energy savings to be higher than predicted by the 2019 Avista TRM values for all but the wall
heater measure. Realization rates relative to 2019 TRM values ranged from 63% for the wall heater
measure to a high of 178%.
Table 26. Residential Fuel Efficiency Analysis Results
Measure
2019 Avista
TRM Unit
Energy
Savings
(kWh)
na
Pre-
Installation
Weather
Normalized
Usage (kWh)
Annual Unit
Energy
Savings
(kWh)
Realizatio
n Rate
Relative
Precision at
90%
Confidence
Participant
State
E Electric To Natural
Gas Furnace 6,104 39 19,054 7,384 121% 16% Idaho
E Electric To Natural
Gas Furnace &
Water Heat
8,513 35 19,284 9,789 115% 12% Idaho
E Electric To Natural
Gas Wall Heatera 10,624 9 17,597 6,745 63% 30% Idaho
E Multifamily Electric
to Natural Gas
Furnace and Water
Heat
4,566 21 12,259 8,133 178% 13%
Idaho and
Washingto
n
a The 2019 Avista TRM does not include the E Electric to Natural Gas Wall Heater measure. The TRM value shown is taken
from the 2018 Avista TRM.

30
Table 27 shows reported, adjusted, and evaluated electric energy savings for the Residential Fuel
Efficiency measures. Based on the measure-level billing analysis results listed in Table 26, Cadmus
calculated a 102% realization rate for evaluated electric energy savings for the Residential Fuel Efficiency
path.
Table 27. Residential Fuel Efficiency Electric Impact Findings
Fuel Efficiency Measure Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Evaluated Electric
Savings (kWh) Realization Rate
Residential Fuel Efficiency 1,156,576 1,010,460 1,181,596 102%

Database review of Residential Fuel Efficiency measures resulted in roughly a 12% reduction in adjusted
savings, primarily because reported savings in some instances used a higher UES value than the 2019
TRM value. Because billing analysis produced valid estimates for all Residential Fuel Efficiency measures,
adjusted savings had no effect on evaluated savings.
In reviewing documentation for 50 Residential Fuel Efficiency measures, Cadmus found issues with two
conversions to gas furnaces: documentation for each site indicated that the furnace replaced an oil-fired
heating system. The team eliminated electricity savings for the natural gas furnaces, given that the
replaced system did not use electric heating. These adjustments led to a document review realization
rate of 97%, but Cadmus did not apply document review results to estimate evaluated savings because
billing analysis produced valid estimates for all Residential Fuel Efficiency measures.
Low-Income Fuel Efficiency Impact Findings
Table 28 shows reported and adjusted electric energy savings for Low-Income Fuel Efficiency measures.
Table 28. Low-Income Fuel Efficiency Program Electric Impact Findings
Fuel Efficiency Measure Reported Electric
Savings (kWh)
Adjusted Electric
Savings (kWh)
Evaluated Electric
Savings (kWh)
Realization
Rate
Low-Income Fuel Efficiency 37,808 37,808 37,808 100%

Cadmus found no discrepancies between reported and TRM UES values for electric energy savings with
Low-Income Fuel Efficiency measures, leading to a realization rate of 100% for electric energy savings.
The billing analysis estimated a realization rate of 144% for Low-Income Fuel Efficiency electric savings,
with a relative precision of ±27% at the 90% confidence level. Participation was not high enough to
estimate savings at the measure level, which is necessary for calculating cost-effectiveness, but the
results do indicate greater electric savings for Low-Income Fuel Efficiency measures as a whole than
indicated by 2019 Avista TRM values. This finding also supports the natural gas billing analysis finding
that the natural gas penalties for Low-Income Fuel Efficiency measures are much higher than estimated
by the 2019 Avista TRM (see PY 2019 Idaho Natural Gas Impact Evaluation Report). Together, the
electric and natural gas billing analysis results suggest a much greater heating load than indicated by
TRM values, which is evident as the heating load shifts from electricity to natural gas.

31
Fuel Efficiency Conclusions and Recommendations
Nonresidential Site Specific and Multifamily Market Transformation Fuel Efficiency measures achieved
evaluated savings of 300,230 kWh, yielding a 100% realization rate. The Multifamily Market
Transformation Fuel Efficiency measures achieved 128% of the electric energy savings goal of
234,960 kWh.
Residential Fuel Efficiency measures achieved evaluated savings of 1,181,596 kWh, yielding a 102%
realization rate and achieving 118% of savings goal. Cadmus recommends that Avista update TRM values
to match measure-level UES values calculated by the billing analysis. Cadmus also recommends that
Avista ensure all measures are represented in the TRM.
For Low-Income Fuel Efficiency measures, evaluated savings were 37,808, with a realization rate of
100%, but fell short of Avista’s savings goals, achieving 37% of the savings target and 43% of the
participation target. Billing analysis indicated that program electric savings are likely higher, based on
the billing analysis realization rate of 144% for Low-Income Fuel Efficiency measures as a whole. Based
on this finding, Cadmus recommends increasing the Avista TRM UES values.