20250530Direct Amen.pdf

RECEIVED
May 30, 2025
Preston N. Carter, ISB No. 8462 IDAHO PUBLIC
Megann E. Meier, ISB No. 11948 UTILITIES COMMISSION
GIVENS PURSLEY LLP
601 West Bannock Street
P.O. Box 2720
Boise, Idaho 83701-2720
Office: [phone]
Fax: [phone]
[e-mail]
[e-mail]
Attorneys for Intermountain Gas Company
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
IN THE MATTER OF THE APPLICATION Case No. INT-G-25-02
OF INTERMOUNTAIN GAS COMPANY
FOR THE AUTHORITY TO INCREASE
ITS RATES AND CHARGES FOR GAS
SERVICE IN THE STATE OF IDAHO
DIRECT TESTIMONY OF RONALD J.AMEN
INTERMOUNTAIN GAS COMPANY
MAY 30,2025

TABLE OF CONTENTS
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION................................I
I. Introduction...........................................................................................................2
II. Statement of Qualifications..................................................................................2
III. Purpose of Testimony............................................................................................3
IV. Load Study and Analysis......................................................................................4
V. Theoretical Principles of Cost Allocation............................................................20
VI. Intermountain's Cost of Service Study...............................................................28
VII. Principles of Sound Rate Design..........................................................................40
VIII. Determination of Proposed Class Revenues .......................................................41
IX. Intermountain's Rate Design ...............................................................................45
X. Concluding Remarks.............................................................................................48
PAGE 1 OF 50
R. AMEN,DI
INTERMOUNTAIN GAS COMPANY

I. INTRODUCTION
1 Q. Please state your name and business address.
2 A. My name is Ronald J. Amen and my business address is 10 Hospital Center Commons,
3 Suite 400, Hilton Head Island, SC 29926.
4 Q. By whom are you employed and in what capacity?
5 A. I am employed by Atrium Economics, LLC ("Atrium") as a Managing Partner.
6 Q. On whose behalf are you testifying?
7 A. I am testifying on behalf of Intermountain Gas Company ("Intermountain" or
8 "Company").
II. STATEMENT OF QUALIFICATIONS
9 Q. What has been the nature of your work in the energy utility consulting field?
10 A. I have over 40 years of experience in the utility industry, the last 27 years of which have
11 been in the field of utility management and economic consulting. I have advised and
12 assisted utility management, industry trade organizations, and large energy users in matters
13 pertaining to costing and pricing; competitive market analysis; regulatory planning and
14 policy development; resource planning and acquisition; strategic business planning; merger
15 and acquisition analysis; organizational restructuring; new product and service
16 development; and load research studies. I have prepared and presented expert testimony
17 before utility regulatory bodies across North America and have spoken on utility industry
18 issues and activities dealing with the pricing and marketing of gas utility services, gas and
19 electric resource planning and evaluation, and utility infrastructure replacement. Further
PAGE 2 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I background information summarizing my work experience, presentation of expert
2 testimony, and other industry-related activities is included as Direct Exhibit 34 to my
3 testimony.
4 Q. Have you previously testified before the Idaho Public Utilities Commission?
5 A. Yes. I have provided expert witness testimony before the Idaho Public Utilities
6 Commission("IPUC" or the "Commission") in Intermountain' s general rate case
7 proceeding, Case No. INT-G-22-07.
III. PURPOSE OF TESTIMONY
8 Q. Please summarize your testimony.
9 A. First, I will present the load study analysis for purposes of determining each customer
10 class's contribution to the system's peak load. Next, I present the development of the
11 Company's allocated Cost of Service Study ("COSS") for the test year ended December
12 31, 2024, with pro forma adjustments for 2025, including a comprehensive overview of the
13 schedules created in support of them. Finally, I present the Company's proposed rates and
14 the resulting customer bill impacts based on the Company's requested revenue increase.
15 My testimony consists of the following topics:
16 • Load Study and Analysis
17 • Theoretical Principles of Cost Allocation
18 • Intermountain's COSS
19 • A Summary of the COSS Results by Rate Class
20 • Determination of Proposed Class Revenues
21 • Rate Design
PAGE 3 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

1 • Customer Bill Impacts
2 Q. Are you sponsoring any exhibits to your direct testimony?
3 A. Yes. I am sponsoring the following 5 Exhibits, all of which were prepared by me or under
4 my supervision and direction.:
5 Exhibit 34—Resume of Ronald J. Amen
6 Exhibit 35 —Cost of Service Study
7 Exhibit 36—Proposed Revenue Targets
8 Exhibit 37—Proposed Rate Design and Proof of Revenue
9 Exhibit 38—Customer Bill Impacts
IV. LOAD STUDY AND ANALYSIS
10 Q. What is a load study?
11 A. A load study determines each customer class's contribution to the natural gas utility's
12 pipeline system peak load. This information is used to develop allocators for purposes of
13 allocating shared costs, or costs that cannot be directly assigned, such as plant and
14 equipment, operation, and maintenance expenses ("O&M"), and some administrative costs
15 to each customer class on the basis of peak day usage. Natural gas pipeline systems are
16 designed and constructed to satisfy peak day demand under design weather conditions and
17 a load study identifies each class's relative contribution to the peak day demand.
18 Q. Did Intermountain develop a load study in its previous general rate case proceeding,
19 No. INT-G-22-07 ("2022 Case")?
20 A. Yes. The Company provided a Load Study in its 2022 Case. Prior to 2022 (in its 2016 rate
21 case), the Company had lacked adequate data to perform a detailed load study, and as a
PAGE 4 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I result was encouraged by the Commission to participate with Staff and other interested
2 parties to determine the best way forward as it relates to class cost-of-service and the
3 acquisition of appropriate cost causation and load data.' The Company provided a Load
4 Study in its succeeding 2022 Case, which Commission staff listed as a key factor in
5 obtaining its support for the Company's class allocation of the revenue requirement.2
6 Q. Does Intermountain have sufficient daily AMI data to develop a load study in this
7 filing?
8 A. Yes. The Company has dramatically expanded its daily metering capability through
9 Advanced Metering Infrastructure ("AMI"). Table 1 below shows the percentage of daily
10 metered customers for each rate class for each of Intermountain's seven distinct weather
11 zones.
Table 1 -Percent of Premises with Daily Meter Readings
Canyon Boise Sun Twin Rexburg Idaho Pocatello
Countv Valle Falls Falls
Residential 99.8% 96.6% 47.5% 72.7% 74.3% 55.7% 91.6% 88.1%
Commercial 98.9% 98.5% 53.1% 70.1% 74.3% 69.8% 89.9% 86.9%
LV-1 54.5% 54.5% 0.0% 40.0% 100.0% 66.7% 0.0% 55 33%
T-3 0.0% 0.0% 0.0% 33.3% 0.0% 75.0% 100.0% 55.6%
T-4 50.0% 45.0% 0.0% 59.5% 66.7% 63.6% 90.0% 56.7%
IS-R 0.0% 66.7% 15.1% 75.0% 0.0% 33.3% 83.3% 17.2%
IS-C 0.0% 100.0% 30.2% 62.5% 100.0% 64.7% 75.0% 52.4%
IRR 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
TOTAL 99.7% 96.7% 47.2% 72.3% 74.3% 57.2% 91.4%
12
1 IPUC Order No.33757,Case No.INT-G-16-02(April 28,2017)at 28-29.
2 IPUC Order No.35836,Case No.INT-G-22-07(June 30,2023)at 4.
PAGE 5 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. Please describe the characteristics of Intermountain's gas load.
2 A. Intermountain serves customers throughout a geographically and economically diverse
3 service territory. There are seven primary rate classes: Residential ("RS") Commercial
4 ("GS"), Large Volume ("LV-1"), Transport ("74"), Interruptible Transport ("T-3"),
5 Interruptible Snowmelt-Residential ("IS-R"), and Interruptible Snowmelt-Commercial
6 ("IS-C"). Intermountain's customers are spread across seven diverse geographic areas with
7 differing weather patterns and elevations (Canyon County, Boise, Hailey(or Sun Valley),
8 Twin Falls, Rexburg, Idaho Falls, and Pocatello). Below is a chart showing total monthly
9 consumption for each rate class for the twelve months ended February 2025.
Figure 1 —Intermountain Monthly Consumption by Rate Class
60.",OW
w,000.000
E
ty 40,00g000
c
0
E 30,04000
"c
0
10000,000
0
f
10,0A000
Mar-14 Ap,24 May-24 lur.-14 lul-24 A I-14 5eµ14 Ott-24 Nw-24 Dx-24 J -25 sebb25
—Resldential(RS) —General Semce NIS) —large V01—ILVq —transportation Onterrupdblel(7-31
10 —Transponation 0.)V 4) —Sw melt-Re%&mial ll5n) l6C) —Wwtion ilRRI
11 Intermountain's Residential and Commercial customers are highly weather sensitive and
12 are spread across all seven weather zones. The Company's Large Volume ("LV")
13 customers are comprised of a mix of industrial and commercial loads and use in excess of
14 200,000 therms per year. These customers may be subject to one of three rate classes: a
PAGE 6 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I bundled sales tariff(LV-1), a distribution system only transportation tariff(T-4), and an
2 interruptible transportation tariff(T-3). The LV customers, on average, account for roughly
3 47 percent of Intermountain's annual throughput and approximately 26 percent of the
4 projected design peak day. The vast majority of the LV throughput reflects distribution
5 system-only transportation; and as a whole, the LV gas usage pattern is less weather
6 sensitive than the residential and commercial classes. The Company also has Residential
7 Interruptible Snowmelt Customers, which are separately metered from the premises and
8 are fully interruptible with at least two hours of notice. Similarly, there are Small
9 Commercial Interruptible Snowmelt Service customers that are also interruptible with two
10 hours of notice. Lastly, the Company has Irrigation Customers, which do not contribute to
I I the winter peak and do not factor into the load study.
12 Table 2 below provides a summary of premises and annual consumption projected
13 for the test year ended 2024 as a percentage of Intermountain's whole system throughput.
PAGE 7 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

Table 2 —Test Year Premises and Consumption Data for Intermountain's Gas
System3
PremisesConsumption Consumption
� Premises (Therms)
Residential 402,618 91.45% 308,569,141 35.98%
General Service 37,063 8.42% 148,190,583 17.28%
Large Volume(LV-1) 38 0.01% 15,440,184 1.80%
Transportation(Interruptible)(T-3) 9 0.00% 38,382,448 4.48%
Transportation(Firm)(T-4) 105 0.02% 345,702,053 40.31%
Snowmelt-Residential(IS-R) 343 0.08% 681,467 0.08%
Snowmelt-Commercial(IS-C) 84 0.02% 506,529 0.06%
Irrigation(IRR) 6 0.00% 45,228 0.01%
TOTAL 440,266 857,517,633
1 Q. How does the Company define its design day?
2 A. The Company's design day represents the coldest temperatures that can be expected to
3 occur during an extreme cold or peak weather event. Intermountain used a statistical model
4 to develop probability-derived peak heating-degree-day("HDD") values to characterize its
5 design day, corresponding to an exceedance probability that Intermountain considers
6 appropriate for its intended use. Intermountain used exceedance probability results to
7 review data associated with both a 50-year and 100-year probability weather event, as
8 shown below in Table 3. The Company's practice has been to rely on a 50-year probability
9 event, which results in a 78 heating-degree-day, for use in the design weather model.
3 Based on average monthly customers and total therms for the calendar year ended December 2024,with additional
system growth projected in test year projections.
PAGE 8 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

Table 3 —Peak Day HDD65 Event by Region
Canyon itSun Twin _I Idaho
PocatelloCompany
CountyFalls
50-Year Event 78 75 82 77 88 87 82 78.45
100-Year Event 81 79 85 80 91 89 85 81.76
Max Degree 83 81 88 80 92 88 83 82.90
Da s4
1
2 Q. What is the purpose of the Peak Load Study?
3 A. The purpose of the Peak Load Study is to project the design day peak, i.e., the 50-year
4 event using the results of linear regression equations or another reasonable estimate of
5 peak load by rate class.
6 Q. Please describe the methodology and approach for developing the Peak Load Study.
7 A. The development of the Peak Load Study began with the receipt of daily AMI data for
8 Intermountain's Residential (RS 1 and RS2) and Commercial (G10 and G11) customer
9 classes for the period, January 1, 2022 through February 28, 2025. At Atrium's request,
10 Intermountain identified and removed daily AMI data from the dataset that showed zero
11 usage when temperatures were less than 45 degrees (or greater than 20 HDD65). This
12 resulted in the removal of 1,269,993 daily Residential readings and 906,912 daily
13 Commercial readings with zero usage.
14 The daily AMI reads were measured in units of one hundred cubic feet, or"Ccf',
15 so it was necessary to apply a billing adjustment factor to the daily AMI data to account for
16 the heating value and pressure to arrive at delivered therms. This billing adjustment factor
4 Max Degree Days reflect the coldest day on record.
PAGE 9 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I was determined by comparing daily AMI data to monthly billing data, the latter of which is
2 based on billing cycles and not a calendar month. Though not exact, Atrium expects that
3 Intermountain's gas system is sufficiently stable in pressure and Btu content on contiguous
4 days, such that the billing factor adjustment provides a reasonable approximation of the
5 Ccf to Therms conversion.
6 Atrium then performed regression analyses to measure the historical linear
7 relationship between metered daily volumes/per customer, weather, and day-of-week
8 variables for each customer class and weather zone for the period from January 1, 2022, to
9 February 28, 2025. The following seven independent variables were regressed against
10 average daily use per customer for each customer class and weather zone combination.
11 • Heating degree days (using 65 degrees as the baseline, or HDD65)—adjusted for
12 wind
13 • Heating degree days (using 55 degrees as the baseline, or HDD55)—adjusted for
14 wind
15 • Change in heating degree days from prior day
16 • Friday
17 • Saturday
18 • Sunday
19 • Holiday
20 Q. Please explain the rationale for including the seven independent variables in the daily
21 regression equations.
22 A. The seven variables account for the linear relationship between weather and gas usage but
23 also account for non-linear or behavioral aspects of gas usage. The HDD65 variable is the
24 baseline weather variable. By introducing a HDD55 variable, we are measuring how
25 customer behavior(and heating system usage) changes when it is already cold versus when
PAGE 10 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I it is marginally cold. In our analysis,we found that the HDD55 coefficient could be higher
2 or lower than the HDD65 variable, tending to be lower in more densely populated regions
3 suggesting that heating systems may retain heat and/or run more efficiently in conditions
4 of sustained cold; and tending to be higher in the less densely populated regions,
5 suggesting that the heat usage between 65 and 55, was less predictable, i.e., customers may
6 not use heat when temperatures are in this range. In all cases, these two weather variables
7 were highly significant (by review of the associated t-statistic5). The AHDD variable
8 measures the difference between the current day HDD65 and the prior day. This variable
9 captures customer response to weather volatility. In our analysis the regression coefficient
10 was negative for the RS and GS rate classes and weather zones, and highly significant in
11 all cases (by review of the t-statistic), suggesting that customer response to sudden changes
12 in HDD is lagged. Atrium also included variables for weekends and holidays to capture the
13 change in customer behavior, attributable to the day of the week. As one would expect, this
14 series of variables were much more significant for the Commercial class of customers than
15 Residential,but the variables were retained in the Residential regressions since they
16 strengthened the overall regression result.
17 Q. Please describe the adjustment that Atrium performed on HDD to account for wind.
18 A. Atrium adjusted HDD by applying the following standard formula to each daily HDD
19 calculation: (too+Averagee0Wind Speed l Because wind data was not available for every
ic
s A t-statistic in excess of 1.96 indicates that the regression coefficient is statistically significant at the 95%
confidence level.A t-statistic in excess of 1.645 indicates that the regression coefficient is statistically significant at
the 90%confidence level.
PAGE 11 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I weather zone, wind data from the nearest weather zones where wind data was available
2 was substituted. Boise wind data was substituted for Canyon County weather zone 350;
3 Twin Falls wind data for Hailey (Sun Valley) weather zone 500; and Idaho Falls data for
4 Rexburg weather zone 700.
5 Q. Please describe the regression analyses using daily AMI metered data for the
6 residential and commercial customer classes and the development of the "Daily" peak
7 load sendout model.
8 A. As indicated in Table I above, there is significant penetration of daily AMI meters for the
9 residential and commercial classes, 88 percent, and 87 percent, respectively. However, two
10 mountainous weather zones, 500 Sun Valley (or Hailey) and Idaho Falls (750)had lower
11 penetration of AMI metering, with Hailey(47.5% RS and 53.1% GS) and Idaho Falls at
12 (55.7%RS and 69.8% GS). In general, the more mountainous and remote areas of service
13 territory have a lower penetration of AMI, which does lead to daily AMI data slightly
14 under-predicting weather sensitivity in those regions. The daily regression results were
15 relied upon to project design day load for the residential and commercial classes, or
16 "Core"6 customer classes. For the large volume classes, either due to lack of weather
17 sensitivity or due to lack of data, other means of estimating peak day results were used.
18 The results of the daily regressions are listed below in Table 4.
6 Core customers are defined on page 168 of Intermountain's 2023-2028 IRP as,"All residential and commercial
customers of Intermountain Gas Company. Includes all customers receiving service under the RS and GS tariffs."
PAGE 12 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

Table 4-Daily Regression Results (January 2022 -February 2025)
Residential Class
R2 0.975 0.977 0.944 0.955 0.975 0.954 0.965
x1(HDD65)coefficient 0.054 0.076 0.047 0.048 0.033 0.056 0.033
xt-stat 22.708 31.550 10.039 11.054 18.072 13.701 9.956
x std.error 0.002 0.002 0.005 0.004 0.002 0.004 0.003
x2(HDD55)coefficient 0.068 0.052 0.067 0.054 0.035 0.032 0.059
x t-stat 21.360 16.051 12.284 9.883 16.155 6.488 14.560
x std.error 0.003 0.003 0.005 0.005 0.002 0.005 0.004
x3(OHDD)coefficient (0.018) (0.015) (0.019) (0.016) (0.008) (0.012) (0.010)
x t-stat (9.504) (8.219) (6.645) (6.149) (6.585) (4.007) (4.246)
x std.error 0.002 0.002 0.003 0.003 0.001 0.003 0.002
x4(Fri)coefficient (0.075) (0.040) (0.037) (0.036) (0.060) (0.046) (0.024)
x t-stat (3.364) (1.737) (0.864) (0.898) (3.492) (1.134) (0.773)
x std.error 0.022 0.023 0.042 0.040 0.017 0.040 0.031
x5(Sat)coefficient (0.026) 0.006 0.020 0.009 (0.008) (0.034) (0.007)
x t-stat (1.190) 0.269 0.468 0.224 (0.462) (0.837) (0.237)
x std.error 0.022 0.023 0.042 0.040 0.017 0.040 0.032
x6(Sun)coefficient 0.050 0.067 0.084 0.038 0.021 0.024 0.060
x t-stat 2.261 2.919 1.992 0.942 1.207 0.584 1.899
x std.error 0.022 0.023 0.042 0.040 0.017 0.040 0.032
x7(Holiday)coefficient (0.083) 0.004 0.141 (0.008) (0.079) (0.075) 0.020
x t-stat (1.391) 0.067 1.259 (0.080) (1.715) (0.702) 0.247
x std.error 0.059 0.061 0.112 0.105 0.046 0.106 0.081
y(Constant)coefficient 0.446 0.521 0.593 0.391 0.537 0.440 0.392
yt-stat 30.399 35.415 17.654 12.922 42.219 15.553 17.806
y std.error 0.015 0.015 0.034 0.030 0.013 0.028 0.022
Commercial Class
R2 0.973 0.977 0.963 0.938 0.978 0.967 0.965
x1(HDD65)coefficient 0.179 0.264 0.034 0.201 0.119 0.193 0.108
x t-stat 15.112 24.387 3.757 8.565 12.392 12.098 6.785
x std.error 0.012 0.011 0.009 0.023 0.010 0.016 0.016
x2(HDD55)coefficient 0.433 0.326 0.253 0.266 0.268 0.232 0.347
xt-stat 27.667 22.649 23.926 9.120 23.736 11.998 17.646
x std.error 0.016 0.014 0.011 0.029 0.011 0.019 0.020
x3(OHDD)coefficient (0.109) (0.073) (0.054) (0.081) (0.052) (0.066) (0.079)
x t-stat (11.607) (8.661) (9.974) (5.825) (8.629) (5.832) (7.206)
x std.error 0.009 0.008 0.005 0.014 0.006 0.011 0.011
x4(Fri)coefficient (0.579) (0.118) (0.034) (0.407) (0.271) (0.213) (0.096)
x t-stat (5.250) (1.155) (0.412) (1.910) (2.990) (1.363) (0.631)
x std.error 0.110 0.102 0.082 0.213 0.091 0.156 0.152
x5(Sat)coefficient (1.239) (0.505) (0.169) (1.030) (0.464) (0.614) (0.458)
x t-stat (11.259) (4.946) (2.054) (4.807) (5.122) (3.927) (2.994)
x std.error 0.110 0.102 0.082 0.214 0.091 0.156 0.153
x6(Sun)coefficient (1.015) (0.474) (0.255) (1.091) (0.703) (0.686) (0.452)
x t-stat (9.233) (4.650) (3.104) (5.091) (7.771) (4.385) (2.947)
x std.error 0.110 0.102 0.082 0.214 0.090 0.157 0.153
x7(Holiday)coefficient (1.621) (0.791) (0.191) (1.270) (0.803) (1.031) (0.423)
x t-stat (5.518) (2.902) (0.874) (2.245) (3.333) (2.494) (1.079)
x std.error 0.294 0.272 0.218 0.566 0.241 0.413 0.392
y(Constant)coefficient 2.754 2.819 1.416 3.083 2.546 2.205 2.103
yt-stat 37.971 42.881 21.670 18.948 38.027 20.073 19.731
1 y std.error 0.073 0.066 0.065 0.163 0.067 0.110 0.107
PAGE 13 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. Did you also perform regression analyses using monthly billing data for the
2 residential and commercial customer classes?
3 A. Yes. Monthly data regressions were performed on Intermountain's monthly billing data.
4 This data had the advantage of covering all customers within the class and weather zone,
5 and was already expressed in therms, so no adjustments to the data were necessary. In the
6 monthly data regressions, average daily HDD (computed based on 65 degrees)was
7 regressed against average daily use per customer for the month, for each class and weather
8 zone. The results of the monthly data regressions for the residential and commercial classes
9 are reported in Table 5. These results are referred to as the "Monthly Model."
Table 5-Monthly Regression Results (January 2022-February 2025)
Residential Class
R2 0.986 0.992 0.972 0.987 0.992 0.991 0.986
x coefficient 0.115 0.126 0.162 0.101 0.070 0.086 0.092
xt-stat 50.497 67.007 35.652 52.222 66.826 63.935 50.528
x std.error 0.002 0.002 0.005 0.002 0.001 0.001 0.002
y coefficient 0.261 0.402 0.326 0.209 0.385 0.368 0.176
y t-stat 5.575 10.637 2.546 4.639 12.791 9.739 3.742
y std.error 0.047 0.038 0.128 0.045 0.030 0.038 0.047
Commercial Class
R2 0.945 0.985 0.948 0.979 0.989 0.988 0.979
x coefficient 0.560 0.603 0.321 0.539 0.418 0.467 0.478
xt-stat 24.924 48.818 25.638 40.490 57.404 53.741 40.791
x std.error 0.022 0.012 0.013 0.013 0.007 0.009 0.012
y coefficient 3.072 2.629 1.915 2.454 2.057 1.576 1.133
y t-stat 6.666 10.586 5.464 7.966 10.029 6.464 3.740
10 y std.error 0.461 0.248 0.350 0.308 0.205 0.244 0.303
11 Q. Was there a validation step performed to check the accuracy of the "Daily" or the
12 "Monthly" peak load sendout models in predicting the winter peak load?
13 A. Yes. To check the appropriateness of the modeling results, "Daily" and"Monthly"peak
14 load sendout models were validated by comparing each to actual historical sendout,using
15 actual historical HDD by weather zone and the class/weather zone regressions for the
PAGE 14 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I period January 1, 2022, to February 28, 2025. The results of that comparison are illustrated
2 below.
Figure 2 —Blended and Monthly Models versus Company Core Sendout
Core Sendout vs Regression Results
450.000
4W,0W
3S0,0W
300.000
150,oW
LW,WU
150,OW
100,000
S0,000
"s � � "s � "so "s � o � ossgsssgsgo "' � `sosgg � `ss � o � � � "s
5 ;Z5 5 a aa > a a5 Z� 'a as > a aM aaa ;Z� a as a a
3 —SENOOUT —DAI/YREGRESSIONS —MONTHLY REGRESSIONS
4 As illustrated in Figure 2 above, the peak use during the illustrated period occurred on
5 January 15, 2024, with an average HDD across all weather zones of 56.94, and core market
6 load of 389,921 MMBtu. This HDD was slightly lower than the coldest day of the period,
7 January 14, 2024, at 5 7.10 HDD,but since January 141h was a Sunday, the sendout was
8 lower than on January 151h, even though the HDD was higher. As Figure 2 shows, the
9 Monthly data predicts a slightly higher peak than the Daily data,but only slightly so. This
10 minor disparity could be explained by the fact that daily data contained a lesser proportion
I I of daily readings for the more mountainous and colder weather regions of Intermountain's
PAGE 15 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I service territory, i.e., there are likely large, highly weather-sensitive customers in cold
2 regions that lacked data for the entire period, either due to meters being added during the
3 period or readings being removed for zero usage when HDD was greater than or equal to
4 20 HDD.
5 Q. Did you validate your daily modeling results against the daily data received from
6 Intermountain?
7 A. Yes. I also compared the daily regression results against the daily data I used to develop
8 the daily regression models to verify that the regression models were properly reflecting
9 the data used to develop them. Figure 3 below, shows the results of that analysis.
Figure 3—Daily Regressions vs. Daily AMI data
Core Sendout vs Regression Results
400,000
350,000
300,000
250,000 I� I
200,000
150,000
100,000
50,000 -
N N ry N N M M N M HI � O O Y O O O O O O Jf Uf
O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
a a W a a a ; 1 W m a S - < a - - - - S r
10 -DAILY REGRESSIONS -DAILY UPC"TRAPCRATIONS
PAGE 16 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I As Figure 3 shows, the daily regressions revealed greater weather sensitivity than was
2 reflected in the actual data. This could be due to the averaging of use per customer data
3 across all customers in the weather zone, in the daily data. The daily regressions were more
4 closely aligned with the monthly billing data, and sendout data, and are better predictors of
5 design day peak weather than the raw daily data itself. Atrium has used Intermountain's
6 daily AMI data for its calculation of Intermountain's design day peak.
7 Q. What were the results of the Daily Peak Load Sendout Model for Intermountain's
8 Core Residential and Commercial Customers?
9 A. The daily regression results were extrapolated to the average test year number of customers
10 for each weather zone for each of the Core classes, RS and GS. Because the test year
11 customer counts were slightly higher(due to adjustments for expected customer growth)
12 than the customer counts used for the analysis, I extrapolated peak load therms to the
13 average test year customers by multiplying the therms calculated in the analysis by the
14 ratio of test year customers over the number of customers used in the analysis by weather
15 zone. The results are shown in Table 6 below.
Table 6—Peak Load Sendout for Core Customers
Core Rate Class Customers7Peak Load
�-211
Residential 402,618 3,598,775
Commercial 37,063 1,490,975
Total Core Customers 439,681 5,089,750
16
7 Based on average monthly customers projected for the test year(January 2024—December 2024). Totals exclude
interruptible snowmelt classes,CNG,and Irrigation.
PAGE 17 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. How did you estimate the peak day sendout for the LV rate classes?
2 A. Because the LV customers are not as weather sensitive as the residential and commercial
3 customers, forecasting their volumes using standard regression techniques based on
4 projected weather may not provide statistically significant results. As such, the maximum
5 contract demand was used for these large volume customers to project loads at peak. For
6 the LV-1 class and the T-4 class, the maximum daily firm quantity ("MDFQ") as of
7 December 2024 was used. The MDFQ reflects the maximum amount of daily gas and/or
8 capacity Intermountain must be prepared to provide to its firm LV customers on any given
9 day, including the projected system peak day. These amounts represent a contracted daily
10 requirement and reflects the known peak day obligation for each customer. The December
11 2024 MDFQ amounts were 1,526,530 therms for the T-4 class, and 82,005 therms for the
12 LV-1 class. It is reasonable to expect that on a peak day these customers will be using their
13 full contracted MDFQ. I note that this treatment is consistent with how the Peak Day
14 Sendout was developed in the 2023 IRP.B
15 The daily peak sendout for the Interruptible Transport Class, T-3,was determined
16 based on the test year average daily load for the twelve months ending December 2024. T-
17 3 customers are interruptible and as such there are no assurances of the amount of capacity
18 that they may be granted on any given day. However, given that Intermountain has rarely
19 interrupted these customers, it is reasonable to provide a peak day allocation for their
20 contribution to the system peak. Peak day sendout results have been provided with and
8 Intermountain Gas Company,Integrated Resource Plan 2023-2028,at pp.33-34.
PAGE 18 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I without the interruptible customers; and note that interruptible customers have previously
2 been excluded from Intermountain's peak load analyses. The average daily usage for the T-
3 3 customers was 104,870 therms for the test year twelve-month period ending December
4 31, 2024.
5 Q. Was the peak day sendout estimated for the Interruptible Snowmelt Classes?
6 A. Yes. The peak day sendout for the Interruptible Snowmelt classes (IS-R and IS-C)were
7 estimated based on their average daily use for the month of February 2025. These classes
8 are also fully interruptible with two hours of notice and could not be assured of capacity
9 during any given peak day. However, as the Company has rarely interrupted these
10 customers, they have been included in the Peak Load Study for reference.
I I Q. Please provide the results for Intermountain's total peak day sendout.
12 A. The results of the peak load study and the resulting allocations with and without the
13 inclusion of interruptible customers were prepared and summarized in Table 7 below.
Table 7—Peak Day Sendout with and without Interruptible Classes—Daily Model
I 1Model
Firm & Interruptible Firm Only
Rate Class: Therms % Therms %
Residential RS 3,598,775 52.8% 3,598,775 53.7%
General Service GS 1,490,975 21.9% 1,490,975 22.3%
Large Volume LV-1 82,005 1.2% 82,005 1.2%
Transportation(Interruptible) (T-3) 104,870 1.5% - 0.0%
Transportation Firm T-4 1,526,530 22.4% 1,526,530 22.8%
Snowmelt-Residential IS-R 6,821 0.1% - 0%
Snowmelt-Commercial IS-C 5,601 1 0.1% - 0%
TOTAL 6,815,578 1 1 6,698,285
14 For purposes of this allocated class cost of service study, the results shown in Table 7 were
15 selected, which use the Daily peak load regression model to determine the Core peak day
PAGE 19 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I sendout since this data reflects actual daily use per customer and provides robust results in
2 predicting peak day sendout as illustrated above in Figures 2 and 3. These results are
3 aligned with Intermountain's projections of peak day sendout for firm demand(RS, GS,
4 LV-1, and T-4) in their 2023-2028 IRP, which projected 643,154 MMBtu for 2024,
5 654,749 MMBtu for 2025, and 666,639 MMBtu for 2026.9 Because Intermountain is
6 incorporating some future growth estimates into its test year customer count, it is
7 appropriate to compare to the 2025 peak demand estimate. The difference in Atrium's
8 Design Day projection differs from Intermountain's estimate of design day load by 3,190
9 MMBtu for 2025. In addition, the contract demands utilized in the Company's IRP were
10 150,254 MMBtu for LV-1 and T-4 combined, whereas by the year end 2024 (the data used
11 for Atrium's load study), those daily contract demands (MDFQ) had increased to 160,854
12 MMBtu—a difference of 10,600 MMBtu, which more than accounts for the difference
13 between Intermountain's 2025 Design Day Demand calculation and that calculated by
14 Atrium.
V. THEORETICAL PRINCIPLES OF COST ALLOCATION
15 Q. Why do utilities conduct cost allocation studies as part of the regulatory process?
16 A. There are many purposes for utilities to conduct cost allocation studies, ranging from
17 designing appropriate price signals in rates to determining the share of costs or revenue
18 requirements borne by the utility's various rate or customer classes. In this case, an
19 embedded COSS is a useful tool for determining the allocation of Intermountain's revenue
9 Intermountain Gas Company,Integrated Resource Plan 2023-2028,Table 29,p. 122
PAGE 20 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I requirement among its customer classes. It is also a useful tool for rate design because it
2 can identify the important cost drivers associated with serving customers and satisfying
3 their design day demands.
4 Embedded cost studies analyze the costs for a test period based on either the book
5 value of accounting costs (a historical period) or the estimated book value of costs for a
6 forecasted test year or some combination of historical and future costs. Typically,
7 embedded cost studies are used to allocate the revenue requirement between jurisdictions,
8 classes, and between customers within a class.
9 Q. Please discuss the reasons that cost of service studies are utilized in regulatory
10 proceedings.
11 A. Cost of service studies represent an attempt to analyze which customer or group of
12 customers cause the utility to incur the costs to provide service. The requirement to
13 develop cost studies results from the nature of utility costs. Utility costs are characterized
14 by the existence of common costs. Common costs occur when the fixed costs of providing
15 service to one or more classes, or the cost of providing multiple products to the same class,
16 use the same facilities and the use by one class precludes the use by another class.
17 In addition,utility costs may be fixed or variable in nature. Fixed costs do not
18 change with the level of throughput, while variable costs change directly with changes in
19 throughput. Most non-fuel related utility costs are fixed in the short run and do not vary
20 with changes in customers' loads. This includes the cost of distribution mains and service
21 lines, meters, and regulators. The distribution assets of a gas utility do not vary with the
PAGE 21 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

1 level of throughput in the short run. In the long run, main costs vary with either growing
2 design day demand or a growing number of customers.
3 Finally, utility costs exhibit significant economies of scale. Scale economies result
4 in declining average cost as gas throughput increases and marginal costs must be below
5 average costs. These characteristics have implications for both cost analysis and rate
6 design from a theoretical and practical perspective. The development of cost studies
7 requires an understanding of the operating characteristics of the utility system. Further, as
8 discussed below, different cost studies provide different contributions to the development
9 of economically efficient rates and the cost responsibility by customer class.
10 Q. Please discuss the application of economic theory to cost allocation.
11 A. The allocation of costs using cost of service studies is not a theoretical economic exercise.
12 It is rather a practical requirement of regulation since rates must be set based on the cost of
13 service for the utility under cost-based regulatory models. As a general matter, utilities
14 must be allowed a reasonable opportunity to earn a return of and on the assets used to serve
15 their customers. This is the cost of service standard and equates to the revenue
16 requirements for utility service. The opportunity for the utility to earn its allowed rate of
17 return depends on the rates applied to customers producing that revenue requirement.
18 Using the cost information per unit of demand, customer, and energy developed in the cost
19 of service study to understand and quantify the allocated costs in each customer class is a
20 useful step in the rate design process to guide the development of rates.
21 However, the existence of common costs makes any allocation of costs problematic
22 from a strict economic perspective. This is theoretically true for any of the various utility
PAGE 22 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I costing methods that may be used to allocate costs. Theoretical economists have developed
2 the theory of subsidy-free prices to evaluate traditional regulatory cost allocations. Prices
3 are said to be subsidy-free so long as the price exceeds the incremental cost of providing
4 service but is less than stand-alone costs ("SAC"). The logic for this concept is that if
5 customers' prices exceed incremental cost, those customers contribute to the fixed costs of
6 the utility. All other customers benefit from this contribution to fixed costs because it
7 reduces the cost they are required to bear. Prices must be below SAC because the customer
8 would not be willing to participate in the service offering if prices exceed SAC.
9 SAC is an important concept for Intermountain because certain customers have
10 competitive options for the end uses supplied by natural gas through the use of alternative
11 fuels. As a result, subsidy-free prices permit all customers to benefit from the system's
12 scale and common costs, and all customers are better off because the system is sustainable.
13 If strict application of the cost allocation study suggests rates that exceed SAC for some
14 customers, prices must nevertheless be set below the SAC,but above marginal cost, to
15 ensure that those customers make the maximum practical contribution to the common costs
16 of the utility.
17 Q. If any allocation of common cost is problematic from a theoretical perspective, how is
18 it possible to meet the practical requirements of cost allocation?
19 A. As noted above, the practical reality of regulation often requires that common costs be
20 allocated among jurisdictions, classes of service, rate schedules, and customers within rate
21 schedules. The key to a reasonable cost allocation is an understanding of cost causation.
22 Cost causation, as alluded to earlier, addresses the need to identify which customer or
PAGE 23 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I group of customers causes the utility to incur particular types of costs. To answer this
2 question, it is necessary to establish a linkage between a Local Distribution Company's
3 ("LDC's") customers and the particular costs incurred by the utility in serving those
4 customers.
5 An important element in the selection and development of a reasonable COSS
6 allocation methodology is the establishment of relationships between customer
7 requirements, load profiles and usage characteristics on the one hand and the costs incurred
8 by the Company in serving those requirements on the other hand. For example,providing a
9 customer with gas service during peak periods can have much different cost implications
10 for the utility than service to a customer who requires off-peak gas service.
11 Q. Why are the relationships between customer requirements, load profiles, and usage
12 characteristics significant to cost causation?
13 A. The Company's distribution system is designed to meet three primary objectives: (1) to
14 extend distribution services to all customers entitled to be attached to the system; (2) to
15 meet the aggregate design day peak capacity requirements of all customers entitled to
16 service on the peak day; and(3) to deliver volumes of natural gas to those customers either
17 on a sales or transportation basis. There are certain costs associated with each of these
18 objectives. Also, there is generally a direct link between the manner in which such costs
19 are defined and their subsequent allocation.
20 Customer related costs are incurred to attach a customer to the distribution system,
21 meter any gas usage and maintain the customer's account. Customer costs are a function of
22 the number of customers served and continue to be incurred whether or not the customer
PAGE 24 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I uses any gas. They generally include capital costs associated with minimum size
2 distribution mains, services, meters, regulators and customer service and accounting
3 expenses.
4 Demand or capacity related costs are associated with plant that is designed,
5 installed, and operated to meet maximum hourly or daily gas flow requirements, such as
6 the transmission and distribution mains, or more localized distribution facilities that are
7 designed to satisfy individual customer maximum demands. Gas supply contracts also
8 have a capacity related component of cost relative to the Company's requirements for
9 serving daily peak demands and the winter peaking season.
10 Commodity related costs are those costs that vary with the throughput sold to, or
11 transported for, customers. Costs related to gas supply are classified as commodity-related
12 to the extent they vary with the amount of gas volumes purchased by the Company for its
13 sales service customers.
14 From a cost of service perspective, the best approach is a direct assignment of costs
15 where costs are incurred for a customer or class of customers and can be so identified.
16 Where costs cannot be directly assigned, the development of allocation factors by customer
17 class uses principles of both economics and engineering. This results in appropriate
18 allocation factors for different elements of costs based on cost causation. For example, we
19 know from the manner in which customers are billed that each customer requires a meter.
20 Meters differ in size and type depending on the customer's load characteristics. These
21 meters have different costs based on size and type. Therefore, meter costs are customer-
22 related,but differences in the cost of meters are reflected by using a different meter cost
PAGE 25 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I for each class of service. For some classes such as the largest customers, the meter cost
2 may be unique for each customer.
3 Q. How does one establish the cost and utility service relationships you previously
4 discussed?
5 A. To establish these relationships, the Company must analyze its gas system design and
6 operations, and its accounting records as well as its system and customer load data(e.g.,
7 annual, and peak period gas consumption levels). From the results of those analyses,
8 methods of direct assignment and common cost allocation methodologies can be chosen
9 for all of the utility's plant and expense elements.
10 Q. Please explain what you mean by the term "direct assignment."
11 A. The term direct assignment relates to a specific identification and isolation of plant and/or
12 expense incurred exclusively to serve a specific customer or group of customers. Direct
13 assignments best reflect the cost causation characteristics of serving individual customers
14 or groups of customers. Therefore, in performing a COSS, the cost analyst seeks to
15 maximize the amount of plant and expense directly assigned to particular customer groups
16 to avoid the need to rely upon other more generalized allocation methods. An alternative to
17 direct assignment is an allocation methodology supported by a special study as is done
18 with costs associated with meters and services.
19 Q. What prompts the analyst to elect to perform a special study?
20 A. When direct assignment is not readily apparent from the description of the costs recorded
21 in the various utility plant and expense accounts, then further analysis may be conducted to
22 derive an appropriate basis for cost allocation. For example, in evaluating the costs charged
PAGE 26 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I to certain operating or administrative expense accounts, it is customary to assess the
2 underlying activities, the related services provided, and for whose benefit the services were
3 performed.
4 Q. How do you determine whether to directly assign costs to a particular customer or
5 customer class?
6 A. Direct assignments of plant and expenses to particular customers or classes of customers
7 are made on the basis of special studies wherever the necessary data are available. These
8 assignments are developed by detailed analyses of the utility's maps and records, work
9 order descriptions, property records and customer accounting records. Within time and
10 budgetary constraints, the greater the magnitude of cost responsibility based upon direct
11 assignments, the less reliance need be placed on common plant allocation methodologies
12 associated with joint use plant.
13 Q. Is it realistic to assume that a large portion of the plant and expenses of a utility can
14 be directly assigned?
15 A. No. The nature of utility operations is characterized by the existence of common or joint
16 use facilities, as mentioned earlier. Out of necessity, then, to the extent a utility's plant and
17 expense cannot be directly assigned to customer groups, common allocation methods must
18 be derived to assign or allocate the remaining costs to the customer classes. The analyses
19 discussed above facilitate the derivation of reasonable allocation factors for cost allocation
20 purposes.
PAGE 27 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

VI. INTERMOUNTAIN'S COST OF SERVICE STUDY
1 Q. Please describe the process of performing Intermountain's COSS analysis.
2 A. Three broad steps were followed to perform the Company's COSS: (1) functionalization,
3 (2) classification, and(3) allocation. The first step, functionalization, identifies and
4 separates plant and expenses into specific categories based on the various characteristics of
5 utility operation. The Company's functional cost categories associated with gas service
6 include storage, transmission, distribution, and general (customer). The general function
7 includes costs that cannot be directly assigned to the primary operating functions of
8 storage, transmission, and distribution. These costs are functionalized in accordance with
9 the Federal Energy Regulatory Commission(FERC)Uniform System of Accounts
10 (USOA). Classification of costs, the second step, further separates the functionalized plant
11 and expenses into the three cost-defining characteristics previously discussed: (1)
12 customer, (2) demand or capacity, and(3) commodity, along with an additional revenue
13 classification consisting of working capital items and revenue. The final step is the
14 allocation of each functionalized and classified cost element to the individual customer
15 class. Costs typically are allocated on customer, demand, commodity, or revenue allocation
16 factors.
17 Q. Are there factors that can influence the overall cost allocation framework utilized by
18 a gas utility when performing a COSS?
19 A. Yes. The factors which can influence the cost allocation used to perform a COSS include:
20 (1)the physical configuration of the utility's gas system; (2) the availability of data within
PAGE 28 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

1 the utility; and(3) the state legislative and regulatory policies and evidentiary requirements
2 applicable to the utility.
3 Q. Why are these considerations relevant to conducting Intermountain's COSS?
4 A. It is important to understand these considerations because they influence the overall
5 context within which a utility's cost study was conducted. In particular, they provide an
6 indication of where efforts should be focused for purposes of conducting a more detailed
7 analysis of the utility's gas system design and operations and understanding the regulatory
8 environment in the State of Idaho as it pertains to cost of service studies and gas
9 ratemaking issues.
10 Q. Please explain why the physical configuration of the system is an important
11 consideration.
12 A. The particulars of the physical configuration of the transmission and distribution system
13 are important. The specific characteristics of the system configuration, such as whether the
14 distribution system is a centralized or a dispersed one, should be identified. Other such
15 characteristics are whether the utility has a single city-gate or a multiple city-gate
16 configuration, whether the utility has an integrated transmission and distribution system or
17 a distribution-only operation, and whether the system is a multiple pressure based or a
18 single pressure-based operation.
19 Q. What are the specific physical characteristics of Intermountain's system?
20 A. The physical configuration of Intermountain's system is a dispersed/multiple city-gate,
21 storage, transmission, distribution, and multi pressure-based system.
PAGE 29 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. What was the source of the cost data analyzed in the Company's COSS?
2 A. All cost of service data has been extracted from the Company's total cost of service (i.e.,
3 total revenue requirement) and subsidiary schedules contained in this filing.
4 Q. How does the availability of data influence a COSS?
5 A. The structure of the utility's books and records can influence the cost study framework.
6 This structure relates to attributes such as the level of detail, segregation of data by
7 operating unit or geographic region, and the types of load data available. Intermountain
8 maintains many detailed plant accounting records for its distribution-related facilities.
9 Q. How are Intermountain's classes structured for purposes of the COSS?
10 A. The COSS evaluated five customer classes: Residential (RS, IS-R), General (GS, IS-C),
11 Large Volume (LV-1), Interruptible Transport (T-3), and Firm Transport(T-4).
12 Q. Do you propose any modifications to the current classes?
13 A. No.
14 Q. Please describe the process of performing Intermountain's COSS analysis.
15 A. The detailed process description of Intermountain's COSS analysis is presented in Exhibit
16 35 - Cost of Service Study. Exhibit 35 provides a full scope of the COSS development
17 process and the results.
18 Q. Please discuss the content of Exhibit 35.
19 A. Exhibit 35 —Cost of Service Study consists of three sections detailing the process of
20 developing the COSS. The first section includes an introduction, the general purpose, and
21 an overview of the Excel-based fully functional COSS model presented in this proceeding.
22 The second section presents the COSS development process specific to the Company
PAGE 30 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I including Functionalization, Classification, and Allocation. The Allocation section
2 specifically describes all internal and external allocation factors and development bases
3 and processes used in the COSS. The last section depicts the results of the cost of service
4 study, including revenue requirement apportionment, comparison of cost of service with
5 revenues under present and proposed rates, and development of rate of return by customer
6 class under present and proposed rates.
7 Q. Please describe the schedules included in Exhibit 35.
8 A. The following is the list of Schedules included in Exhibit 35:
9 • Schedule 1 -Account Balances, Functionalization, Classification and Allocation
10 • Schedule 2 - External Allocation Factors
11 • Schedule 3 -Internal Allocation Factors
12 • Schedule 4 - Cost of Service and Rate of Return under Present and Proposed Rates
13 • Schedule 5 - Cost of Service Allocation Study Detail by Account
14 • Schedule 6 -Functionalized and Classified Rate Base and Revenue Requirement,
15 and Unit Costs by Customer Class
16 Q. Please explain the COSS information contained in Schedules 1 through 6.
17 Schedule 1 displays revenue requirements presented by FERC accounts with
18 corresponding selections of functions, classifications, and allocations methods applied to
19 the accounts. Schedule 2 and Schedule 3 depict the derivation of external and internal
20 allocation factors that are explained in detail in Exhibit 35. Schedule 4 is a summary of the
21 cost to serve as compared to revenues under present and proposed rates. Schedule 5 is a
22 detailed cost of service study presented by the FERC accounts for the individual rate
PAGE 31 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I classes. Schedule 6 presents a summary of functionalized and classified rate base and
2 revenue requirements along with unit cost derived by customer class.
3 Q. How did the COSS classify and allocate intangible plant?
4 A. The intangible plant accounts contain the costs related to the Company's software and
5 other intangible assets. Utility software includes customer billing software, plant
6 operations software and internal employee related software among other types of software
7 needed to operate the utility day to day. A review of records in miscellaneous intangible
8 plant account 303 was done to determine which plant was related to either utility plant
9 functions, customers, or labor. Plant-related intangible plant was then allocated to the
10 customer classes based on the allocation of storage, transmission, and distribution plant.
11 Customer-related intangible plant was allocated on the basis of customer counts, while
12 labor related intangible plant was allocated based on labor allocations.
13 Q. How did the COSS classify and allocate underground storage plant?
14 A. The storage plant accounts contain the costs related to the Company's LNG facilities.
15 These facilities are needed to provide deliverability and reliability during peak periods.
16 Because of the cost and cycle characteristics, LNG withdrawals are typically reserved for
17 needle peaking during very cold weather events or for system integrity events. Therefore,
18 the storage plant accounts are classified as demand and allocated on a peak day basis.
19 Q. How did the COSS classify and allocate transmission plant?
20 A. The transmission plant accounts contain the costs related to the Company's high pressure
21 transmission facilities. These facilities were designed and sized to provide deliverability
PAGE 32 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I during peak periods. Therefore, the transmission plant accounts are classified as demand
2 and allocated on a peak day basis.
3 Q. How did the Company's COSS classify and allocate investment in Distribution
4 Mains?
5 A. The Company classified 58 percent of its investment in distribution mains as customer-
6 related and 42 percent of the investment as demand-related. The customer-related portion
7 of the distribution mains investment was then allocated based on the number of customers
8 on Intermountain's distribution system. The demand-related investment was allocated to
9 the customer classes based on the respective contributions to peak day demand.
10 Q. Please explain the basis for the Company's choice of classification and allocation
11 methods?
12 A. It is widely accepted that distribution mains are installed to meet both system peak period
13 load requirements and to connect customers to the LDC's gas system. Therefore, to ensure
14 that the rate classes that cause the Company to incur this plant investment or expense are
15 charged with its cost, distribution mains should be allocated to the rate classes in
16 proportion to their peak period load requirements and number of customers.
17 There are two cost factors that influence the level of distribution mains facilities
18 installed by an LDC in expanding its gas distribution system. First, the size of the
19 distribution main(i.e., the diameter of the main) is directly influenced by the sum of the
20 peak period gas demands placed on the LDC's gas system by its customers. Secondly, the
21 total installed footage of distribution mains is influenced by the need to expand the
22 distribution system grid to connect new customers to the system. Therefore, to recognize
PAGE 33 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I that these two cost factors influence the level of investment in distribution mains, it is
2 appropriate to allocate such investment based on both peak period demands and the
3 number of customers served by the LDC.
4 Q. Is the method used by the Company to determine a customer cost component of
5 distribution mains a generally accepted technique for determining customer costs?
6 A. Yes. The two most commonly used methods for determining the customer cost component
7 of distribution mains facilities consist of the following: (1) the zero-intercept approach and
8 (2)the most commonly installed, minimum-sized unit of plant investment. Under the zero-
9 intercept approach, a customer cost component is developed through regression analyses to
10 determine the unit cost associated with a zero-inch diameter distribution main. The method
11 regresses current unit costs associated with the various sized distribution mains installed on
12 the LDC's gas system against the size (diameter squared inches) of the weighted
13 distribution mains installed. The zero-intercept method seeks to identify that portion of
14 plant representing the smallest size pipe required merely to connect any customer to the
15 LDC's distribution system, regardless of the customer's peak or annual gas consumption.
16 The most commonly installed, minimum-sized unit approach is intended to reflect
17 the engineering considerations associated with installing distribution mains to serve gas
18 customers. That is, the method utilizes actual current installed investment units to
19 determine the minimum distribution system rather than a statistical analysis based upon
20 investment characteristics of the entire distribution system.
21 Two of the more commonly accepted literary references relied upon when
22 preparing embedded cost of service studies, Electric Utility Cost Allocation Manual,by
PAGE 34 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I John J. Doran et al,National Association of Regulatory Utility Commissioners
2 ("NARUC"), and Gas Rate Fundamentals, American Gas Association, both describe
3 minimum system concepts and methods as an appropriate technique for determining the
4 customer component of utility distribution facilities.
5 Clearly, the existence and utilization of a customer component of distribution
6 facilities, specifically for distribution mains, is a fully supportable and commonly used
7 approach in the gas industry.
8 For purposes of determining the customer component of distribution mains to be
9 used in Intermountain's COSS, the zero-intercept method was employed, the detailed
10 development process of which is presented in Exhibit 35.
11 Q. Was the same method to classify and allocate distribution mains utilized in the 2022
12 Case?
13 A. Yes. The Company used similar classification and allocation methods in its previous
14 general rate case proceeding.
15 Q. How did the COSS classify and allocate the remainder of the distribution plant?
16 A. Special studies were performed for the allocation of Accounts 380 (Services), 381
17 (Meters), and 385 (Industrial Measuring and Regulating Station Equipment). The costs in
18 Account 383 (House Regulators) were classified and allocated to the Large Volume and
19 Transportation classes based upon the meters study. The development steps of these are
20 discussed in Exhibit 35.
21 The plant costs in Account 378 (Measuring and Regulating Station Equipment—
22 General) were allocated based on the allocation of distribution mains plant account.
PAGE 35 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Account 379 (Measuring and Regulating Station Equipment—City Gas Stations)was
2 classified as capacity or demand-related and allocated on a peak demand allocator.
3 Account 374 (Land and Land Rights) are associated with distribution mains and
4 therefore, were allocated based on the allocation of distribution mains plant account.
5 Account 375 (Structures and Improvements) and Account 387 (Other Equipment)were
6 allocated based on the allocation of the distribution plant accounts 376 through 385.
7 Q. Were there any changes in underlying allocation factors to distribution plant from
8 the last case?
9 A. Yes. Through review of the underlying data and discussion with company personnel,
10 changes were made to the Meters special study and Industrial Measuring and Regulation
11 Station Equipment ("Industrial M&R") special study. Specifically, the Meter study was
12 conducted in a manner that considered certain customer classes can have more than one
13 meter per customer(Large Volume and Transport classes). The Industrial M&R study was
14 refined from the prior case to exclude the General Service class because it was determined
15 that no General Service customer has meters or related meter station equipment of a size to
16 be accounted for in Account 385 Industrial M&R. The impact of these changes results in
17 more costs assigned to larger customers, specifically, Large Volume, Interruptible
18 Transport, and Firm Transport classes, but are a better reflection of cost causation on the
19 underlying plant.
20 Q. How did the COSS classify and allocate general plant?
21 A. General Plant was classified and allocated to the rate schedules based upon the allocation
22 of storage, transmission, and distribution plant. Mathematically, this is the sum of storage,
PAGE 36 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I transmission, and distribution plant accounts that were allocated by rate class. That total by
2 rate class is then divided by the total company amount to find each rate class's percentage
3 allocation. Account 391 (Office Furniture and Equipment) was allocated based on the
4 factor derived based on the Company's labor cost records.
5 Q. How are other rate base components classified and allocated in the COSS?
6 A. Accumulated Provision for Depreciation and Amortization is presented by FERC accounts
7 and allocated based on the same allocation factor as the related plant in service accounts.
8 This treatment ensures that the net plant for each FERC account is allocated consistently to
9 each customer class. Accumulated Deferred Income Taxes are presented on a functional
10 level and allocated based on the relevant internal plant allocator as shown in Exhibit 35.
11 Account 154 (Material and Supplies) was allocated based on the allocation of
12 storage, transmission, and distribution plant. Account 164 (LNG Inventory)balance was
13 allocated based on the peak day factor as the inventory exists to ensure reliability during
14 peak periods. Customer Account 252 (Advances for Construction)was allocated based on
15 the mains and services plant balances.
16 Q. How are operation and maintenance ("O&M"), customer accounts, customer services
17 and information ("Customer"), and administrative and general("A&G") expenses
18 classified and allocated in COSS?
19 A. A utility's O&M expenses support the corresponding plant in service accounts. In general,
20 O&M expenses are allocated based on the cost allocation methods used for the Company's
21 corresponding plant accounts. The majority of Customer expenses were classified as
22 customer-related costs and allocated based on the average number of distribution
PAGE 37 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I customers by class, except for Account No. 904 (Uncollectible Accounts Expense), which
2 is allocated based upon the three-year average of uncollectible write-offs.A&G expenses
3 were allocated on an account-by-account basis. Items related to labor costs, such as
4 employee pensions and benefits, were allocated based on O&M labor costs. Items related
5 to the plant in service, such as maintenance of the general plant and property taxes, were
6 allocated based on the plant allocator. The detailed classification and allocation methods
7 applied to these expense categories can be found on Schedule 1 of Exhibit 35.
8 Q. Were any additional studies performed in Intermountain's COSS?
9 A. Yes. Certain categories of gas supply and gas system control related O&M expenses
10 include salaries and benefits of personnel in the following responsibility centers: Gas
11 Supply Resource Planning, Gas Supply, and Gas Control. The corresponding labor
12 expenses were distributed among the three categories of Gas Planning, Gas Supply, and
13 Gas Control based on the labor hours allocations reported by the personnel in these
14 responsibility centers. These expenses were first segregated between sales and
15 transportation classes and then allocated to customer classes as discussed in Exhibit 35.
16 Q. Please discuss the classification and allocation of the remaining expenses.
17 A. Depreciation and amortization expense is presented on the functional level and allocated
18 based on the relevant internal plant allocator, as demonstrated in Exhibit 35. Taxes other
19 than income are allocated in a manner that reflects the specific cost associated with each
20 tax expense category. Generally, taxes can be cost classified on the basis of the tax
21 assessment method established for each tax category and can be grouped into the following
PAGE 38 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I categories: (1) labor; (2)plant; and(3)revenue. In Intermountain's COSS, all non-income
2 taxes were assigned to one of the above stated categories and relevant allocation factors.
3 Current income taxes were allocated based on each class's net income before taxes.
4 Income taxes for the total revenue requirement were allocated to each class based on the
5 allocation of the required net income by rate class. Income taxes at proposed revenues by
6 class were allocated to each class based on the proposed income prior to taxes for each
7 class.
8 Q. Please summarize the results of Intermountain's COSS.
9 A. Table 8 below presents a summary of the results of the Company's COSS that can be
10 reviewed in detail in Schedule 4 of Exhibit 35. The COSS shows an overall revenue
11 deficiency for the Company of$26.5 million.
Table 8—Summary Results of the COSS
Customer Classes Current Revenues Cost to Serve Current Rate of Class Revenue Current
-
Return Excess Cost Ratio Parity Ratio
Residential Service $ 79,817,679 $ 112,527,159 1.89/6 $ (32,709,480) 0.72 0.87
General Service 28,264,261 22,085,001 15.49/6 6,179,260 1.27 1.54
Large Volume 777,024 1,010,145 2.89/6 (233,121) 0.78 0.94
Transport Service(Interruptible) 563,913 218,937 66.1% 344,976 2.52 3.05
TransportService(Firm) 10,172,451 10,253,575 8.69/6 (81,124) 0.99 1.20
Subotal I $ 119,595,328 $ 146,094,817 1 $ (26,499,489)
Other Revenues 1 5,320,360 5,320,360
Total System 1 $ 124,915,688 $ 151,415,177 r 4.5% $ (26,499,489)1 0.82 1.00
12 Table 8 presents the revenue deficiency/excess for each rate class, the class rate of return
13 on net rate base at current rates, the revenue to cost ratio, and the associated parity ratio.
14 Regarding class revenue levels, the results show that the Residential, and Large Volume
15 classes are charged rates that do not recover their indicated costs of service, the General
16 Service and Interruptible Transport Service classes are charged rates that recover more
PAGE 39 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I than their indicated cost of service, and the Firm Transportation Service class is charged
2 rates that recovery within one percent of its cost of service.
VII. PRINCIPLES OF SOUND RATE DESIGN
3 Q. Please identify the principles of rate design utilized in development of the Company's
4 rate design proposals.
5 A. Several rate design principles find broad acceptance in the recognized literature on utility
6 ratemaking and regulatory policy. These principles include:
7 (1) Cost of Service;
8 (2) Efficiency;
9 (3) Value of Service;
10 (4) Stability/Gradualism;
11 (5) Non-Discrimination;
12 (6) Administrative Simplicity; and
13 (7) Balanced Budget.
14 These rate design principles draw heavily upon the "Attributes of a Sound Rate Structure"
15 developed by James Bonbright in Principles of Public Utility Rates.10
16 Q. Can the objectives inherent in these principles compete with each other at times?
17 A. Yes. These principles can compete with each other, and this tension requires further
18 judgment to strike the right balance between the principles. Detailed evaluation of rate
19 design recommendations must recognize the potential and actual tension between these
20 principles. Indeed, Bonbright discusses this tension in detail. Rate design recommendations
10 principles of Public Utility Rates,Second Edition,Page 111-113 James C.Bonbright,Albert L. Danielson,David
R.Kamerschen,Public Utility Reports,Inc., 1988.
PAGE 40 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I must deal effectively with such tension. There are tensions between cost and value of
2 service principles as well as efficiency and simplicity. There are potential conflicts between
3 simplicity and non-discrimination and between value of service and non-discrimination.
4 Other potential conflicts arise where utilities face unique circumstances that must be
5 considered as part of the rate design process.
6 Q. How are these principles translated into the design of rates?
7 A. The overall rate design process, which includes both the apportionment of the revenues to
8 be recovered among rate classes and the determination of rate structures within rate
9 classes, consists of finding a reasonable balance between the above-described criteria or
10 guidelines that relate to the design of utility rates. Economic, regulatory, historical, and
11 social factors all enter the process. In other words,both quantitative and qualitative
12 information is evaluated before reaching a final rate design determination. Out of necessity
13 then, the rate design process must be, in part, influenced by judgmental evaluations.
VIII. DETERMINATION OF PROPOSED CLASS REVENUES
14 Q. Please describe the approach generally followed to allocate Intermountain's proposed
15 revenue increase of$26.5 million to its rate schedules.
16 A. The apportionment of revenues among rate schedules consists of deriving a reasonable
17 balance between various criteria or guidelines that relate to the design of utility rates. The
18 various criteria that were considered in the process included: (1) cost of service; (2) rate
19 schedule contribution to present revenue levels; and(3) customer impact considerations.
20 These criteria were evaluated for Intermountain's rate schedules.
PAGE 41 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. Have various rate schedule revenue options been considered in conjunction with your
2 evaluation and determination of Intermountain's interclass revenue proposal?
3 A. Yes. Using Intermountain's proposed revenue increase, and the results of its COSS, a few
4 options were evaluated for the assignment of that increase among its rate schedules and, in
5 conjunction with Intermountain personnel and management, ultimately decided upon one
6 of those options as the preferred resolution of the interclass revenue issue. The benchmark
7 option that was evaluated under Intermountain's proposed total revenue level was to adjust
8 the revenue level for each rate schedule so that the R:C ratio for each class was equal to
9 parity or 1.00 (Unity), as shown in Exhibit 36, under Scenario A: Revenues at Equalized
10 Rates of Return. Rate schedules above parity would suggest the need for revenue decreases
11 in order to move them closer to cost(i.e., a convergence of the resulting revenue-to-cost
12 ratios towards unity or 1.00).
13 The resulting customer impact implications for the Residential Service class have
14 led to the conclusion, in consultation with the Company, to refrain from revenue reductions
15 for the remaining customer classes. From a policy perspective, Intermountain believed that
16 every rate schedule should participate in the proposed overall revenue increase. Therefore,
17 as a matter of judgment, it was decided that this fully cost-based option was not the
18 preferred solution to the interclass revenue question. It should be pointed out, however,
19 that those class revenue results represented an important guide for purposes of evaluating
20 subsequent rate design options from a cost of service perspective.
21 A second option considered was assigning the increase in revenues to
22 Intermountain's rate schedules based on an equal percentage basis of its current margin
PAGE 42 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I revenues (see Scenario B, Equal Percentage Increase), in Exhibit 36. By definition, this
2 option resulted in each rate schedule receiving an increase in revenues equal to the system
3 average. However, when this option was evaluated against the COSS Study results (as
4 measured by changes in the revenue-to-cost ratio for each customer class); there was no
5 movement towards cost for most of Intermountain's rate schedules (i.e., there was no
6 convergence of the resulting revenue-to-cost ratios towards unity or 1.00). While this
7 option was not the preferred solution to the interclass revenue issue, together with the fully
8 cost-based option, it defined a range of results that provides further guidance to develop
9 Intermountain's class revenue proposal.
10 A third option considered was moderately assigning the increase in revenues to all
11 Intermountain's rate schedules (Scenario C:Moderated based on Current Parity Ratio),
12 which is the proposed revenue allocation method in this proceeding.
13 Q. What was the result of this process?
14 A. The various criteria that were considered in the process included: (1) cost of service; (2)
15 class contribution to present revenue levels; and(3) customer impact considerations. After
16 further discussions with Intermountain, the conclusion reached was that the appropriate
17 interclass revenue proposal would consist of adjustments, in varying proportions, to the
18 present revenue levels in all of Intermountain's rate schedules.
19 The Residential and Large Volume margin revenue increase was limited to 26.59
20 percent or 1.20 times the average system increase of 22.16 percent. The minimum increase
21 was applied to the Interruptible Transport of 0.25 of the average system increase, which
22 resulted in 5.54 percent of margin revenue increase. The remainder of the margin revenue
PAGE 43 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I increase was allocated among General Service and Firm Transport rate schedules, which
2 resulted in a 13.11 percent margin revenue increase or 0.59 of the average system increase.
3 This revenue apportion is shown in Direct Exhibit 36 as Proposed Scenario C:Moderated
4 based on the Current Parity Ratio.
5 Q. What is the recommended increase for each rate class?
6 A. In summary, this preferred revenue allocation approach resulted in reasonable movement
7 of the customer classes' revenue-to-cost ratio toward unity as shown on Table 9 below,
8 while providing moderation of the revenue impact by requiring some level of revenue
9 increase responsibility from all rate schedules for the Company's total proposed revenue
10 requirement.
Table 9—Current and Proposed Parity Ratios
ProposedCustomer Classes Current Parity
Ratio Ratio
Residential Service 0.87 0.90
General Service 1.54 1.43
Large Volume 0.94 0.97
Transport Service(Interruptible) 3.05 2.66
Transport Service(Firm) 1.20 1.12
Total System 1.00 1.00
11 From a class cost of service standpoint, this type of rate schedule movement, and modest
12 reduction in the existing class rate subsidies, is desirable.
13 The following Table 10 summarizes the proposed distribution margin revenue
14 change for each rate class and the percent change in distribution margin revenues resulting
15 from the above-described process.
PAGE 44 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

Table 10—Proposed Class Revenue Apportionment
Customer Classes Margin Revenues Margin Revenues Proposed Revenue Percent Change Relative to Proposed
at Current Rates at Proposed Rates Change Syste m Parity Ratio
Increase
Residential Service $ 79,817,679 $ 101,040,525 $ 21,222,846 26.59% 1.20 0.90
General Service 28,264,261 31,969,521 3,705,260 13.11% 0.59 1.43
Large Volume 777,024 983,628 206,604 26.59% 1.20 0.97
Transport Service(Interruptible) 563,913 595,150 31,237 5.50o' 0.25 2.66
TransportService(Firm) 10,172,451 11,505,993 1,333,542 13.11% 0.59 1.12
Subotal I $ 119,595,328 1$ 146,094,817 1 $ 26,499,489 1 22.16%1 1.00
Other Revenues 1 5,320,360 1 5,320,360
Total System 1 $ 124,915,688 1$ 151,415,177 1 $ 26,499,489 1 21.21%1 11.00
IX. INTERMOUNTAIN'S RATE DESIGN
1 Q. Please summarize the rate design changes Intermountain has proposed in this rate
2 proceeding.
3 A. The proposed rate design includes (1) increases in the fixed monthly customer charges for
4 all rate classes, (2) a two-step process for increasing fixed monthly customer charge for
5 Residential class, and(3) increases in demand rates to Large Volume and Firm Transport
6 classes. Once the fixed monthly customer charge targets and demand rates were set for
7 each rate class, the remaining proposed revenues for each rate class were recovered
8 through the volumetric charges.
9 Q. Please describe the changes to the monthly customer charge levels.
10 A. Table 11 provides a summary of current and proposed customer charges by rate schedule as
11 compared to the COSS results:
PAGE 45 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

Table 11 -Current and Proposed Customer Charge
Rate Classes Customer COSS Unit Cost Customer Percent
Charge Cust-Related Current Proposed
-
Residential Service Step 1 $ 14.00 $ 6.00 75%
Residential Service Step 2* $ 8.00 $ 20.33 $ 24.12 $ 20.00 $ 6.00 43%
Residential Service(Interruptible)Step 1 $ 14.00 $ 6.00 75%
Residential Service(Interruptible)Step 2* $ 8.00 $ 20.33 $ 24.12 $ 20.00 $ 6.00 43%
General Service $ 15.00 $ 34.28 $ 51.34 $ 40.00 $ 25.00 167%
General Service(Interruptible) $ 15.00 $ 34.28 1 $ 51.34 $ 40.00 $ 25.00 167%
Large Volume $ 150.00 $ 1,317.80 $ 2,295.90 $ 375.00 $ 225.00 150%
Transport Service(Firm) $ 150.00 $ 1,897.83 $ 8,434.11 $ 300.00 $ 150.00 100%
Transport Service(Interruptible) $ 300.00 $ 1,879.68 $ 2,101.02 $ 600.00 1$ 300.00 100%
*change columns compare to Step 1 proposed
1 Overall, the proposed customer charges are within reasonable range of increases
2 considering the customer and customer/demand unit costs per rate class supported by the
3 COSS results, as indicated on Schedule 6 of Exhibit 35. These increases to the basic
4 customer charges will provide significant improvement in the recovery of the fixed
5 customer-related costs via fixed charges. To offset the foregoing increases to the basic
6 customer charges, all blocks of the volumetric rates in the respective tariff schedules were
7 reduced ratably based on the margin revenue in each block.
8 Q. Why is the Company proposing to increase the fixed monthly customer charges?
9 A. The primary goal of rate design was to move towards recovery of fixed costs by increasing
10 all customer charges. This resulted in better alignment between the fixed costs incurred by
11 Intermountain and the charges incurred by customers. Regarding the General Service class,
12 while the proposed customer charge is above the customer-related COSS unit costs, it is
13 less than the customer& demand-related COSS unit costs, which is reasonable because the
14 General Service class has no demand charge. In the absence of a demand charge, which is
PAGE 46 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I designed as a means to recover the demand-related costs of the utility, it is reasonable to
2 include those fixed costs in the customer charge.
3 Q. Please explain the two-step process for increasing the fixed monthly customer charge
4 for the Residential class.
5 A. Intermountain proposes to increase the residential customer charge in two steps, over a
6 two-year period. As noted in Table 11 above, the first step would take the Residential class
7 from its current customer charge of$8.00 to $14.00 per month. In step two, a revenue
8 neutral change will increase the customer charge from $14.00 to $20.00 per month, with an
9 offsetting reduction to the delivery charge. As shown in Table 11 above, the Residential
10 class's Customer-related costs are $20.33 per customer per month, which is above the
11 proposed $20.00 Residential customer charge in step two. The reasoning for the two-step
12 process is to allow customers, namely low use customers, to adjust to their bill including a
13 higher customer charge. It should be noted that an increase in the fixed recovery of costs
14 through a fixed customer charge in Step 2 will simultaneously see a decrease in the
15 volumetric recovery of costs through volumetric charges. Said another way, an average
16 customer's bill will not dramatically change because as the customer charge increases, the
17 volumetric charge will decrease proportionally. The Company's gradual approach to
18 increasing the residential customer charge should not be confused with the principle of
19 gradualism as discussed above. Gradualism should be considered in the context of the total
20 revenue responsibility and total customer bill for which I just described as including a
21 fixed and volumetric component.
PAGE 47 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I Q. Please describe the changes proposed to the demand rate.
2 A. The current demand charge in Large Volume and Firm Transportation classes of$0.32 per
3 therm per month is proposed to be raised to $0.44, which will recover approximately 100
4 percent of the unit demand-related costs for these customer classes.
5 Q. Have you provided an exhibit detailing the proposed rates and corresponding
6 revenues?
7 A. Yes. Exhibit 37 shows the derivation of each rate component for each of Intermountain's
8 tariff schedules and the corresponding revenues generated from those proposed rates.
9 Q. Have you prepared bill impacts?
10 A. Yes. Exhibit 38 provides monthly bill impacts for Residential, General, and Interruptible
11 Transportation rate classes presented as a range of monthly usage (therms) and
12 corresponding bills under current and proposed rates. The bill impacts for Large Volume
13 and Firm Transportation customers are presented as various scenarios of monthly usage
14 and MDFQ with corresponding bills under current and proposed rates.
X. CONCLUDING REMARKS
15 Q. Please summarize your recommendations.
16 For purposes of Intermountain's allocated class cost of service study, the Load Study
17 results which use the Monthly peak load sendout model to determine the Core peak day
18 sendout are recommended. It provides superior results in predicting peak day sendout.
19 These results are aligned with Intermountain's projections of peak day sendout in its 2023-
20 2028IRP.
PAGE 48 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY

I I recommend the Commission accept the COSS presented in Section VI of this
2 testimony, including the proposed class revenue apportionment. The COSS represents a
3 fair and reasonable allocation of cost responsibility for each rate class,based on the
4 Company's proposed total system revenue increase. The Company's proposed COSS
5 allocation method for distribution mains best reflects the cost causative characteristics of
6 extending service to new customers and sized to meet peak demand requirements. As such,
7 the Commission should rely on the Company's proposed COSS to guide revenue targets
8 for each rate class.
9 The revenue targets proposed by Intermountain reasonably balance the concepts of
10 cost of service, current revenue contributions, and gradualism, while moving all classes
11 closer to parity. Lastly, the COSS model demonstrates that fixed costs, both customer-
12 related and demand-related are materially higher than the current level of customer
13 charges; therefore, the proposed increases to customer charges should be approved by the
14 Commission to better align fixed cost occurrence with fixed cost recovery and price signals
15 received by customers.
16 Q. Does this conclude your testimony?
17 A. Yes, although I reserve the right to supplement or amend my testimony before or during
18 the Commission's hearing in this proceeding.
PAGE 49 OF 50
R.AMEN,DI
INTERMOUNTAIN GAS COMPANY