20040209Knox Direct.pdf

DAVIDJ. MEYER
SENIOR VICE PRESIDENT AND GENERAL COUNSEL
A VISTA CORPORATION
O. BOX 3727
1411 EAST MISSION AVENUE
SPOKANE, W AS IllNGTON 99220-3727
TELEPHONE: [phone]
FACSIMILE: [phone]
BEFORE THE IDAHO IJUBLIC UTILITIES COMMISSION
IN THE MATIER OF THE APPLICATION
OF A VISTA CORPORATION FOR THE
AUTHORITY TO INCREASE ITS RATES
AND CHARGES FOR ELECTRIC AND
NATURAL GAS SERVICE TO ELECTRIC AND
NATURAL GAS CUSTOMERS IN THE STATE OF IDAHO
CASE NO. A VU-04-
CASE NO. A VU-04-
DIRECT TESTIM:ONY
TARA L. KNOX
FOR A VISTA CORPORATION
(ELECTRIC AND NATURAL GAS)
Please state your name, business address and present position with A vista
Corporation?
My name is Tara L. Knox and my business address is 1411 East Mission
Avenue, Spokane, Washington.I am employed as a Rate Analyst in the Rates and
Regulation Department.
Would you briefly describe your duties?
I am responsible for preparing data for and maintaining the regulatory cost of
service models for the Company as well as providing support in the preparation of results of
operations reports and miscellaneous other duties as required.
Would you describe your educational background and professional
experience?
I graduated from Washington State University with a Bachelor of Arts degree
in General Humanities in 1982 and a Master of Accounting degree in 1990. As an employee
in the Rate Department at Avista since 1991 I have attended several ratemaking classes
including the EEl Electric Rates Advanced Course that specializes in cost allocation and cost
of service issues. I have also been a member of the Cost of Service Working Group since
1999, which is a discussion group made up of technical professionals from utilities
throughout the United States and Canada concerned with cost of service issues.
What is the scope of your testimony in these proceedings?
My testimony and exhibits will cover the Company s electric and natural gas
cost of service studies performed for these proceedings and the weather normalization
adjustments to retail usage.
Knox, Di
A vista Corporation
ELECTRIC SERVICE
ELECTRIC WEATHER NORMALIZATION
Would you please briefly summarize your testimony related to electric
weather normalization?
The Company s weather normalization adjustment incorporates the effect of
both heating and cooling on weather sensitive customer groups. The weather adjustment
developed from regression analysis of five years of billed usage per customer, billing period
heating degree day data and billing period cooling degree day data. The resulting weather
sensitivity coefficients for each customer subgroup are multiplied by the average number of
customers in each subgroup during the test period and the difference between normal
heating/cooling degree days and test period observed heating/cooling degree days. This
calculation produces the change in kWh usage required to adjust existing loads to the amount
expected if weather had been normal. Mr. Hirschkorn includes the adjustment to normal
usage as part of the Revenue Adjustment for pro forma results of operations. Mr. Kalich
includes the adjustment to normal loads in the modeling for the Pro Forma Power Supply
Adjustment for pro forma results of operations.
Is this different from the method employed in the Company s prior cases?
Yes, although the actual methodology has changed very little. The prior
method did not include the effect of weather sensitive cooling. During the regression phase
of the process, more combinations of variables are tested to arrive at the best fit. Also, the
time period used for the analysis was modified to reflect exactly five heating seasons, July
through June, rather than the five and one-half heating seasons included in the prior method.
Knox, Di
A vista Corporation
The application of the results of the regression analysis is the same as the prior method, only
now we apply both the difference between normal and actual cooling degree days as well as
normal and actual heating degree days.
Why is it important to include cooling sensitivity in the electric weather
normalization process?
Analysis of the billed usage data since the late 1990's have indicated that
summer weather sensitive usage has become significant for many of the customer groups.
This is most likely reflective of increased saturation of the air conditioning market in the
region.Although normally a winter peaking utility, in recent years the Company has
experienced summer peaks near the same level as the winter peaks. In fact, in 2002 the
annual system peak occurred during July. Without incorporating cooling sensitivity the prior
method would add usage during an abnormally hot summer due to fewer than normal heating
degree days.
ELECTRIC COST OF SERVICE
Would you please briefly summarize your testimony related to the electric
cost of service study?
I believe the Base Case cost of service study presented in this case is a fair
representation of the costs to serve each customer group. For comparison purposes, I have
also provided results of an alternative scenario to illustrate the impact of different allocation
decisions in the cost of service process.
Knox, Di
A vista Corporation
The Base Case study shows Residential Service Schedule 1 and Extra Large General
Service Schedule 25 (not including the Potlatch Lewiston plant) earn substantially less than
the overall rate of return under present rates. General Service Schedule 11, Large General
Service Schedule 21, and Pumping Service Schedule 31 earn substantially more than the
overall rate of return under present rates (although less that the requested rate of return). The
Potlatch Lewiston plant (at Schedule 25 rates) and Street and Area Lights earn close to the
overall rate of return under present rates.
Are you sponsoring any exhibits related to the electric cost of service
study?
Yes. I am sponsoring Exhibit No. 16 divided into the following Schedules:
Schedule 1, electric cost of service study process description; Schedule 2, Electric Base Case
cost of service study model output; and Schedule 3 , alternate scenario summary results.
Was this exhibit prepared by you or under your supervision?
Yes.
What is an electric cost of service study and what is its purpose?
An electric cost of service study is an engineering-economic study, which
apportions the revenue, expenses, and rate base associated with providing electric service to
designated groups of customers. The groups are made up of customers with similar load
characteristics and facilities requirements. Costs are assigned in relation to each groups
characteristics, resulting in an evaluation the cost of the service provided to each group. The
rate of return by customer group indicates whether the revenue provided by the customers in
each group recovers the cost to serve those customers. The study results are used as a guide
Knox, Di
A vista Corporation
in determining the appropriate rate spread among the groups of customers. Schedule 1 of
Exhibit No. 16 explains the basic concepts involved in performing an electric cost of service
study. It also details the specific methodology and assumptions utilized in the Company
Base Case cost of service study.
What is the basis for the electric cost of service study provided in this
case?
The electric cost of service study provided by the Company as Exhibit No. 16,
Schedule 2 is based on the 2002 test year pro-forma results of operations presented by Mr.
Falkner in Exhibit No. 14.
Would you please describe what is shown in Schedule 2?
Exhibit No. 16, Schedule 2 is the Electric Cost of Service Study. The exhibit
shows the Excel spreadsheet model calculation of the cost of service results. This detail has
been divided into three distinct segments.
Part 1 is composed of a series of summaries of the study results. The summary on
page 1 shows the results of the study by FERC account category. The rate of return by rate
schedule and the ratio of each schedule s return to the overall return are shown on Lines 39
and 40. This summary was provided to Mr. Hirschkorn for his work on rate spread and rate
design. The results will be discussed in more detail later in my testimony.
Pages 2 and 3 are both summaries that show the revenue to cost relationship at current
and proposed revenue. Costs by category are shown first at the existing schedule returns
(revenue); next the costs are shown as if all schedules were providing equal recovery (cost).
These comparisons show how far current and proposed rates are from rates that would be in
Knox, Di
A vista Corporation
alignment with the cost study.Page 2 shows the costs segregated into production,
transmission, distribution, and common functional categories. Page 3 segregates the costs
into demand, energy, and customer classifications.
Part 2 is the cost of service calculations from the spreadsheet called "Assign" showing
the functionalization, classification, and allocation of each line item in the study. The
supporting schedules required to run the model made up of the allocation and classification
factors used in the study are shown on pages 31 through 35.
Finally, Part 3 is the spreadsheet called "Proforma.This worksheet shows the
segregation of Mr. Falkner s pro forma results of operations into the detailed accounting data
used in this study.
Base Case Cost of Service. Electric
Does the Company s electric Base Case cost of service study follow the
methodology filed in the Company s last electric general rate case in Idaho?
The methodology is the same as the cost of service study filed in Case No.
WWP-98-11 with one modification.
Please explain this modification.
Administrative and general costs that cannot be directly assigned to
production, transmission, distribution, or customer relation s functions are left in the
common cost category.In Avista s 1998 case these common costs were allocated to
customer groups by a 60% customer-40% energy allocation factor. In this case the allocation
factor for these common costs has been modified to reflect a four-factor allocation based on
direct O&M, direct labor, net direct plant, and number of customers. With this change the
Knox, Di
A vista Corporation
same four-factor allocation used on common costs at the utility and jurisdictional levels is
now also applied at the customer group level.
Why did you choose to make this modification?
As I was replicating the methodology from WWP-98-11 to prepare the cost
studies for this case, I considered the need to update the common cost allocator. The four-
factor allocator is accepted in all of the Company s jurisdictions for determining the
appropriate sharing of common costs for results of operations. It is primarily based on other
costs within the study, and reflects a variety of relationships rather than being solely
dependent on a single comparison. The four-factor provides a balanced approach that I
consider more appropriate than the factor used in the last case.
What are the results of the Company s Base Case cost of service study?
The following table shows the rate of return and the ratio of the schedule
return to the overall return (relative return ratio) at present rates for each rate schedule:
Table 1
Customer Class
Residential Service Schedule 1
Rate of Return Return Ratio
97%
70%
12%1.73
1.17%
24%1.11
24%1.54
55%
71%1.00
General Service Schedule 11
Large General Service Schedule 21
Extra Large General Service Schedule 25
Potlatch Ex Lg Gen Service Schedule 25P
Pumping Service Schedule 31
Lighting Schedules 41 - 49
Total Idaho Electric
Knox, Di
Avista Corporation
As can be observed from the above table, residential and extra large general service
schedules (1 and 25) show significant under-recovery of the costs to serve them. The
summary results of this study were provided to Mr. Hirschkorn as an input into development
of the proposed rates.
Would you please explain the significance of Schedules 25 and 25P in the
table above?
There are currently 15 customers served on Schedule 25 including the 100
average megawatt load from the Potlatch facilities in Lewiston, Idaho (potlatch Lewiston).
Potlatch Lewiston alone has nearly three times the usage of the other fourteen Schedule 25
customers combined. Prior to 2002 Potlatch Lewiston was served on a special contract with
a sharing of their retail revenue between the Idaho and Washington jurisdictions. Since
January of 2002 Potlatch Lewiston has been served at Schedule 25 rates. This is the first
Idaho embedded cost study to reflect Potlatch Lewiston' s full 100 average megawatt load. In
this case Potlatch Lewiston has been evaluated as a separate cost of service class, due
primarily to the load being significantly higher than other Schedule 25 customers.
Why is the rate of return for Potlatch Lewiston higher than the rate of
return for the remainder of Schedule 25?
There are two primary factors driving the cost differences between the
Potlatch Lewiston plant and the other Schedule 25 customers.
First, Potlatch Lewiston has a significantly higher load factor (98% at the time of the
system peak compared to 77% for the rest of Schedule 25). The cost study reflects load
factor through the relative allocation of energy related costs compared to demand related
Knox, Di
A vista Corporation
costs. Schedule 25 customers are allocated approximately ten percent of energy related costs
and nine percent of demand related costs.Potlatch on the other hand is allocated
approximately twenty-eight percent of energy related costs but only twenty percent of
demand related costs. The net effect is less demand related production and transmission
costs are allocated to the Potlatch Lewiston plant relative to their consumption than the rest
of schedule 25 customers that have a lower load factor.
Second, Potlatch Lewiston is excluded from allocations of demand related primary
distribution plant. This includes FERC accounts 364 through 368 comprised of poles,
conduit, and overhead or underground conductors & devices. The situation at the Potlatch
Lewiston plant is unique in that they receive primary voltage power at the 115 kV substation
that is dedicated to them. No Company owned primary distribution plant is interconnected
with that substation, therefore exclusion from the allocation is appropriate. The cost of the
substation is directly assigned to Potlatch. The net effect is less distribution facility costs are
assigned or allocated to the Potlatch Lewiston plant relative to their consumption than the rest
of Schedule 25 customers who do receive the benefit of the interconnected primary
distribution system.
Alternative Scenario
Were the results of the Base Case methodology compared to the
methodology from Case No. WWP-98-11 (A vista's last general rate filing)?
Yes, the alternative scenario shown in Exhibit No. 16, Schedule 3 represents
the results using the methodology from Case No. WWP-98-11. The only difference is the
Knox, Di
A vista Corporation
allocation factor used for common costs. In this scenario common costs are allocated 60% by
number of customers and 40% by annual customer level consumption. The effect of the prior
methodology is to have a heavier emphasis on number of customers. Table 2 below shows
the relative return ratios from this scenario in comparison to the Base Case.
Table 2
Customer Grou Base Case WWP-98-Difference
Residential Service Schedule 1 0.42
General Service Schedule 11 1.99
Large General Service Schedule 21 1.73 1.97 +0.24
Extra Large General Service Schedule 25 +0.
Potlatch Ex Lg Gen Service Schedule 25P 1.11 1.19 +0.
Pumping Service Schedule 31 1.54 1.66 +0.
Lighting Schedules 41 - 49 1.39 +0.42
Residential and General Service schedules that have relatively low usage per
customer show lower relative rates of return in this scenario which emphasizes the number of
customers. The change in non-metered lighting service is dramatic because most individual
area light customers that also take service on another schedule are counted as customers only
on their metered service schedule. The municipal street lighting customers that do receive
separate bills for lighting service and therefore are counted as lighting customers tend to have
hundreds of lights. This phenomenon causes changes in the amount of customer allocation to
have a greater effect on the results for lighting service.
Knox, Di
A vista Corporation
What conclusions do you draw from the results of these two cost of
service studies?
In both scenarios residential and extra large general service customers are providing
less than the cost to serve them. General service, large general service, and pumping service
customers are consistently above the overall rate of return. Potlatch Lewiston is close to the
overall return. Lighting service is also close to the overall return in the Base Case.
NATURAL GAS SERVICE
NATURAL GAS WEATHER NORMALIZATION
Would you please briefly summarize your natural gas weather
normalization testimony?
No change has been made to the historical methodology used to calculate
natural gas weather sensitivity.The weather adjustment is developed from regression
analysis of five and one-half years of billed usage per customer and billing period heating
degree day data. The resulting weather sensitivity coefficient for each customer subgroup is
multiplied by the average number of customers in the subgroup during the test period and the
difference between normal heating degree days and test period heating degree days. This
calculation produces the change in therm usage required to adjust existing loads to the
amount expected if weather had been normal. Mr. Hirschkorn includes the adjustment to
normal usage as part of the Revenue/Gas Supply Adjustment for pro forma results of
operations.
Knox , Di
Avista Corporation
Is this different from the method employed in the Company s prior cases?
No. This method has been utilized in the Company s last Idaho natural gas
general rate filing as well as the semi-annual commission basis reporting.
The Company is proposing to modify the weather normalization
methodology for electric usage, why not for natural gas usage as well?
The change to the electric methodology was necessary to reflect the impact of
air conditioning load during the summer months. Natural gas is not used for air conditioning,
the usage per customer data shows no cooling sensitivity, and the current regression fit
statistics for the weather sensitive subgroups are excellent. Therefore, there is no need to
change the existing methodology.
NA TURAL GAS COST OF SERVICE
Would you please briefly summarize your testimony related to the
Company s natural gas cost of service study?
Yes. I believe the Base Case cost of service study presented in this case is a
fair representation of the costs to serve each customer group. The study indicates that
General Service Schedule 101 (primarily residential customers) is earning slightly less than
the overall return, all other schedules are earning more than the overall return but less than
the requested return.
Knox, Di
A vista Corporation
Are you sponsoring any exhibits related to the natural gas cost of service
study?
Yes. I am sponsoring Exhibit No. 17 divided into the following Schedules:
Schedule 4, natural gas cost of service study process description; and Schedule 5, natural gas
cost of service study model output.
Was this exhibit prepared by you or under your supervision?
Yes.
Please describe the natural gas cost of service study and its purpose.
A natural gas cost of service study is an engineering-economic study which
apportions the revenue, expenses, and rate base associated with providing natural gas service
to designated groups of customers. The groups are made up of customers with similar usage
characteristics and facility requirements. Costs are assigned in relation to each groups
characteristics, resulting in an evaluation of the cost of the service provided to each group.
The rate of return by customer group indicates whether the revenue provided by the
customers in each group recovers the cost to serve those customers. The study results are
used as a guide in determining the appropriate rate spread among the groups of customers.
Exhibit No. 17, Schedule 4 explains the basic concepts involved in performing a natural gas
cost of service study. It also details the specific methodology and assumptions utilized in the
Company s Base Case cost of service study.
Knox, Di
A vista Corporation
What is the basis for the natural gas cost of service study provided in this
case?
The cost of service study provided by the Company as Exhibit No. 17,
Schedule 5 is based on the 2002 test year pro-forma results of operations presented by Mr.
Falkner in Exhibit No. 15.
Would you please describe what is shown in Schedule 5?
Exhibit No. 17, Schedule 5 is the Natural Gas Cost of Service Study. The
exhibit shows the Excel spreadsheet model calculation of the cost of service results. This
detail has been divided into three distinct segments.
Part 1 is composed of a series of summaries of the study results. Page 1 shows the
results of the study by FERC account category. The rate of return and the ratio of each
schedule s return to the overall return are shown on lines 38 and 39. This summary is
provided to Mr. Hirschkorn for his work on rate spread and rate design. The results will be
discussed in more detail later in my testimony. The additional summaries show the costs
organized by functional category (page 2) and classification (page 3), including margin and
unit cost analysis at current and proposed rates.
Part 2 is the cost of service calculation from the spreadsheet called "Assign" showing
the functionalization, classification, and allocation of each line item in the study. The
supporting schedules required to run the model are shown on pages 28 through 44.
Finally, Part 3 is the spreadsheet called "Proforma.This worksheet shows the
segregation of Mr. Falkner s pro-forma results of operations into the detailed accounting data
used in this study.
Knox, Di
A vista Corporation
When was the last time the Company filed a natural gas cost of service
study with the Idaho Public Utilities Commission?
The last natural gas cost of service study was filed with Case No. WWP-88-
5. Purchased gas cost allocations have been examined and approved by the Commission with
each periodic purchased gas tracker filing that has occurred in the interim. Distribution base
rates have not been adjusted since February of 1990.
Does the Natural Gas Base Case cost of service study utilize the
methodology from A vista's last Idaho natural gas case?
No. The Base Case cost of service methodology for distribution, customer
services, and administrative & general costs is based on the most recent methodology
employed by A vista in the Washington jurisdiction.This methodology, accepted in
Washington since 1994, resulted from a fully litigated cost of service case specifically
intended to determine appropriate natural gas distribution rates in the era of transportation
service (WUTC Docket No. UG-940814). The result was a compromise methodology
accepting ideas promulgated by Washington Natural Gas Company (now Puget Sound
Energy), the Commission Staff, and Public Counsel.
Purchased gas costs reflect the Idaho accepted methodology that has evolved with
industry changes through periodic tracker filings.Underground storage costs use the
allocation method required by the Idaho Commission order in Case No. WWP-88-
What are the key elements that define this methodology?
Natural gas main investment has been segregated into large and small mains.
Large usage customers that take service from large mains do not receive an allocation of
Knox, Di
A vista Corporation
small mains. Meter installation and services investment is allocated by number of customers
weighted by the relative current cost of those items. System facilities that serve all customers
are classified by the peak and average ratio that reflects the system load factor, then allocated
by coincident peak demand and throughput, respectively. Demand side management costs
are treated in the same way as system facilities. General plant is allocated by the sum of all
other plant. Administrative & general expenses are segregated into labor related, plant
related, revenue related, and "other . The costs are then allocated by factors associated with
labor, plant in service, or revenue, respectively. The "other" A&G amounts get a combined
allocation that is one-half based on O&M expenses and one-half based on throughput. A
detailed description of the methodology is included in Exhibit No. 17, Schedule 4.
What are the results of the Company s natural gas cost of service study?
The following table shows the rate of return and relative return ratio at present
rates for each rate schedule:
Table 3
Customer Class
Residential Service Schedule 101
Rate of Return Return Ratio
76%
04%1.21
27%1.25
44%1.49
88%1.57
00%1.00
Small Firm Service Schedule 111
Large Firm Service Schedule 121
Interruptible Service Schedule 131
Transportation Service Schedule 146
Total Idaho Natural Gas System
These results indicate that Schedule 101 is currently earning slightly less than the
overall return. The other schedules are earning more than the overall return by varying
Knox, Di
A vista Corporation
degrees. The summary results of this study were provided to Mr. Hirschkorn as an input into
development of the proposed rates.
Does this conclude your pre-filed direct testimony?
Yes.
Knox, Di
A vista Corporation