20170215Gorman Rebuttal.pdf

Cuao M. Srores
-*l*--*
CABLE HUSTON,,,
.1! a --h.'il I i'rlj
rlitllilVED
15 ff? 9:hl
s!0i,,r
[e-mail]
February 14,2017
VIA FEDERAL EXPRESS
Diane Hanian
Commission Secretary
Idaho Public Utilities Commission
472 W . Washington Street
Boise,lD 83702
Northwest Industrial Gas Users' Testimony and Exhibits
Case No. INT-G-16-02
Dear Ms. Hanian:
Enclosed for filing with the Commission please find ten copies of the Rebuttal Testimony
and Exhibits on behalf of Northwest Industrial Gas Users. Please note that one copy of the
Rebuttal Testimony and Exhibits has been designated as a Reporter's Copy.
Please let me know if you have any questions. Thank you.
V y yours,
Chad M. Stokes
CMS/sk
Enclosures
cc: Service List via E-Mail
26678. 885\483 5 -237 0 -617 9.v I
Re
Suite 2000, 1 001 SW Fifth Avenue, Portland, 0regon 97204-1 1 36 r Phone: [phone] r Fax: 503.224,31 76 . wwwcablehuston.com
CERTIFICATE OF SERVICE
I CERTIFY that on this l4th day of February 2017, I served the foregoing document upon
all parties of record in this proceeding via electronic mail and/or by mailing a copy properly
addressed and first class postage prepaid.
Peter J. Richardson
Gregory M. Adams
Richardson Adams, PLLC
515 N 27th Street
Boise,ID 83702
peter@ richardsonandoleary.com
gre g@richardsonandoleary. com
Ronald L. Williams
Williams Bradbury, P.C.
1015 W. Hays Street
Boise, ID 83702
ron@wi I liamsbradbury. com
Benjamin Otto
Idaho Conservation League
710 N 6th Street
Boise, ID 83702
botto@ idahoconservation.org
Ken Miller
Snake River Alliance
P.O. Box 1731
Boise,ID 83701
kmiller@ snakeriveralliance. org
Michael P. McGrath
Director, Regulatory Affairs
Intermountain Gas Company
P O Box 7608
Boise,ID 83707
Mike.mc grather@ intergas.com
Scott Dale Blickenstaff
Amalgamated Sugar Co LLC
1951 S Satum Way Ste 100
Boise, ID 83702
sblickenstaff@ amal su gar. com
F. Diego Rivas
NW Energy Coalition
1101 8th Avenue
Helena, MT 59601
[e-mail]
Karl Klein
Sean Costello
Idaho Public Utilities Commission
PO Box 83720
Boise, ID 83720-0074
[e-mail]. gov
Sean. costello@nuc. idaho. gov
26678. 885\48 I 4-6993 -9523.v I
Brad M. Purdy
2019 N 17th Street
Boise, lD 93702
[e-mail]
Andrew J. Unsicker
Lanny L. Zieman
Natalie A. Cepak
Thomas A. Jernigan
Ebony M. Payton
AFLOA/JA.ULFSC
139 Barnes Drive, Suite 1
Tyndall, AFB FL 32403
Andrew. unsicker@ us. af. mil
Lanny.zieman. 1 @us.af.mkil
Natalie. cepak. 2@us. af.mil
Thomas j erni gan. 3 @us.af.mil
Ebony.payton. ctr@us. af. mi I
Michael C. Creamer
Givens Pursley
mcc@ eivensoursley. com
Dated in Portland, Oregon, this 14fr day of ,2017.
Chad M. Stokes, OSB No. 004007
Tommy A. Brooks, OSB No. 076071
Cable Huston LLP
1001 SW Fifth Ave., Suite 2000
Portland, OR 97204-1 136
Telephone: [phone]
Facsimile: [phone]E-Mail: [e-mail]
tbrooks@cabl ehuston. com
Of Attorneys for the
Northwest Industrial Gas Users
26678.885\48 l 4-6993 -9 523.v t
iil; il i: IVE D
r,r il i:Li l 5 filt g: t+ l
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
I
,,ILit
t,
$si0l{
IN THE MATTER OF THE
APPLICATION OF
INTERMOUNTAIN GAS COMPANY
FOR THE AUTHORITY TO
CHANGE ITS RATES AND
CHARGES FOR NATURAL GAS
SERVICE TO NATURAL GAS
CUSTOMERS IN THE STATE OF
IDAHO
Case No. INT-G-I6-02
)
)
)
)
)
)
)
)
)
)
)
RebuttalTestimony of
Michael P. Gorman
On behalf of
Northwest Industrial Gas Users
February 15,2017
Bnunnren & fusocnrr$ lNc.
Project 10309
Table of Contents to the
Rebuttal Testimonv of Michael P. Gorman
Page
3I. COST OF SERVICE STUDY
LA. Load Study.....,...
l.B. Customer-Related Costs
l.C. Distribution Main Costs (Account 376) - Demand and Customer
l. D. Exclusion of Customer Classes ................
II. REVENUE ALLOCATION
III. PROPOSED RATE DESIGN
Gorman, Di-Reb TOC
Northwest lndustrial Gas Userc
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
o
A
PLEASE STATE YOUR NAME AND BUSINESS ADDRESS.
Michael P. Gorman. My business address is 16690 Swingley Ridge Road,
Suite 140, Chesterfield, MO 63017.
ARE YOU THE SAME MICHAEL P. GORMAN WHO PREVIOUSLY FILED
TESTIMONY IN THIS PROCEEDING?
Yes. On December 16, 2016, I filed Direct Testimony on behalf of Northwest
lndustrial Gas Users (.NWlGU').
WHAT IS THE PURPOSE OF YOUR REBUTTAL TESTIMONY?
I will respond to the Direct Testimony of ldaho Public Utilities Commission Staff
witnesses Michael Morrison and Bentley Erdwurm. I will respond to Mr.
Morrison's positions on lntermountain Gas Company's ('lGC' or "Company") cost
of service methodology, and his proposed allocation of the revenue increase
among the Company's rate classes. I will respond to Staff witness Erdwurm's
proposed Firm Transportation rate design. I will also address some of the
testimony presented by Amalgamated Sugar.
PLEASE SUMMARIZE YOUR CONCLUSIONS AND FINDINGS AS LAID OUT
IN YOUR REBUTTAL TESTIMONY.
I take issue with Staff witness Morrison's conclusion that the Company's class
cost of service study ("COSS") is unreliable. The Company did develop
allocation factors based on actual load studies of the Company's customers. The
Company relied on the best information available to measure each class's
contribution to the system coincident peak and non-coincident peak.
Mr. Morrison proposes alternative methods which would provide improved
estimates of the class contribution to these allocation factors, but notes the data
needed to complete the allocation factors as he prefers is not available.
Gorman, Di-Reb 1
Northwest Industrial Gas Userc
o
A
o
A
o
A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
1B
19
20
I also take issue with Staffs proposed phase-in of the Company's
proposed demand-based pricing for its Large Volume and Firm Transportation
customers. The Company's pricing is consistent with its cost of service, which is
the best estimate available in the record of the Company's cost of providing
service to these large customers. Further, this pricing based on cost of service
will provide efficient price signals to these customers to make efficient
consumption decisions, and allow them to make economic conservation-related
changes in operations or investments in energy assets in a means that allows
them to manage their bills while placing more efficient demands on the Company
system. The superior pricing signals created through the Company's proposed
pricing structure is reasonable, and should be adopted.
I also comment on the need for mitigations for customers that
demonstrate that the change to demand pricing can create difficulties for those
customers. The Commission should investigate whether such customers that
are detrimentally impacted by a demand-based Firm Transportation rate can
alternatively choose lnterruptible Transportation service, or should be considered
for a load retention rate. lf a load retention rate option is selected, the
Commission should require all customers to make contribution to costs not
covered by the load retention rate because retaining the customers would benefit
the system as a whole.
Gorman, Di-Reb 2
Northwest lndustrial Gas Users
1
2
3
4
5
o
7I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
o
A
a
A
I. COST OF SERVICE STUDY
PLEASE SUMMARIZE MR. MORRISON'S POSITIONS ON THE COMPANY'S
COST OF SERVICE METHODOLOGY AND HIS PROPOSED REVENUE
SPREAD.
Mr. Morrison argues that IGC's COSS does not produce a fair allocation of costs
to IGC's rate classes for the following reasons:
1. IGC did not conduct a load study and has not produced a fair allocation of
capacity costs.
2. IGC's distribution service costs (FERC Accounts 380 through 385) are
properly regarded as customer-related, but were not allocated across rate
classes reasonably by IGC (Morrison Direct Testimony at 9 and 10).
3. IGC's proposal to classify distribution costs in Account 376 as both demand
and customer-related is inappropriate (/d. at 11 and 12).
4. IGC's COSS should not have excluded two lnterruptible Snow Melt classes
(lS-R and lS-C) (/d. at 16).
DID MR. MORRISON PROPOSE CHANGES TO CORRECT THE COMPANY'S
CLASS COSS BASED ON THESE GONCERNS?
No. Mr. Morrison concluded that without a load study, it is impossible to develop
a suitable alternative cost of service methodology. Therefore, he recommends
that the revenue requirement be allocated in proportion to the normalized
revenue currently being collected from each customer class, Also, Mr. Morrison
recommends that lGC, Commission Staff and other stakeholders hold workshops
to develop a load study and cost of service methodology after this rate case (/d.
at 3).
IS MR. MORRISON'S PROPOSAL FOR AN EQUAL PERCENT ALLOCATION
ON NORMALIZED REVENUES REASONABLE?
No. The Company's current rates were developed 30 years ago and much has
changed in the industry and on the Company's system since that time. To
Gorman, Di-Reb 3
Northwest lndustrial Gas Users
o
A
1
2
3
4
5
o
7
I
9
'10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
continue the rate spread that has existed for such a long period of time is
untenable unless that rate spread is justified by a class COSS. ln contrast, the
Company's class COSS is based on the best, most recent information available
to the Company, Commission Staff, and other parties to develop a class COSS in
this proceeding. While certain arguments made by Mr. Morrison have some
merit, they are not adequate to abandon the Company's COSS which is, again,
based on the best information available in the record. For these reasons, I
recommend Mr. Morrison's proposed equal percent allocation on normal
revenues across all rate classes be rejected,
It is important to adjust rates toward cost of service for several purposes
including: (1) to send accurate price signals forefficient use and conservation of
gas; (2) to allow customers to implement effective utility management initiatives;
and (3) to equitably adjust rates across rate classes in this proceeding. Mr.
Morrison's proposal fails to meet these objectives.
Accurate price signals are particularly important in this case, as outlined
by the proposed demand-side management and conservation efforts undertaken
by IGC and described in the direct testimony of IGC witness Allison Spector.
Effectively implementing efficient and accurate price signals that reflect cost of
service better supports economic demand-side management and conservation
programs which do not create uneconomic impacts on either the Company or
customers. That is, if prices are set based on cost of service, then incremental
revenue loss from demand-side management and conservation efforts should
more closely align with reductions in incremental utility cost. NWIGU witness Ed
Finklea is separately providing testimony addressing price signal issues.
Gorman, Di-Reb 4
Northwest lndustrial Gas Users
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
o
A
For these reasons, the Company's efforts to move rates toward cost of
service should be approved by the Commission, and encouraged in this and
future cases.
O WILL YOU RESPOND TO THE COSS DEFICIENCIES IDENTIFIED BY MR.
MORRISON?
A Yes. I respond to each of Mr. Morrison's concerns related to the Company's
class COSS, and I show that while his overview of the COSS is reasonable, his
criticisms of the Company's COSS are overly critical, and do not diminish the
usefulness of the Company's class COSS nor provide a sufficient basis for
utilizing a different methodology.
l.A. Load Study
PLEASE RESPOND TO MR. MORRISON'S CONCERNS WITH THE
COMPANY'S CLASS COSS.
Mr. Morrison argues that the Company's class COSS is not a reasonable
methodology of allocating costs across rate classes because the allocators are
not based on a load study.
WHAT DOES MR. MORRISON MEAN BY LOAD STUDY?
He states that a load study determines peak usage by class which can then be
used to allocate common equipment costs across rate classes that cannot
otherwise be directly allocated. (Morrison Direct Testimony at 4, lines 16-18).
He further states that a load study generally is used to determine coincident peak
('CP') allocators and non-coincident peak ("NCP") allocators. He opines that CP
allocators generally are used for larger transmission and storage costs, and NCP
allocators are more appropriate for capacity-related portions of distribution plant
Gorman, Di-Reb 5
Northwest lndustrial Gas Users
o
A
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
o
A
which are generally sized for specific customers connected to distribution circuits
rather than the system as a whole. (/d. at 5-6).
WHY DOES MR. MORRISON BELIEVE THAT THE COMPANY'S COSS DID
NOT MEASURE CAPACITY ALLOCATION FACTORS REASONABLY?
He states at page 7 that the Company's development of capacity allocators was
based on both monthly billing data for certain classes, and daily peak data for
other classes. He states at page 7, lines 8-21 that the Company combined peak
day information from its approximately 150 lndustrial and Transportation
customers who are equipped with meters capable of recording daily demand,
with the monthly billing information from approximately 340,000 Residential and
General Service ("GS') customers who are not equipped with metering
equipment that can measure daily demands.
Mr. Morrison states that the Company subtracted the peak usage of the
lndustrial and Transpo(ation customers from the system peak, and the amount
of peak that was not serving these two customer groups was then allocated to
the Residential and GS classes. The allocation of the peak daily usage for the
Residential and GS classes was then based on their January monthly
consumption data.
DO YOU BELIEVE THAT THE COMPANY'S METHOD OF MEASURING EAGH
CLASS'S CONTRIBUTION TO THE SYSTEM PEAK IS UNRELIABLE AS
MR. MORRISON SUGGESTS?
No. I believe the Company's methodology is based on the best information
available, and has produced the most accurate description of class contributions
to the system peak that has been presented in the record. Further, I believe this
methodology is generally consistent with industry practices and produces a
reasonable method of allocating peak day capacity costs.
Gorman, Di-Reb 6
Northwest lndustrial Gas Users
A
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
lmportantly, the methodology accurately measures the system daily peak
and the peak for the Transportation and lndustrial customer classes. The only
limitation, if any, in measuring contributions to peak day demand in the
Company's study is its estimate for the Residential and GS classes' peak day
demand.
As Mr. Morrison properly notes, however, it is not possible to measure the
Residential and GS classes' contribution to the system CP demand directly.
These classes do not have the metering equipment that allows for daily
consumption measurement. (Morrison Direct Testimony at 7, lines 13-14). As a
result, in order to measure each class's contribution to IGC's CP, assumptions
would have to be made for the Residential and GS classes. Therefore, Staffs
concern with the Company's COSS simply cannot be cured in this proceeding.
However, and more importantly, the Company's limitation on metering for the
Residential and GS classes is not unique, but it is generally the standard across
the industry. Therefore, the Company's efforts to utilize existing infrastructure to
measure each class's contribution to CP is generally consistent with industry
practice, and is reliable,
IF THE RESIDENTIAL AND GS CLASSES DO NOT HAVE METERS
CAPABLE OF MEASURING THEIR CONTRIBUTION TO CP, DOES THAT
MEAN IT IS NOT POSSIBLE TO ACCURATELY ALLOCATE CAPACITY
COSTS TO THESE CLASSES?
No. ln my experience, gas utilities' service to Residential and GS customers is
normally based on meters that are not capable of measuring daily consumption.
lndeed, measuring daily consumption requires a more sophisticated and more
expensive meter. Therefore, it is normal for gas utilities to estimate these
classes' contribution to the CP of the system.
Gorman, Di-Reb 7
Northwest lndustria! Gas Users
A
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
'16
17
18
19
20
21
22
23
24
25
26
o
However, because these classes have predictable weather-sensitive load
characteristics, utilities can project usage using weather data for historical
periods reasonably well. I believe these estimates are reliable.
SHOULD THE COMPANY HAVE INSTALLED METERING EQUIPMENT THAT
WOULD BE CAPABLE OF MEASURING DAILY CONTRIBUTIONS TO THE
SYSTEM DEMAND FOR THE RESIDENTIAL AND GS CLASSES SO IT
COULD COMPLETE A LOAD STUDY AS CONTEMPLATED BY
MR. MORRISON?
Only if such metering investment would be found to be a prudent investment and
economically justified. lnstalling meters for the purpose of doing a class COSS
simply would be unlikely to meet this standard unless done on only a subset or
sample of these smaller customers. lnstalling a subset of this equipment would
allow for a statistical measurement of classes' CP. However, alternative
statistical evaluations can be performed using monthly sales data. The
Company's methodology relies on existing infrastructure and metering equipment
to make these approximations in a similar but different manner than a statistical
test of specific customers on the system. Again, the Company's methodology is
reasonable.
Mr. Morrison states at page 8, lines 17-22 of his testimony, that the
Company is replacing Encoder Receiver Transmitter ("ERT") meters with ERT
meters capable of recording hourly consumption information for each of its
customers. He states that a relatively small number of these meters could be
used to obtain the peak information needed to develop accurate CP and NCP
allocators. Residential and GS customers generally do not monitor their gas
consumption on a daily basis. Rather, they monitor and manage their monthly
gas bills. Generally speaking, these classes do not need more detailed interval
Gorman, Di-Reb I
Northwest Industrial Gas Userc
A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
o
A
data than monthly consumption. More specifically, however, metering equipment
for these smaller customers that is capable of measuring daily consumption
levels generally is far more expensive than the existing metering equipment that
is not capable of this measurement detail.
DO YOU BELIEVE THAT THE COMPANY'S METHOD OF DEVELOPING
COINCIDENT DEMAND ALLOCATORS IS GENERALLY CONSISTENT WITH
INDUSTRY PRACTICE?
Yes. lt is normal for utility companies not to have more sophisticated metering
equipment for smaller customers such as GS and Residential customers. For
these weather-sensitive customers, the Company's practice of approximating
their contribution to system peak through monthly volumes is a well regarded and
normal methodology for measuring coincident demand allocators. This is
recognized by the National Association of Regulatory Utility Commissioners
('NARUC',).
The NARUC Gas Distribution Rate Design Manual ('NARUC Manual")
recognizes that installing meters capable of daily usage can be very costly. ln its
Manual, NARUC indicates that instead for small customers' classes, CPs can be
measured as follows,
However, since system peaks in the gas industry are
highly weather sensitive, a fairly reliable correlation
between temperature versus gas consumption can be
developed from utility records. By applying a least square
fit to "average degree day" and "use per day" data for each
customer group, one can calculate with reasonable
accuracy the demands to be placed on the system.
19
20
21
22
23
24
25
26
27
28
29
30
b. Determination of Load Curves Bv Billinq Records
Load curves can be determined for some classes from the
billino records of customers who arc equipped with
standard recordinq instruments. This is feasible for
Gorman, Di-Reb 9
Northwest lndustria! Gas Users
1
2
3
4
classes in which all, or nearly all, the customers are so
equipped. Normally, this is the case for interruptible and
large industrial customers, a tiny fraction of all customers
served by a utility.l
DO YOU HAVE ANY CONCERNS WITH MR. MORRISON'S PROPOSAL FOR
WORKSHOPS FOLLOWING THIS RATE CASE TO DISCUSS THE
IMPORTANCE OF ACCURATELY MEASURING CP AND NCP?
No. I believe workshops would be beneficial if all stakeholders are able to
participate in the process. However, that does not take away from the fact that
the Company's class COSS is the best information available in this record and is
reasonable to use to allocate IGC's cost of service across rate classes consistent
with the best information on cost causation.
DOES MR. MORRISON TAKE ISSUE WITH ANY OTHER ASPECTS OF THE
LOAD DATA USED TO MEASURE PEAK DAY ALLOCATORS?
Yes. At page 8 of his testimony, he notes that the Company measured its peak
day on January 1,2016. He states that this is a holiday. He opines that holiday
usage may not reflect normalized load characteristics of all the customers on the
system.
PLEASE RESPOND.
I generally agree with Mr. Morrison that the Company can consider a normalized
review of load characteristics on the system in measuring CP and NCP. As long
as the data used in the study is generally accepted, and generally reflects the
load characteristics on the system, then I would agree. However, no witness in
this proceeding has challenged the Company's findings on the reasonableness
and accuracy of its CP allocation factors based on the actual load data used by
the Company.
'N{RUC Gas Distibution Rate Design Manual, June 1989, pages 28-29, emphasis
Gorman, Di-Reb 10
Northwest lndustrial Gas Users
5
6
7
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
A
o
A
o
A
added.
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
o
A
o
A
l.B. Customer-Related Costs
DOES MR. MORRISON TAKE ISSUE WITH THE COMPANY'S DISTRIBUTION
SERVICES AND THE GOMPANY'S CLASSIFICATION OF THEM AS
CUSTOMER.RELATED?
ln part. Mr. Morrison does not dispute that costs associated with FERC Accounts
380 through 385 are properly classified as customer-related, However, he
disagrees with the Company's method of allocating these costs across
customers based on a weighted meter method. lnstead, he believes that these
costs should be directly assigned across rate classes.
IS THE COMPANY'S USE OF A WEIGHTED METER ALLOCATION OF
CUSTOMER.RELATED COSTS REASONABLY CONSISTENT WITH
ACCEPTED INDUSTRY PRACT!CE?
Yes. Again, the NARUC Manualsupports a weighted meter customer allocation
of costs. The NARUC Manualstates as follows:
a. Customer Costs
Customer costs may be distributed in proportion to the number of
customers in a class, or a more detailed study may be made
whereby certain components of the customer costs may be
distributed on a per-customer basis, directly assigned or
distributed on a weighted per-customer basis. The latter method
permits recognition of known or ascertainable customer cost
differences such as the frequency of meter readings, complexity in
obtaining readings or integrating meter reading charts, and the
individiual [sic] attention which may be given to large customers,
such as separate meter reading schedules. (NARUC Manual,
page 24.)
DO YOU AGREE WITH MR. MORRISON'S CRITICISM?
No. Again, I think he is being too critical of the Company's methodology. Absent
the accounting records necessary to allocate costs across rate classes as
Gorman, Di-Reb 11
Northwest lndustrial Gas Users
o
A
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
16
17
1B
19
20
21
22
23
24
25
26
o
A
Mr. Morrison proposed, the best methodology is to use the weighted metering
methodology used by the Company in this proceeding.
Mr, Morrison acknowledges that the Company does not have the
accounting records available to directly allocate these costs across rate classes
as he proposes. (Page 10, lines 13-16). Again, the Company's class COSS
reflects the best information available to develop an accurate measurement of
the Company's total system cost, and allocate that cost across rate classes
based on cost causation.
DID MR. MORRISON OFFER AN ALTERNATIVE TO A WEIGHTED METER
ALLOCATION OF THESE CUSTOMER.RELATED COSTS?
He offered a concept as an alternative to the weighted meters. At page 11 of his
testimony, he suggested that absent accounting data necessary for a direct
allocation of these costs, a better methodology for allocating distribution service
costs would be based on the relative cost of installing distribution services for
each class. He states that he would expect "the costs of regulators to be
allocated in proportion to the costs of regulators used by each class," and "the
costs of ERT devices to be allocated in proportion to the costs of ERT devices
used by each class, and so-on." He believes that the information required to
create these allocators is readily available through the system the Company uses
to estimate line extension costs. (Lines 7-18).
IS MR. MORRISON'S ALTERNATIVE METHOD OF ALLOCAT!NG
DISTRIBUTION SERVICE COSTS REASONABLE?
The concept is generally reasonable, yes. However, Mr. Morrison does not
attempt to actually show the difference between the results that would result from
his method and the Company's results. As Mr. Morrison notes, the information is
readily available, and it could have been used to challenge the Company's
Gorman, Di-Reb 12
Northwest lndustria! Gas Users
o
A
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
'19
20
21
22
23
24
weighted meter allocation method. However, no information has been offered
that disputes the accuracy of the Company's derived customer allocators.
As noted above, the Company's practice is generally consistent with
industry practices. Mr. Morrison's concept certainly has merit, but he has not
shown that it would produce allocators that are materially different than the
Company's allocators. While Mr. Morrison's critique does not justify rejecting the
Company's method in this case, it could be considered by the Company as an
alternative method of allocating distribution costs in the next rate case. As such,
if the Commission desires additional information on developing customer
allocation factors, then the Company should be directed to provide both its
weighted meter allocation methodology, and Mr, Morrison's proposed
incremental cost allocation methodology consistent with the Company's line
extension policies. After a review of the two alternative analyses in the next rate
case, the Commission can consider evidence on which one produces a more
reasonable allocation.
l.C. Distribution Main Gosts (Account 376) - Demand and Customer
O DID MR. MORRISON TAKE EXCEPTION TO THE COMPANY'S
CLASSIFICATION OF DISTRIBUTION SERVICE COSTS ON BOTH DEMAND
AND CUSTOMER?
A Yes. Mr. Morrison takes issue with the Company's proposal to classiff 47.160/o
of its distribution mains (FERC Account 376) as customer-related, and the
remaining 52.84o/o as demand-related. The customer-related portion of
distribution mains is allocated on the number of customers in each class, and the
demand-related portion is allocated on each class'single CP demand.
Gorman, Di-Reb 13
Northwest lndustrial Gas Userc
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
o
A
o
A
o
A
WHY DOES MR. MORRISON TAKE ISSUE WITH THE COMPANY'S
PROPOSAL TO CLASSIFY A PORTION OF DISTRIBUTION MAINS AS
CUSTOMER RELATED?
At page 12 of his testimony, Mr. Morrison argues that the customer-related
classification of plant should be limited to the incremental costs that can be
identified with individual customers. He further argues that because distribution
mains serve multiple customers, it is not appropriate to classifiT any portion of
distribution mains as customer-related. He also argues that the cost of
connecting customers to the system is already captured in the distribution
services (FERC Accounts 380 through 385) which are properly classified as
customer-related.
DO YOU AGREE WITH MR. MORRISON'S ASSESSMENT?
No. I disagree with Mr. Morrison's position that it is not reasonable to classify a
portion of distribution main costs contained in FERC Account 376 as customer-
related.
WHY WOULD SMALLER DISTRIBUTION MAINS BE PROPERLY
CLASSIFIED AS BOTH CUSTOMER AND DEMAND RELATED?
The Company designs and incurs costs for its distribution mains to both meet the
peak day demands of the customers connected to that system, and to have
adequate length of distribution main in order to connect all customers to the
system. As such, distribution main costs are driven by both number of customers
and their locations, as well as the coincident demands of the customers
connected to the distribution main.
Gorman, Di-Reb 14
Northwest lndustrial Gas Users
1
2
3
4
5
6
7
I
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
A
CAN YOU PROVIDE AN EXAMPLE THAT ILLUSTRATES WHY LENGTH OF
MAIN VARIES BY CUSTOMERS AND NOT ONLY COMBINED PEAK DAY
DEMANDS?
Yes. Consider an example where the Company has two customers with the
same peak day demand connected to a distribution loop, and the customers are
two miles apart from each other. For this distribution loop, IGC would need to
install two miles of distribution main, that have adequate capacity to meet the
peak day demands of the two customers. The length of main is determined by
the geographic distance between the two customers.
Now assume IGC has another portion of its distribution system where
again there are two customers with the same peak day demand but they are
10 miles apart. For this distribution loop, the Company would need to install
10 miles of distribution main to connect these customers to the distribution
system, again using main sized to meet the combined peak day demands of the
two customers.
While in each of the two distribution loops, the mains are sized to meet
the combined peak day demand, the second loop would require considerably
more distribution investment because it will require five times greater length of
distribution main to connect the customers to the system. This length of main is
not driven by the customer's peak demand but, rather, is driven by the customer
location and length of main needed to connect the customer to the system. As
such, the number of customers and the location of customers on this distribution
loop are important engineering design features, as well as cost-causation bases
for determining the utility's cost of providing distribution service to all customers.
ln this example, IGC designs its distribution system both to meet the peak
day demands of the customers on its distribution loops, and to have adequate
Gorman, Di-Reb 15
Northwest lndustrial Gas Userc
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
A
length of main to connect all customers to its distribution system. Hence, the
cost-causation factors, or the engineering design parameters, reflect both peak
day demands of the customers and length of main needed to connect the
customers to the system.
DOES MR. MORRISON TAKE ISSUE WITH THE COMPANY'S ZERO
MINIMUM INTERCEPT METHODOLOGY TO ARRIVE AT THE CUSTOMER
DEMAND SPLIT FOR ITS DISTRIBUTION MAIN COSTS?
Yes. At pages 13 and 14 of his testimony, Mr. Morrison states that the Company
should not have included the minimum sized pipe in this study, but rather should
have included a capacity component as a regression modeling factor. He opines
that the different sizes in pipe throughout the system should have been a factor
in the regression study. He opines that the Company's use of a nominal pipe
diameter without a capacity factor in its regression model is a concern.
PLEASE RESPOND TO MR. MORRISON'S CONCERN WITH THE
COMPANY'S MINIMUM INTERCEPT METHOD.
I do not agree with Mr. Morrison that a capacity component should have been a
factor in the minimum size study. ln my experience, the minimum pipe size is a
proper factor used to estimate the zero capacity intercept point for determining
appropriate classification of distribution costs as capacity and customer-related.
The purpose of the study is to identifo how much distribution main cost
the Company would incur regardless of the capacity demands of the customers
on the system. lt is this portion of distribution main costs that should then be
classified as customer-related. All other distribution main costs should be
classified as demand-related. For these reasons, I believe the Company's
methodology to determine the split in classification for distribution mains cost
between customer and demand is reasonable.
Gorman, Di-Reb 16
Northwest lndustrial Gas Users
a
A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
'15
16
17
18
19
o
A
o
A
ARE THERE ADDITIONAL BENEFITS TO THE SYSTEM TO CLASSIFY
DISTRIBUTION MAIN COSTS BETWEEN CUSTOMER AND DEMAND?
Yes. lnterruptible customers receive service from IGC throughout most of the
year, but they are subject to interruption on IGC's peak days. Without a
customer component of distribution mains, interruptible customers would not be
obligated to pay for any portion of the Company's distribution main costs,
because they do not contribute to IGC's peak day demand. Under IGC's
proposal, interruptible customers would pay a customer allocation component of
the distribution system, which represents its costs for connecting interruptible
customers to the system.
IS IGC'S COST OF SERVICE PRACTICE OF ALLOCATING DISTRIBUTION
MAIN COSTS ON CUSTOMER AND DEMAND CONSISTENT WITH
INDUSTRY PRACTICE?
Yes. As described in my direct testimony, NARUC recognizes that demand or
capacity-related costs can be allocated on both peak day demand and the
number of the customers.
Also, in a recent annual update on gas rate structures, the American Gas
Association ('AGA') stated the following on classifying a portion of distribution
mains as a cost of customer component.
The largest part of a natural gas customer's bill is the cost of the
gas itself, over which the utility has little control. This cost
accounts for about 41 cents of every dollar of revenue received by
a distribution utility.[footnote omitted] The bill amount for the gas
portion varies with price as well as amount consumed. Natural
gas utilities also incur costs that are not dependent on a
customer's consumption. These "fixed" costs may include:. Meter readingo Billing. Fixed costs on plant and equipmento Depreciation and taxeso Distribution mains, meters, and service lines
Gorman, Di-Reb 17
Northwest !ndustrial Gas Userc
20
21
22
23
24
25
26
27
28
29
30
31
1
2
3
4
5
6
7
o Most administrative and general expenseso Wageso Buildings, energy, etc.o Natural gas storage. Customer and service O&M
Most utilities recover at least a portion of these costs through a
fixed charge on a customer's bill.2
DID MR. MORRISON HAVE ANY OTHER COMMENTS RELATED TO
APPROPRIATE ALLOCATIONS CONCERNING DISTRIBUTION MAIN COSTS
IN ACCOUNT 376?
Yes. At pages 6 and 26 of his testimony, Mr, Morrison suggests that rather than
using a customer/demand split to allocate the costs of distribution mains in
Account 376, it may be appropriate to use NCP or peak and average allocators.
While distribution main costs are partially demand related, it is never appropriate
to use a peak and average ("P&A") allocator to allocate these costs.
IS IT REASONABLE TO USE A PEAK AND AVERAGE ALLOCATOR FOR
DISTRIBUTION MAIN COSTS?
No. lt is not appropriate to use a P&A methodology as an appropriate demand
allocation factor for any capacity-related costs, including small distribution mains
in Account 376. Using a P&A allocator distorts the allocation of distribution main
costs to be predominantly based on volumetric use rather than the need for
capacity during the peak day to provide firm service to the customers. The effect
of the P&A is to shift capacity-related costs to higher load factor customers from
lower load factor customers.
lndeed, NARUC recognizes that the P&A methodology is generally used
to mitigate impacts on low load factor customers, rather than to properly allocate
demand-related costs. NARUC describes the P&A methodology as:
2American Gas Associalion Energy Anatysis, "Natural Gas Utility Rate Structure: The
Customer Charge Component - 2015 Update," May 28, 2015.
Gorman, Di-Reb 18
Northwest lndustrial Gas Userc
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
o
A
a
A
1
2
3
4
5
This method reflects a compromise between the coincident and
noncoincident demand methods. . . . This method allocates cost to
all classes of customers and temoers the apportionment of costs
between the high and low load factor customers. (NARUC
Manual al27-28, emphasis added).
NARUC's characterization of the P&A method is in stark contrast to its
description of allocating capacity-related costs based on coincident demands and
non-coincident demands. NARUC characterizes the demand-based allocation of
these factors as apportioning capacity-related costs in proportion to the demands
that customers place on the system. (ld. at27).
Further, Mr. Morrison's own testimony makes it clear that the Company
incurs capacity-related costs based on peak day usage of customer classes. At
page 4 of his testimony he states that "ln general, plant equipment is designed to
meet the maximum load that will be placed on individual pieces of plant
equipment, so costs are caused by the need to meet system peak." The P&A
allocator simply does not reflect class contributions to the system peak and does
not allocate cost based on cost causation.
AT PAGES 15 AND 16 OF MR. MORRISON'S TESTIMONY, HE STATES
CONCERN ABOUT THE COMPANY'S ALLOCATION OF GROSS PLANT
IN.SERVICE. PLEASE RESPOND.
Mr. Morrison states that an allocation in the COSS of gross plant in-service
includes an opportunity to earn a fair rate of return on the Company's investment,
which is only a fraction of the Company's costs. He believes that the Company's
gross plant-in-service introduces factors other than the Company's actual
investment in the Company's allocation methodology.
I do not follow Mr. Morrison's testimony on this point. The Company's
financial statements as far as I can tell are audited financial statements that
follow appropriate Generally Accepted Accounting Principles 1"GAAP") and
Gorman, Di-Reb 19
Northwest lndustrial Gas Users
6
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
o
A
1
2
3
4
5
o
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
regulatory accounting principles. As such, allocating its gross plant in-service
across rate classes for plant that is prudently incurred, and used and useful will
accurately determine the Company's cost of providing service to its rate classes.
l.D. Exclusion of Customer Classes
O AT PAGE 16 OF MR. MORRISON'S TESTIMONY HE TAKES ISSUE WITH
THE COMPANY'S COSS BECAUSE IT EXCLUDED TWO INTERRUPTIBLE
SNOW MELT CLASSES - IS.R AND IS.C. PLEASE RESPOND.
A He states that the Company explained that these classes are relatively small,
and the consumption for these customers was included with Residential and GS
classes. He states the Company also did not provide information related to
Schedule H-1, Ketchum-Sun Valley Area Hook-up Fee.
A DO YOU BELIEVE THE EXCLUSION OF THESE CLASSES RENDERS THE
COMPANY'S COSS UNRELIABLE?
A No. These classes are relatively small, and including them in with the
Residential and GS classes which unlikely had any effect on the allocation.
While I would agree that Mr. Morrison's proposal to separate these interruptible
customers from firm service customers in a class COSS is reasonable, I do not
believe combining these customers in one class will skew the reliability of the
cost study I recommend that the Commission direct the Company to do this in its
next rate case. lncluding them in its COSS as a component of other classes in
this case is not shown to diminish the value or accuracy of the Company's COSS
in this case.
Gorman, Di-Reb 20
Northwest lndustrial Gas Users
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
o
A
o
A
II. REVENUE ALLOCATION
PLEASE DESCRIBE MR. MORRISON'S PROPOSED REVENUE
ALLOCATION.
Mr. Morrison recommends allocating the approved non-gas revenue increase in
proportion to the normalized base rate non-gas revenue currently being collected
from each rate class. This results in an equal percentage increase to all classes
of about 4.2o/o, based on Staffs revenue deficiency.
DO YOU AGREE WITH MR. MORRISON'S PROPOSED REVENUE
ALLOCATION?
No. His proposed allocation is based on his belief that the Company's COSS
unfairly allocates costs among customer classes. As described in my direct
testimony, the Company's proposed class revenue allocation is based on the
results of its class COSS. Since the COSS moves rates towards cost of service,
I agree with the Company's proposal to base its class revenue allocation on the
results of its class COSS.
PLEASE SUMMARIZE YOUR CONCLUSIONS AND RECOMMENDATIONS.
The Company's COSS should be used as the basis for allocating any approved
revenue increase in this case, instead of Mr. Morrison's proposed equal
percentage increase. Distribution main costs included in Account 376 should be
allocated using a customer/demand split as proposed by the Company. IGC
should continue to develop peak day demand allocators based on accepted
industry practice. However, Mr. Morrison's proposal for consideration of more
data, including a normalized assessment of contributions to peak day, should be
investigated in the next rate case to determine if it produces a better estimate of
classes' contribution to CPs and NCPs. Lastly, I am not opposed to Company
Gorman, Di-Reb 21
Northwest lndustrial Gas Users
o
A
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
o
A
and stakeholder workshops to discuss cost of service methodology and the
development of a load study.
III. PROPOSED RATE DESIGN
DOES STAFF COMMENT ON THE COMPANY'S PROPOSED RATE DESIGN
IN THIS PROCEEDING?
Yes. Staff witness Bentley Erdwurm comments on the Company's proposed
revisions to the structure of the Large Volume and Transportation classes' rate
designs. He notes at page 18 of his testimony that the Company proposes a
new charge of Maximum Daily Firm Quantity ('MDFQ') for its LV-1, T-4 and T-5
customers. He also observes that the Company is proposing to combine
Transportation rates T-4 and T-5.
Mr. Erdwurm opines that introducing a demand charge into the
Company's Large Volume and Transportation rates recognizes the Company's
costs to serve these customers are derived in large part by demands they place
on the system. However, he also opines that the Company has not supported
the amount of its proposed MDFQ charge with a COSS. (Erdwurm Direct
Testimony at 18, lines 22-24). Therefore, he concludes that Staff recommends
that the amount of MDFQ charge be addressed at the workshop proposed by
Staff.
ln this case, Staff is willing to move in part toward an MDFQ charge for
the LV-1 and T-4 classes. He states that a demand charge will better match
customers' payments to the cost the Company incurs. Therefore, he
recommends a demand charge of 201/therm per month for nominated MDFQ.
The remaining part of the charges would be based on a four-tiered volumetric
rate structure as shown on his Exhibit No. 1 16, page 2.
Gorman, Di-Reb 22
Northwest lndustrial Gas Users
1
2
3
4
5
6
7
I
I
10
11
12
13
14
15
16
17
18
'19
20
21
22
23
24
25
26
o
A
IS STAFF'S PROPOSED RATE DESIGN FOR LARGE VOLUME AND
TRANSPORTATION GUSTOMERS REASONABLE?
No. For the reasons outlined above, I disagree with Staffs conclusion that the
Company has not provided a reasonable and reliable class COSS. lndeed,
Staffs contention that the Company has not performed a load study is overly
critical and simply does not recognize that the Company actually did perform a
load study using the best load data that was available to the Company. That
same data is generally used by utilities to perform load studies and class cost of
service studies throughout the industry. Staffs proposal for more detailed data is
a worthwhile goal, but the reality is that that meter data desired by Staff is simply
not available for the Residential and General Service classes.
I also agree with Staff witness Erdwurm that it is appropriate to price the
Large Volume and Firm Transportation customers largely on the basis of a
demand charge because that is how the Company incurs costs to serve these
customers. However, the same is also true for the delivery costs for all other
classes. For these reasons, pricing Firm Transportation customers based on
predominantly a demand charge is consistent with cost of service, is consistent
with the Company's cost of providing service to these customers, is balanced
and produces an accurate price signal for these customers.
IS THERE A SIGNIFICANT DIFFERENCE BETWEEN THE AMOUNT OF FIRM
TRANSPORTATION REVENUES RECOVERED ON DEMAND CHARGES AND
ON VOLUME CHARGES BETWEEN THE COMPANY'S AND STAFF'S
PROPOSALS?
Yes, While there appear to be some problems with Mr. Erdwurm's proof of
revenue for the Transportation class, it is clear that he intends to collect about
70% of the delivery revenue for the T-4 class based on volumetric charges. This
Gorman, Di-Reb 23
Northwest lndustrial Gas Userc
o
A
1
2
3
4
5
6
7
8
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
o
is shown on his proof of revenue attached to Exhibit No. 116, page 2 of his direct
testimony. There, he shows demand charge recoveries based on a 20dltherm
per month for MDFQ of approximately $3.1 million out of a total approximately
$10 million collected from the combination of Firm Transportation classes T-4
and T-5.
In significant contrast, under the Company's pricing proposal, the
Company proposes to recover approximately $5.5 million through a demand
charge, out of a class revenue assignment to the combined Firm Transportation
class of $7.6 million. This results in over 70% of the class revenue being
collected in demand charges, which is materially different than the Staffs
proposed 30% of delivery charges. (Direct Testimony of Lori A. Blattner, Exhibit
No. 24, page 2).
Because the Company's rate design comports more accurately with the
undisputed fact that delivery charges for the Firm Transportation group are
primarily based on demands of this group, the Company's proposed pricing
structure produces a far more reliable and accurate price signal to LV-1 and T-4
classes than the pricing structure proposed by Staff. Because accurate pricing
signals encourage conservation and efficient procurement of utility services, I
strongly recommend the Commission approve the Company's proposed pricing
structure for the T-4 customers.
HOW DO YOU RESPOND TO MR. ERDWURM'S CONCERN ABOUT LOW
LOAD FACTOR AND HIGH LOAD FACTOR TRANSPORTATION
CUSTOMERS IF THE RATE IS MODIFIED TO A DEMAND CHARGE ONLY?
I appreciate Mr. Erdwurm's concern about rate impacts on customers. While
there could be impacts on customers based on variance in load factor, neither
Mr. Erdwurm nor the Company has proven that there would be any detrimental
Gorman, Di-Reb 24
Northwest lndustria! Gas Userc
A
1
2
3
4
5
6
7
B
I
10
11
12
13
14
15
16
17
18
19
20
A
o
rate impact on any customer. For these reasons, I believe movement to the
demand-based pricing structure as proposed by the Company is appropriate.
DID THE AMALGAMATED SUGAR COMPANY, LLC ("AMALGAMATED")
WITNESS DR. DON READING COMMENT ON THE POTENTIAL NEGATIVE
IMPACT ON IT AS A RESULT OF A DEMAND.BASED PRICING
STRUCTURE?
Yes. Amalgamated witness Dr. Don Reading stated that the proposed
movement to a demand-based pricing structure would create a significant rate
impact on Amalgamated. As a result, Dr. Reading proposes a five rate case
phase-in to a demand charge rate structure for the T-4 tariff. (Direct Testimony
of Dr. Reading at 13).
PLEASE COMMENT.
I appreciate Amalgamated's concern about impacts on its facilities associated
with more accurately pricing Firm Transportation service based on a demand-
based charge. However, rather than creating a phase-in to this rate structure, I
believe a more balanced and equitable method would be simply to allow
Amalgamated to consider different tariff rate alternatives, or a load retention rate
if the Commission finds one to be in the public interest. With regard to this the
Commission should consider, and Amalgamated could comment or, the
following:
1. Can Amalgamated move to IGC's lnterruptible Transportation rate T-3? This
would continue to provide delivery service priced on a volumetric usage
structure, but service would be subject to interruption. The volumetric rate
structure of T-3 could mitigate the impact on Amalgamated.
2. IGC could consider a seasonal or alternative lnterruptible rate that would be
appropriate for Amalgamated and other low load factor Transportation
customers.
3. To the extent Amalgamated is not capable of taking lnterruptible
Transportation service (T-3), and the Commission believes that it is in the
Gorman, Di-Reb 25
Northwest lndustrial Gas Users
o
A
21
22
23
24
25
26
27
28
29
1
2
3
4
5
6
7
8
public interest to mitigate the impact on Amalgamated in this filing, it could be
appropriate to design an Amalgamated-specific load retention rate. This rate
could produce a phase-in for Amalgamated toward a demand-based pricing
structure on the Firm Transportation T-4 rate. However, during the period of
the phase-in, any discount provided to Amalgamated would be spread across
all rate classes, rather than only require other Firm Transportation customers
in class T-4 to subsidize Amalgamated during the tariff rate structure phase-in
period.
To the extent the Commission believes maintaining Amalgamated on its
Firm Transportation rate is in the public interest, and other customers are better
off retaining Amalgamated on the system as opposed to potentially losing this
customer on the system, then all customers could support the discount provided
to Amalgamated during the phase-in period. ln this manner, all customers would
share equally in any load retention benefits the Commission finds to be in the
public interest and appropriate costs for IGC's customers.
PLEASE SUMMARIZE YOUR POSITION ON RATE DESIGN.
I recommend the Company's proposed rate design for Large Volume and Firm
Transportation customers being set on a demand basis or MDFQ be approved.
The demand rate, however, should be set based on the cost of service the
Commission finds to be appropriate for the LV-1 and T-4 rate classes. To the
extent the Commission finds it is in the public interest to provide a phase-in to
any large Transportation customers that are not able to take lnterruptible service,
then any discounts provided to such customers should be spread over all
customers and not simply other Transportation customers. Load retention efforts
benefit all customers equally, and all customers should share in the cost of load
retention,
DOES THIS CONCLUDE YOUR REBUTTAL TESTIMONY?
Yes, it does.
Gorman, Di-Reb 26
Northwest lndustrial Gas Users
I
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
o
A
o
A
\\d oc\sha ros\prclawdocs\sdw\1 0309\testi mony-bai\3 1 4 1 44. doq