20100929PAC to IIPA 124-133.pdf

~~,.~OUNTAIN
September 28, 2010
Eric L. Olsen ISB# 4811
Raine, Olson Nye, Budge & Baíley, Chared
P.O. Box 1391; 201 E. Center
Pocatello, Idaho 83204-1391
RE: ID PAC-E-10-07
IIPA Data Request (124-133)
rH:('t:í'ü- v'"
r~j
zumSEP 29 AM 10:20 ~~ ~~~hc~:~nù~~i~4~~o
\Dß~i40'
~JTIUT\ES
Please find enclosed Rocky Mountan Power's responses to IIPA Data Requests 124-133.
Provided on the enclosed CD are Attchments IIPA 124, 127, 129 and 132.
If you have any questions, pleae feel free to cal me at [phone].
Sincerely,
:JTu!tl~/~
J. Ted Weston
Manager, Regulation
Enclosure:
cc: Rady Budge/Monsato (C)
Jean Jewell/UC (C)/ 3 copies
Anthony Y anell A (C)
BenOtto/ICL (C)
James R. Smithonsto (C)
Richad Anderson/onsanto (C)
George C. Carer, ITonsto (C)
Dens Peseau/onsto (C)
Gareth R. Kajander/Monsanto (C)
Maurce Brubaker/Monsanto (C)
Brian Collins/onsanto (C)
Michael Gormanonsanto (C)
Kath Iversn/onsanto (C)
Mark Widmer/Monsanto (C)
P AC-E-1 0-07/Rocky Mountain Power '
September 28, 2010
IIP A Data Request 124
nPA Data Request 124
In response to lIP A Request 4 the Company provided hourly border load data for
Idaho as well as for all rate schedules (and special contract customers) for 2008
and 2009. Please provide similar information for each of the other jursdictions
during this same timeframe and for all jurisdictions (including Idaho) for 2004-2007. .
Response to np A Data Request 124
The followig border load information is provided as Attachment lIPA 124. This
data was provided in response to discovery requests in past Idaho general rate
cases. No other information is readily available. The following time periods are
provided:
ApriI2003-March 2004 - from PAC:-E-05-01
January2006-December 2006 -from PAè-E-07-05
Januar 2007-December 2007 - from PAC-E-08-07
Recordholder:
Sponsor:
Scott Thornton
To Be Determined~)1¡.... "LJ'
j' ,'l
r
',)
, ~~
PAC-E-10-07/Rocky Mountain Power
September 28,2010
LIP A Data Request 125
IIPA Data Request 125
Is the border load data supplied in response to lIPA Request 4 at the same "level"
as the Metered Load data on pages 10.13 and 10.14 of Company Exhibit 2, or do
additional losses need to be added? If additional losses need to be added, what
are they?
Response to lIP A Data Request 125
The data are at the same "level" but represent different time periods. The data
presented in the Company's response to lIPA Data Request 4 presents actual
jurisdictional load measurements for both the 2008 and 2009 calendar years. The
data presented in Company Exhibit 2, pages 10.13 and 10.14, represent
jurisdictional load estimates for the forecast year 2010.
Recordholder:
Sponsor:
Scott Thornton
To Be Determined
. ~'- ; ~' ~
r
.1
i. /,
:c. \ ,.
: .~.
PAC-E-10-07/Rocky Mountain Power
September 28, 2010
IIP A Data Request 126
LIP A Data Request 126
The flowchar provided in response to IIP A Request 104 contains an item near the
bottom left that portrays the inclusion/introduction of losses into the sales data.
What loss factors were used for each month for each jurisdiction?
Response to lIP A Data Request 126
Annual line loss factors were applied to each month of sales data with the
following breakdown by jurisdiction:
. Oregon - 10.89%
. Washington - 10.90%
. California - 15.00%
. Uta -7.28%
. Idaho-9.88%
. Wyoming -6.89%
Recordholder:
Sponsor:
Romita Biswas
Pete Eelkema
PAC-E-10-07/Rocky Mountain Power
September 28,2010
LIP A Data Request 127
LIP A Data Request 127
The response to lIP A Request 5 lists "pricing Kwh" data for each class/rate
schedule in Idaho by month for 2009. Please provide similar "pricing Kwh" data
for all jursdictions on a monthly basis from January 2004 through December
2009.
Response to lIP A Data Request 127
The following "pricing kwh" data is provided as Attachment IIPA 127. This data
was provided in response to requests in past Idaho general rate cases. No other
information is readily available. The following time periods are provided:
April2003-March 2004 - from PAC-E-05-01
Janua 2006-December 2006 - from PAC-E-07-05
Januar 2007-December 2007 - from PAC-E-08-07
Recordholder:
Sponsor:
Scott Thornton
To Be Determined
¡ '.~ i,j ¡ " "l, :.) .,'It
!', r:;' :¡~~ ¡," ,"!
;.,1 !
j l:t
:i
" .,1,I., ;,"
P AC-E-1 0-07/Rocky Mountain Power
September 28, 2010
LIP A Data Request 128
IIPA Data Request 128
With respect to Monsanto Request 13-1c, raj please explain how the losses for
each jursdiction were developed and rb) how those values relate to border load
and "pricing Kwh".
Response to lIP A Data Request 128
a. PacifiCorp uses a five year average to develop line loss factors. The years
2005 - 2009 were used since that was the most recent five-year period
available at the time the load forecast was developed. Anual line losses are
determined as jurisdictional load (sales with line losses) minus total retail
sales by jurisdiction divided by total retail sales by jurisdiction, where retail
sales are measured at the customer meter.
b. Please refer to the response to subpar "a" above. Border loads include losses
since they are measured at system input. "Pricing K whs" are measured at the
customer meter.
Recordholder:
Sponsor:
Romita Biswas
Pete Eelkema
,,,:1 \;,
, i.
d.
.J:i;.
.i:-l
PAC-E-10-07/Rocky Mountain Power
September 28,2010
lIPA Data Request 129
LIP A Data Request 129
With respect to Monsanto Request 1J.:1c, please provide the losses for each
jurisdiction on a monthly basis for the five years in question.
Response to lIPA Data Request 129
Please refer to Attchment lIPA 129'for the monthly line losses as requested, and
also refer to the Company's response to IIPA Data Request 126 for the annua line
loss factors that were applied to the monthly forecasted sales. PacifiCorp uses an
anual five year (2005-2009) average line loss factor. Monthly line loss factors
are not used for planing puroses.
Recordholder:
Sponsor:
Romita Biswas
Pete Eelkema
J.i', L..
. I;I:~
~,~~,
~..
~t
i' i'
P AC-E-1 0-07/Rocky Mountain Power
September 28, 2010
LIP A Data Request 130
LIP A Data Request 130
Comparing the total hourly border loads and the hourly sales levels (using energy
loss factors to scale the sales values up) found in response to lIP A Request 4 on a
monthly basis for 2009, it can be calculated that on an average monthly basis for
10 months (excluding June and August) that the scaled up sales figures are
consistently below the border load data. Since the sales figures have been grossed
up for losses as well as adjusted for actual monthly sales, this suggests that either
the border load data is too high or the loss factors are too low. Please comment
on the reason for the general discrepancy that occurs between these two sets of
figures on a monthly basis.
Response to lIP A Data Request 130
It is unealistic to expect the sum of the class loads to approximate the
jursdictional loads with any degree of certinty. The requestor has identified
what would be considered the two primar factors:
Loss factors - Jursdictional loads are directly measured at their point of input
into the system. As such, they are considered to include actual losses. Class
loads are measured at the customer meter level, so do not include losses. Losses
are assigned to the classes based on single, voltage level specific factors
developed and summarzed in the Company's loss factor studies. The same factor
is applied to all 8,760 hours of a given year for a given class/voltage leveL. In any
hour where the measured losses differ from those statically derived losses, the
difference will be reflected in the comparson of the sum of class loads to the
jursdictional measurements. . .', .,',
Sample error - The Company utlÌì~é.s a:cpipbination of census metering and load
studies to estimate class contrbution):o system peak. The load studies are used to
provide estimates for approximately' 90% of tota Idaho load. These load studies
have been designed to meet the PURP A stadard of:: 1 0% accuracy at the 90%
confidence level for the variable of interest. While we expect these load studies to
provide a reasonable estimate, they are not designed to provide an exact
meaurement of actual usage. Any difference between actua usage and the
estate of usage will be reflected .in.thecomparison of jursdictional load
meaurements to class load estimates:."' ..
The Company recognzes that these'~ay!not be the only contrbutors to this
mismatch. Among others considered: ~ .
. Cusomer owned generation-s~all residential and commercial generation
facilties (paricularly wind/solar) are not measured in a time series format,
and so are not reflected in the jursdictional load mix. Nor is this resource
being adequately captued in the class load data.
. Curent diversion - Energy resources lost due to theft are stil included in the
jursdictonal level data but not the customer level (class load) data.
~
. ';\
j ,! :l.
PAC-E-IO-07/Rocky Mountan Power
September 28,2010
LIP A Data Request 130
.. Metering error - Meters are designed to meet strict accuracy requirements, but
they do faiL. Depending on the type of customer, the failure can go unoticed
for a period of time. In these situtions, the jurisdictional level data is
captuing usage, but the customer level data is not, and this difference wil
also contrbute to differences between jurisdictional and class load
measurements.
The recognition that there are differences between jursdictional load
measurements and the sum of class load estimates is not new, or even unique to
this Company. In the California market, for example, these differences are
referred to as Unaccounted for Energy (UFE). In a 2002 Load Research Working
Group report fied with the Utah PUblic Service Commission, the issue was
examined and it was recommended that the practice of calibration be
discontinued. It is believed, for the reasons stated above, that a definitive
identification of the source(s) and level of contribution canot be accurately
determined.
Recordholder:
Sponsor:
Scott Thornton.
To Be Determinedi ~
I'
:1,11',.. '1'
'T/"~ ~
;"'
PAC-E-10-07/Rocky Mountain Power
September 28,2010
LIP A Data Request 131
lIPA Data Request 131
Comparng the total hourly border loads and the hourly sales levels (using energy
loss factors to scale the sales values up) found in response to IIP A Request 4 on a
monthly basis for 2009, it can be calculated that on an average monthly basis for
June and August that the scaled up sales figures are consistently and significantly
above the border load data. Since the sales figures have been grossed up for
losses as well as adjusted for actual monthly sales, this suggests that either the
border load data is too low or something else is taking place-the variation is so
great that loss factors would not have a significant impact. Please comment on
the reason for the general discrepancy that occurs between these two sets of
figues on a monthly basis.
Response to IIPA Data Request 131
Please refer to the Company's response to lIPA Data Request 130.
Recordholder:
Sponsor:
Scott Thornton
To Be Determined
i.
. i.; i-:
. ¡.¡'j ¿;,
I' 1.:
,¡'i
.':
P AC-E-1 0-07/Rocky Mountain Power
September 28, 2010
lIPA Data Request 132
LIP A Data Request 132
J. '
The border load data provided in response to lIP A Request 4 is said to be a
compilation of all energy that flows into and out of Idaho as measured at specific
locations. Please list for each hour öf2008 and 2009 the name of each border
location and the value of energy that either went into the Idaho jurisdiction or left
the Idaho jursdiction as well as all energy that was generated within the
jurisdiction.
Response to lIP A Data Request 132
Please refer to Attachment LIP A 132:
Recordholder:
Sponsor:
Tom Beck
Pete Eelkema
. T
1"; ).
PAC-E-10-07/Rocky Mountain Power
September 28,2010
LIP A Data Request 133
LIP A Data Request 133
Exhibit 2, Tab 1 page 1.3 lists the total Idaho normalized General Business test
year revenue at $202,733,163. Exhibit 50 lists the present Idaho revenue at
$202,733,000 associated with 3,325,872 MWH. Exhibit 49, Tab 5, page 14, line
10, lists test year sales (~sale level) as 3,325,873 MWH which is associated with
3,557,594 MWH on line 14 ~ Input. By:comparison, Exhibit 2, Tab 10, page
10.14 uses a load of 3,628,130 MWH to allocate costs to Idaho. Please explain
why the base Idaho revenue in this case is being based upon 3,557,594 MWH ~
Input, while the allocation of jursdictional costs is based upon 3,628,130 MWH.
Response to lIP A Data Request 133
There are thee reasons for this diffeFence:
1. The Cost of Service (COS) study. only includes the losses associated with
retail sales and differentiates line losses by voltage leveL. Also COS uses line
losses based on an anual analysis of Idaho operations applicable to metered
retal sales. Line losses used in the Jursdictional Allocation Model (JAM)
includes total system line losses, which includes losses associated with
serving both retail and wholesale sales obligations. JAM uses a five year
average system loss factor.
2. Please refer to the Company's response to lIPA Data Request 15. COS results
are presented as if there are no, c#railments or interrptions to Monsanto.'., i:. .,_ ,f
3. The COS study includes the weåihernormalization of Januar 2010 actual
sales. Please refer to the Coinp~y's response to Monsanto Data Request
13.1 d for additional information.
~"C
Recordliolder:
Sponsor:
Pete Eelkema
To Be Determined',I. ...
.t.Vi:;j¡-:
..1
.r~.,
.l
..U' .\ j.' I.i.,
; . :;'" ~..