ML20003F011
| ML20003F011 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 07/31/1980 |
| From: | Vogel T TEXAS, STATE OF |
| To: | |
| Shared Package | |
| ML19240B984 | List:
|
| References | |
| NUDOCS 8104170675 | |
| Download: ML20003F011 (11) | |
Text
_ _
9 DOCKET NO. 3250 APPLICATION OF TEXAS ELECTRIC l
PUBLIC UTILITY COMMISSION SERVICE COMPANY FOR AUTHORITY l
TO CHANGE RATES 1
0F TEXAS DIRECT TESTIMONY OF TED V0 GEL l
ECONOMIC RESEARCH DIVISION PUBLIC UTILITY COMMISSION OF TEXAS l
l l
JULY, 1980 I810.4170(,75 l\\.-,,
....-.c...
.~,,. -.-----.+.- - ---
~~
-- ~ - - - - ~ ~ - - ' ' ' - ~ ~ ~ ' ~ ' ' ~ ' *
- ~~'
' ' '" ' ~ ~ ~ ' ~'
age J of 1 DockCt No.
3250 1
Q.
Please state your name and business address.
2 A.
TedVogel,78d0ShoalCreekBoulevard, Suite 400N, Austin, Texas. 78757.
3 Q.
By whom are you employed and in what capacity?
4 A.
I am employed by the Public. Utility Coemission of Texas as a Senior Financial 5
Analyst and Econometrician in the Economic Research Division.
6 Q.
What are your principal areas of responsibility in this capacity?-
7 A.
I have responsibility for the development and estimation of econometric 8
models of demand, revenues, and other factors concerning public utilities
'9 under the Comission's jurisdiction.
10 Q.
Please state briefly your educational background and professional qualifica-11 tions.
12 A.
I received a B.S.
degree with a major in accountirig from Penn Statt 13 University, State College, Pennsylvania.
After graduation, I studied 14 finance at the University of Southern California in Los Angeles and received
- l 15 a Masters Degree in Business. Upon completion of graduate school, I moved to 16 Little Rock, Arkansas.
There I worked for the Arkansas Public Service 17 Com:ission while studing for a Master's Degree in Public Administration.
i 18 Recently, I was awarded an MPA with a concentration in statistics from 19 the University of Arkansas. Currently, I am taking post-graduate courses at 20 the University of Texas at Austin.
i 21 I am a memoer of the American Finance Association, Financial Management 22 Associatien, American Economic Association, and the Association of Public 23 Administrators. In addition, I have attended several conferences related to 24 public utilities, namely:
(1) the Public Utilities Finance Procram, at the 25 University of Texas at Dallas, and (2) the Public Utility Regulation
1 Docket No.
3250 Page 2 of 8 1
Conference, under the direction of Myron Gorden and Paul Halpern, at the -
2 University of Toronto, Toronto, Canada.
My military experience includes 3
three (3) years in the U.S. Marine Corps.
4 Q. -Have you ever testified before a comission?
5 A.
Yes, I have.
I was recently employed by the Arkansas Public Service 6
Comission where I testified or participated in numerous water, cooperative, 7
gas, telephone, and electric cases.
8 Q.
Would you please state the purpose of your testimony in Docket No. 32507 9
A.
The testimony will involve two adjustments.
First, I will discuss the 10 company's adjustment to "other 0&M expenses" and propose a staff adjustment.
11 Second, I will comment on the customer, weather, and price elasticity 12 adjustments as prepared by the company.
Thus, the prepared testimony has 13 been organized into two sections:
14 1.
"Other operation and maintenance expense" adjustment.
15 2.
Customer, weather, and price elasticity adjustments.
j 16 Q.
You mentioned the testimony involved two adjustments.
Would you please l
17 mention how your testimony fits into the overall staff case in Docket No.
18 3250?
19 A.
The adjustments were calculated to supplement the work of Ms. Cathy Jones, 20 PUC Chief Accountant, and Mr. Milton Lee, Assistant Director of Engineering, 21 indeterminingtherevenuerequUrementsandadjustedlevelofkwhsales.
22 These adjustments are for the purpose of normalizing revenues and l
23 expenses.
l l
24 I.
OTHER OPERATION AND MAINTENANCE ADJUSTMENTS 25 Q.
Did you review the company's adjustment to "other 0&M expense"?
l
Docke2 No. 3250 Page 3 of 8 1
A.
Yes.
2 Q.
What is the purpose of this adjustment?
3 A.
The purpose is to adjust for other types of operation and maintenance 4
. expenses not accounted for elsewhere.
These expenses include c voluminous 5
number of goods rnd services in which individual study.and adjustment is 6
unfeasible. Examples of "other O&M expenses" would be customer installation 7
expenses, meter expenses, and maintenance of overhead lines.
8 Q.
Do you have any concents regarding this adjustment?
9 A.
Yes.
The staff agrees with the general manner in which the adjustment was 10 made by Ms. Judith Voss, Senior Accountant of Texas Electric Service Company.
11 The only refinement the staff proposes is the use of the confidence interval 12 technique.
13 Q.
Would you describe the staff refinement and the resulting adjustment?
~
14 A.
Yes.
The staff refinement consists of the confidence interval technique 15 which enhances the accuracy of the model proposed by the company.
The 16 company used regression analysis to estimate expenses.
It rendered a point 17 estimate, or mean value, for these expenses.
The deficiency of a point 18 estimate is it tells nothing about reliability.
By using a confidence 19 interval, we can find a measure of reliability. Thus, it can be said that 95 20 times out of 100 an estimate will be within a certain range. This is a 95 21 percent confidence interval and represents the confidence interval 22 technique.
The major conceptual change is the company adjusted "other 0&M l
23 expenses" to the sample mean value of the estimated' relationship while the 24 staff recormends an adjustment to the closest boundary of a 95 percent 25 conficence interval.
i
Page 4 of 8 Docket No.
3250 1
Q. What is the effect of the staff adjustment to the company's "other O&M" 2
adjustment?
3 A.
It reduces the estimated expenses from Ms. Voss's $52,758,536 to $49,870,230; 4
the "other O&M expense" adjustment is thus reduced from $4,407,655 to 5
$1,519,350 and has been reflected in the work of Ms. Cathy Jones.
See 6
Schedule I.
7 Q.
Why should the adjustment be made to the closest boundary of a 95 percent 8
confidence interval rather than the sample mean?
9 A.
The sample mean value is only one indicator of confidence at the 95 percent 10 level.
There are many other indicators within that level and.each has the 11 same probability of being the correct estimate. When making adjustments of 12 this type it is prudent to allow for the relative uncertainty inherent in 13 statistical procedures.
By adjusting to the closest boundary of the 14 confidence interval the staff incorporates the above concerns.
15 Q.
What methodology has the commission applied in the past?
16 A.
The commission has applied the staff approach, as recommended in this docket, l
17 in Dockets No. 1903, 2606, and 3006.
l 18 II. CUSTOMER, WEATHER, AND PRICE ELASTICITY ADJUSTMENTS 19 Q.
Have you reviewed the company's proposed adjustments to kilowatt-hours (kwh) 20 for the effects of customers, weather, and price elasticity?
21 A.
Yes.
22 Q.
What is the purpose of these adjustments?
23 A.
The purpose of thee; adjustments is to accurately reflect the kwh sales the 1
24 company will generate. Changes in certain f actors affect the number of kwh's 25 sold. Known and measurable changes to test year kwh sales should be adjusted l
1 l
Docket No.
3250 Page 1 of 8 1
to correctly represent the company's sales 2
Q.
What is the first issue you intend to discuss?
3 A.
The first issue the staff will address is the change in the number of 4
customers. If test year figures are used without adjustment, kwh sales would 5
probably be understated. The company has adjusted the test year figures to 6
reflect the level of kwh sales that would have occurred if the year-end 7
number of customers had been served for the entire test year. The adjustment 8
isstraightforwardandthdstaffagreeswiththecompany'smethodology.
9 Q.
How did the company make the adjustment for weather and price elasticity?
10 A.
The weather coefficients and price elasticities requested were based on a 11 bifurcated analytical process.
In the first instance, the weather 12 coefficients and adjustments to kwh sales were based on the work of Mr. Ted I
13 Kuhn, former PUC staff member, which were approved in Dockets No.1903 and 14 2605.
For the price elasticity, the company relied upon published studies 15 and their own econometric modeling efforts to apply an elasticity of.20 to 16 all rate groups.
17 Q.
Would you briefly describe the company's econometric modeling techniques?
18 A.
Yes. The company developed econometric models for the R, GS, SD, Y6, and W 19 rate classes. Within these models were various exogenous variables such as 20 weather, income, and electricity price. The R rate class was disaggregated 21 by division.
The remaining models were aggregated on a system wide basis.
22 Data was collected on a monthly basis from January 1972 to August 1979 and 23 the company utilized regression analysis to determine the relationships 24 between the variables.
The model specifications were based on the price 25 elasticity work of Houthakker and Taylor, noted researchers in the field of
~
Docket No.
3250 Page 6 of 8 1
electricity demand theory.
2 Q.
Does the st'aff have any comment to make regarding the procedure just 3
outlined?
4 A.
Yes. One of the goals of the regulatory process in consistency. The staff 5
feels it is inconsistent to use Docket No.1903 and 2606 weather coefficients 6
and current model price elasticities when one model should explain both 7
adjustments. The staff has based its recommendations for weather and price 8
elasticity on updated econometric models incorporating both variables.
9 Q.
How was the staff's econometric modeling accomplished?
10 A.
The staff replicated the company's econometric models using data from 11 January 1974 to the end of the test year.
12 Q. Why did the staff use a different data period than the company?
13 A.
The staff wanted to avoid any structural changes in the economy that might 14 have occured as a consequence of the Arab Oil Embargo and the resultant 15 rising price levels.
It was thought the 1972-1973 period could introduce 16 bias into the* estimates.
Also, the staff wanted the most recent data to 17 increase the reliability of the study.
18 Q.
Could you explain the steps in the staff analysis?
19 A.
Yes.
First, the staff obtained data from the company from January 1974 to 20 the end of the test year, March 1980. This data was operationalized to the 21 same level as that utillied by the company for the test yea'r.
22 Q.
So, what you are saying is that you developed and replicated the company's 23 econometric model to explain the impacts of weather and price on kwh sales?
24 A.
Yes.
25 Q.
What were the results of the modeling on the" weather adjustment?
s
Docket No.
3250 Page 7 of 8 1
A.
In regard to the weather adjustment, the model adjusted kwh sales to a number 2
greater than' that requested by the company.
The staff thus agrees to the 3
lower company adjustment of 11,791,169 kwh sales.
4 Q.
What about the price adjustment?
5 A.
The company uses a price elasticity of.20 for all rate classes. The staff 6
supports an elasticity of
.20 for only the R and GS rate classes.
See 7
Schedule II.
8 Q.
Does an elasticity recommendation for only the R and GS rate classes mean 9
price elasticity does not exist within the other rate classes?
10 A.
No, it does not.
Specifically, I mean that the t-statistic on the price 11 elasticity variables were not significant at the 95 percent level.
The lack 12 of a recommendation of an elasticity coefficient for the other classes is the 13 result of an unacceptable t-statistic for those rate classes.
14 Q.
What is a t-statistic and how was it used in Docket No. 32507 15 A.
A t-statistic tells how important a variable is, i.e.,
it tells if its 16 coefficient is significantly different from zero.
The bigger the t-17 statistic, the more important, or significant, the variable is likely to be.
18 A derived t-statistic is compared with a table of t-statistics to see if it 19 is significant at a certain level, like 90, 95, or 99 percent. The 95 percent 20 level of confidence is a standard used by statisticians.
Checking the t-21 statistics, the staff found the price variables to be significant at the 95 22 percent level for the R and GS rate classes.
The W and Y6 price variables 23 were significant at the 90 percent level.
No independent company or staff 24 studies were conducted on the HV6, MP, GL, Y1, Y71, FSL, SL, or SD price 25 variables. Therefore, the staff made no assumptions as to price elasticity
Docket No.
3250 Page 8 of '8 1
for these rate classes.
2 Q.
What action does the staff recommend the commission take in this matter?.
3 A.
The staff posits concurrence with a price elasticity adjustment of.20 for 4
. the R and GS rate classes. The basis of this recommendation is the staff's
~
5 modeling 'and verification efforts on the compa'ny data.
The total price 6
elasticity adjustment then becomes 462',592,762 kwh.
7 Q.
What are the effects of the total price elasticity figures on test period 8
base rate revenues?
9 A.
The test period base rate revenues become $414,999,505.
This number is 10 reflected in the work of Ms. Cathy Jones, in Schedule I, page 1 of 10.
Il Q.
What are the effects of price elasticity on proposed rates?
12 A.
The price adjustment results is a total adjustment of 452,694,125 kwh for the 13 effects of the proposed increase in Docket No. 3250.
This number is 14 reflected in the work of Ms. Cathy Jones and Mr. Milton Lee. Of course, the 15 above figures are highly dependent on the precise percentage increase final.y 16 approved by the commission.
The staff would have to utilize computer 17 f acilities to give the commission the price adjustment based on whatever 18 final approved increase is granted as it is formidible to do these 19 calculations oy hand. Therefore, the above adjustments should be viewed as 20 only preliminary results as they are dependent upon the commission's final' 21 order.
22 Q.
Does this conclude your testimony?
23 A.
Yes, it does.
24 i
25
-.,r g
,..r
-c_,-
c,
,.,..e 4-.._
4
,~, _ -. - -. -.
e---
e Sch:dule I' DOCKET 3250 Page 1 of 1 PUBLIC UTILITY COMMISSION OF TEXAS TEXAS ELECTRIC SERVICE COMPANY OTHER OPERATION AND MAINTENANCE EXPENSE ADJUSTMENT Formula:
y, = a + bx, + e, Regression equation:
= -232,984,900 + 556.11270 (513,823) y Regression equation:
= $52,758,595.85 y
Coefficient t-Statistic Parameters:
-232,984,900
=
b 556.11270 25.36
=
513,823 x
=
n
$52,758,595.85
=
y yLCI =
$49,870,230.82 Standard Error of the Regression
$1,392,654.31
=
2 Coefficient of Determination (R ),
.96693
=
95% Confidence Interval for E(y)
+$2,888,365.03
=
Where:
a = constant y, = other 0&M expense x, = numbered customers e, = error term, assumed random COMPUTATION Estimated expense for 513,823 customers = $49,870,230.82 Actual expense for 501,736 customers 48,350,881.00
=
Total "other O&M expense" adjustment = $ 1,519,350.00
Schedule II DOCKET 3250 Page 1 of 1 Public Utility Comission of Texas Texas Electric Service Company Price Variable Rate Class h
+
R-BSD
.0390
.383/60.0 i
R-ED
.2471
-1.69/95.0 R-FWD
.2899
-1.78/95.0 R-SD
.1879
-1.80/95.0 R-WFD
.3322
-2.56/99.0 R-WD
.0656
.65/70.0 R-AE
.1934
-2.06/97.5 GS
.2183
-2.29/97.5 4
I i
i I
f
.