ML20076J044
| ML20076J044 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 05/31/1983 |
| From: | Rosen R GREATER NEW YORK COUNCIL ON ENERGY |
| To: | |
| References | |
| ISSUANCES-SP, NUDOCS 8306200290 | |
| Download: ML20076J044 (5) | |
Text
'i I
UNITED STATES OF AMERICA
- 1 NUCLEAR REGULATORY COMMISSION ATOMIC SAFETY AND LICENSING BOARD 5 gt '
Before Administrative Judges: [')?
o,., _
c James P. Gleason, Chairman L. f'
\\
N Frederic J. Shon p.
d!/4,17,
(
Dr. Oscar H. Paris h
- {9
- Qp,,
\\\\
\\ /j In the Matter of
.)
)
)
CONSOLIDATED EDISO!! COMPANY OF
)
. NEW YORK, INC.
)
(Indian Point, Unit No. 2)
)
Docket Nos.
)
50-247 SP POWER AUTHORITY OF THE STATE 0F
)
)
(Indian Point, Unit No. 3)
)
)
i AFFIDAVIT OF GREATER NEW YORK COUNCIL ON ENERGY WITNESS RICHARD A. ROSEN ON COMMISSION QUESTION 6.3 May 31, 1983 8306200290 830531 yDRADOCK05000g47 PDR
~
_ L) 9) c.
-n
C
'\\
l Q.
HAVE YOU ATTEMPTED A COMPARISON SIMILAR TO THAT UNDERTAKEN 2
BY DR. FELD OF ACTUAL 1982 CAPACITY FACTORS FOR SALT WATER PW 3
1982 CAPACITY FACTORS PREDICTED BY THE ESRG CAPACITY FACTOR MODEL?
4 A.
Yes, I have.
)
5 Q.
AND WHAT WERE THE RESULTS OF THAT COMPARISON?
6 A.
In order to corroborate the results of Dr. Feld's comparison 7
I pred[cted the adjusted capacity factors in 1982 for all 14 8
operating salt water PWRs using the ESRG model.
Unlike 9
Dr. Feld's analysis, however, I then compared the predicted 10
.re-adjusted capacity factors, incorporating refueling and NRC 11 mandated outage hours, to the actual unadjusted 1982 capacity 12 factors (as of November 30, 1982).
The result of this comparison 13 was that the ESRp model, predicting an average capacity factor 14 over all plants of 54.3% for 1982, underestimated the actual 15 average 1982 capacity factor of 63.8% by 9.5 percentage points.
16 Dr. Feld, on the other hand, calculated an error margin of 17 13.9 percentage points.
18 Q.
DID YOU FIND ANY OTHER MAJOR DIFFERENCES BETWEEN THE RESULTS 19 OF YOUR COMPARISON AND DR. FELD'S?
20 A.
Yes, I did.
Dr. Feld stated that he found two plant predictions 21 that were larger than the actual 1982 capacity factors, one of which was San Onofre and the other of which was unnamed but 22 cited as having a predicted capacity factor that exceeded the 23 24 actual figure by 1.5 percentage points.
I also found over-25 estimates for San Onof re and a plant (Salem 1) whose prediction 26 was slightly over the actual.
More interestingly, and in -
t 1
contr'ast to Dr. Field's results, my calculations also showed 2
a predicted capacity factor in 1982 for Indian Point 3 (56.2%)
3 that exceeds the actual capacity factor in 1982 of 20.1%
4 b'y a significant 36.1 percentage points.
5 Q.
DR. FELD HAS TESTIFIED THAT THE RESULTS OF HIS COMPARISON 6
INDICATE THAT THE ESRG MODEL IS " BIASED IN TERMS OF UNDER-7 ESTIMATING CAPACITY FACTORS."
DOYOUAGREEWITHhMIS 8
CONCLUSION?
9 A.
No, I do not.
Dr. Feld bases his conclusions on the comparison 10 of predicted to actual capacity factors in one specific year.
11 However, the predictions generated by the ESRG model are not 12 '
meant to perfectly track capacity factors for all plahts and 13 years, but instsad serve to reflect the basic time trends 14 of plants with specific characteristics.
By focusing on 15 one year's experience, Dr. Feld ignores the explanatory 16 power of the model in its ability to track plant operating 17 performance over a span of years.
18 To illustrate the inconsistency in Dr. Feld's approach 19 I performed a second comparison of predicted to actual capacity 20 factors for salt water PWRs, this time for 1981.
The results 21 of this comparison, in contrast to the 1982 results, showed 22 an average predicted capacity factor for all plants (58.6%)
23 which almost perfectly matched the actual average 1981 capa-24 city factor of 58.2%.
Furthermore, of the fourteen plants 25 examined, seven plants had predicted capacity factors which 26 exceeded the actual 1981 figures and seven had predicted,
y 1
values which underestimated the actual 1981 performance.
2 Clearly, no bias exists in the 1981 predicted capacity factors 3
either toward over or underestimating actual performance.
4 Two conclusions emerge from this exercise.
First, as 5
mentioned above, there is limited value in undertaking single 6
year comparisons of the model predictions to actual experience.
7 The value of the model should be judged on itq ability to track 8
experience over time.
Second, and more importantly, because of 9
the year to year variation in the models predictive accuracy, 10 the statistical results should be applied to the Indian Point 11 units with caution.
Thus, in our analysis of the early retire-12 ment of the Indian Point units we adapted a mid-range scenario 13 that forecasts a {%. annual decline in capacity factors assuming 14 a 1982 capacity factor of 55% for Indian Point 2 and 53% for i
15 Indian Point 3.
The mid-range approach is a conservative 16 application of the regression results which show a 7% annual 17 decline for the two units.
18 Q.
DOES THIS CONCLUDE YOUR AFFIDAVIT?
19 A.
Yes, it does.
9 l
I UNITED STATES OF AMERICA BEFORE THE NUCLEAR REGULATORY COMMISSION STATE OF MASSACHUSETTS)) ss COUNTY OF SUFFOLK
)
AFFIDAVIT OF RICHARD A. ROSEN
'l Richard A.
Rosen, being first duly sworn, on oath, deposes and says that the attached statement was prepared by him or under his supervision and the information contained in such is true and
, correct to the best of his knowledge, information and belief.
(
'4 %
A Ri6hard A.
Rosen i
Subscribed and sworn to before me this Of day of N
1983.
g b&M Notary Public
}
My Commission Expires Nov. 12, 1987 l
i l
l l
-