ML19350B271
| ML19350B271 | |
| Person / Time | |
|---|---|
| Site: | Yankee Rowe |
| Issue date: | 03/16/1981 |
| From: | Kay J YANKEE ATOMIC ELECTRIC CO. |
| To: | Crutchfield D Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19350B272 | List: |
| References | |
| TASK-02-03.A, TASK-02-03.B, TASK-2-3.A, TASK-RR FYR-81-42, NUDOCS 8103200196 | |
| Download: ML19350B271 (7) | |
Text
.
YANKEE ATOMIC ELECTRIC COMPANY m-g 167I Worcester Road, framingham, Massachusetts 01701 2.C.2.1 YANKEE FYR 81-42 March 16, 1981 cp Y.
9 Y
f' linited States Nuclear Regulatory Commission U
,C Washington, D.C.
20555
(
,g 9
Attention:
Mr. Dennis M. Crutchfield, Chie f Operating Reactors Branch #5 j
Division of Licensing Co g
References:
(a) License No. DPR-3 (Docket No. 50-29)
(b) USNRC Letter (w/ Enclosure) to YAEC dated February 9, 1981 (c) Design Basis Flood Analysis, YAEC-1207, October 1980
Subject:
Yankee Atomic Electric Company Comments on Draf t Safety Evaluation -
SEP Topics II-3.A and II-3.B (Partial)
Dear Sir:
As requested in Re ference (b), we are submitting our comments on the staff's draf t report of Sys tema tic Evaluation Program Topics II-3. A, Hydrologic Description, and II-3.B (Partial), Flooding Potential and Protection Requirements.
Yankee has rigorously analyzed the flood potential in the Rowe area and is sebmitting its final report as Reference (c) of this letter. The conclusion of this study is th a t the Somerset and Harriman Dams will not be overtopped. Although the Searsburg and Sherman Dams would be overtopped, the peak water level at the Yankee Rowe site would be below the crest of the on-site flood protection barriers.
Based on Yankee's in-depth analysis of the potential maximum rainfall in this specific region, we must strongly dispute the conclusions and recommendations made by the NRC in Reference (b).
Yankee's basic disagreement with the Dra ft Safety Evaluation (Reference b) is with the conclusion that the rainfall from Hydrometeorological Report No. 33 should be used as the design basis rainfall. Use of that report produces a generalized Probable Maximum Precipitation (PMP) estimate which is inappropriate for the Rowe site as explained in our attached report. Fur th e rmore, the NRC staff's evaluation clearly states that "NRC regulations do not specifically require a PMP for the Design Basis Flood for nuclear power plants."
NRC's Draft Safety Evaluation includes a storm transposition analysis, the results of which difier from our design maximum rainfall by only 13 percent. This relatively minor difference is apparently attributable to certain assumed storm transposition adjustment factors which are not supported 903200 kb
O-Regula tory Commission March 16, 1981 Mr. Dennis M. Crutchfield Page 2 in the staff report. The corresponding storm adjustment factors are fully
' documented and supported by tne results of our extensive investigations as described in the attached Yankee report. The Sta ff report states that since the transposed storm results do "not alter the conclusions of this study, it was concluded that the rainfall from H.R. No. 33 would be used as the design basis rainf all for this study." We submit tha t this conclusion will change if the proper storm adjustment factors from the Yankee final report (Re ference (c)) are used.
. More specifically, it is concluded in the Draft Safety Evaluation, Reference (b), that the rainfall fron Hydrometeorological Report No. 33 would be used as the design basis rainfall.
But neither this report nor its updated version, Hydrometeorological Report No. 51,
.a applicable to the Rowe site secause.(as pointed out in each document) the data have not been evaluated with respect to terrain effects. We recognized this limitation and conducted a regional rainfall study to demonstrate the presence of terrain effects and determine the adjus tments required in the data. The results of this study sh ow th a t terrain effects produce rainfall values considerably different from those presented in the generalized maps of H. R. No. 33.
We, there fore,
assert th a t H. R. No. 33 should not be used directly to evaluate the flood potential in the Rowe area.
It is also apparent that the hydrologic concepts required to adequately eval ua te flooding potential at Rowe have been misinterpreted.
For example, pages 19 and 20 of the draft report point out the differences in Clark coe fficients between the sta ff and Yankee. values. We did not, however, provide the Clark coefficient values shown. Therefore, r.he conclusions made by the staf f regarding these coe fficients are misleading.
The Draft 1 Safety Evaluation repeatedly states that the basin rainfall "does not meet current licensing criteria." This also is not true.
Not only are-all. appropria te criteria met, but we feel that.we have incorporated many l
conserva tisms. in our analysis (Reference c).
Regulatory Guide 1.59, which includes American National Standards Institute Standard N170-1976, provides the applicable licensing criteria. These criteria are met.
For example, ANSI l
N170-1976 states. that where the generalized probable maximum precipitation i
estimates do not adequately cover an area,Lspecific studies may be made. We j
have done' this by performing a comprehensive regional rainfall study.
In addition, as suggested in ANSI N170-1976,'our. conclusions represent a risk probability. In Reference (c), we have presented a statistical analysis of maximum tainfall for the Rowe site that shows the annual exceedance probability of our-design basis rainfall to be in the order of one in ten million (1 x 10-7).
When combined with other flood design basis factors, l'
the annual exceedance probability of the design basis flood will be substantially lower and, therefore, lower than the acceptable ANSI standard.
i r?
L
O i
Regulatory Commission March 16, 1981 Mr. Dennis M. Crutchfield Page 3 4
P l
These problems with the NRC draft report are disconcerting in light of the fact that the staff had been advised of new and updated information i
subsequent to submittal of Yankee's draft report in January 1980. Yet thit l
informa tion, which in many cases provided additional conservatisms, was excluded from the draft evalua tion.
In addition, the staff study demonstrates a. pronounced preference for subjective rather than quantitative analysis.
Accordingly, the staff should review Yankee's final flood report (10 copies of which are enclosed).
Since Yankee's study is much more comprehensive and quantitative than the staff's evaluation, and since it is site specific, Yankee strongly recommends that it be adopted by NRC as the Final Safety Evaluation of SEP Topics II-3. A and II-3.B.
Yankee's detailed comments on the staf f Draf t Safety Evaluation Report are included as. Enclosure A of this letter. We are prepared to discuss our analysis with the staff leading to its adoption as a technically defensible 4
Final Safety Evaluation Report on the flooding potential of the Yankee plant site.
In closir.g, we wish-to point out that Yankee does not own or operate, and 1
is not in.any way responsible - for Harriman Dam.
That dam.is owned and operated by New England Power Company (NEP). Although NEP is a part owner of Ya nk ee, it is a separate corporation both legally and financially, with (1) di f f eren t. owner sh ip, (2) different management, and (3) a different Board of Directors. JEP.is regulated with respect to Harriman Dam by the Federal Energy Regulatory Commission.(FERC) under License No. 2323.
FERC is responsible for. protecting Yankee, just as they are for protecting any other downs tream activity, such as farms and factories built along the river. If there really were. a' problem with the dam, FERC can eliminate it by requiring the reservoir to 'oe drawn down. This is a controllable matter, unlike the uncontrollable - natura1' phenomena that NRC' usually considers, and the legal f authorityJand legal responsibility for its control rests with the Federal
' Energy Regula tory Commission under 18CFR12 (rev. 3-1-81).
Very truly yours, YANKEE ATOMIC ELECTRIC COMPANY J. A. Kay Senior Engineer'- Licensing Enclosures w
r
'-J',
4
- -, +. - - - >
Yankee Atomic Electric Company Comments On SEP DRAFT SAFETY TOPIC EVALUATON YANKEE ROWE NUCLEAR POWER STATION TOPICS II-3. A, HYDROLOGIC DESCRIPTION, AND II-3.B. (PARTIAL), FLOODING POTENTIAL AND PROTECTION REQUIREMENTS e
1.
Section 2, Discussior )f Problem Page 4, first paragraph - The NRC Draft Sa fety Evaluation report, Reference (b), states that Harriman Dam does not meet current licensing criteria.because the design basis Probable Maximum Precipitation (PMP) will overtop the dam.
But on page 15, paragraph three, the report states "NRC regula tions do not speciitcally require a PMP... (emphasis added).
This contradiction between statements of whether a PHP is required or not illustrates the ccafusion as t6 what rainfall is appropriate. Re gula tory Guide 1.59 and ANSI Standard N170-1976 provide the appropriate criteria for determining the design basis rainfall. These governing documents clearly point out that where areas are not adequately covered by generalized PMP estimates, specific studies may be used. This is the
-case for the' Rowe site (see item 3a). The NRC Draft Sa fety Evaluation, therefore, should reflect this and be rephrased accordingly.
2.
Section-3.3.1, Flood His tory Page 13, first paragraph - We agree that the Smethport, PA storm cannot be transposed across the Appalachian Mountains. Consequently,- the las t two sentences of the first paragraph on page 13 should be deleted since the Smethport storm does not contribute to an evaluation of the design basis rainfall 'at the Rowe site. The storm' should also be deleted from
~
. Table 3.3.2.
3.
Section 4.2.1,-Probable Maximum Precipitation a)
Page'14, last paragraph and page 15 continuation - We strongly
. disagree with : the use of Hydrometeorological Report Number 33
_ generalized' maps'as.a source for design baris rainfall estimate at the Rowe site. As clearly pointed out in 'H.R. No. 51, the. updated version of H.R. No. 33, terrain effects can yield rain estimates significantly di fferent from the generalized. maps because the. normal-smoothing -procedure used to develop the maps' does not account-for these eff ec^:s. ;The -generalized maps,Etherefore,- havelbeen "s tippled (shaded) ~ to highlight-the. terrain-e f fected regions.
Since"the Rowe site' is. located in a stippled zone, the use of-the
- generalized maps to fes tima te the PMP cannot be done' with a high _
.de gree of confidence. Accordingly, it is our contention 'that a
comprehensive regional rainfall study incorporating terrain ef fects is the appropriate approach to estimate the design basis rainfall at the Rowe site. This is supported by the statement on page 29 of the s taff report. We have done this as detailed in Reference (c).
b)
Page 15, second paragraph - The final Yankee storm transposition values are correctly derived and are explained in greater detail in Reference (c).
c)
Page 16, second paragraph - There is a contradiction with design basin rainfall values.
On page 16, for instance, the 236 square mile basin rainfall is shown as 16.5 inches in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, but on page 20a the value is stated as 13 inches.
A 4.
Seciion 4.2.3, Unit Hydrograph Coe fficients Pages 18 We concur with the staff's conclusion that in the interest of conservatism the use of the Snyder coefficients furGished by us, derived and verified with actual flood events, are appropriate. However, several points in the staff's discussion warrant clarification.
As noted on page 19, we furnished to the staff the Snyder Unit Hydrograph coe fficients for each of,the four subbasins. These coefficients are accurately depicted on the left side of the table comparing NRC and Yankee Atomic values. However, we did not furnish the Clark coefficients a ttribu ted to Yankee Atomic on the right ' side of the same table.
- Therefore, the staff 's comparison. between the NRC and the " Yankee Atomic" Clark coe fficients is misleading.
In addition, we believe that the staff misinterpreted the coefficients.
For example, when using the ' HEC-1 ' program to convert from Snyder to Clark coe fficients,' the. units are directly. related to the tabulation interval chosen 'in the model. ~ Therefore, care should be taken not to compare Clark icoe fficients generated by the HEC-1. program unless the same time interval was used.. Dur work was performed using a 15-minute time interval; al though not clearly de fined in the staff's evaluation, a time interval other than 15 minutes ' appears - to have been used.
' In light of this. dis'cussion on the Clark coefficienes, the statement in -
~
'the first paragraph on page - 20. that ~the runoff hydrographs developed from
" Yankee: Atomic" coe fficients' will be broader l and fla tter (lower peak -
discharge). than the hydrographs. developed with the NRC coefficients should be' corrected. In fact, Yankee hydrographs used in Reference (c) are more critical' and, therefore, yield higher peak discharges for equivalent rainfalls.
s c
7
Finally, if the sta ff's statements concerning the Clark coe fficients were correct, the s tatement in the second paragraph on page 20 concerning the interest of conservatism is illogical because, on the Deerfield River, broad and flat (lower peak discharge) runoff hydrographs wculd not be as conservative as ones with a higher peak discharge.
5.
Section 4. 3.1, Reservoir Rou ting Pages 22 and 23 - Based on our work to date, we disagree with the staff's conclusion regarding a 25% reduction in capacity at higher heads for the Harriman Dam morning glory spillway. For a detailed discussion of.>ur position on the spillway, refer to Reference (c), pages 38 through 44.
6.
Section 4.4, Erosional Dam-Failure a)
Page 25, second and third paragraphs - The staff in its analysis predicts substantial overtopping of Harriman Dam during the PMP.
We feel this is a result of inappropriate input parameters placed into the HEC-1 model. To aid in quantifying the duration of the postulated dam breach, tLe staff chose to use the TVA Dam Breaching Program. The use of the model as applied is inappropriate at Harriman for two reasons.
First, the' staff obtained the Harriman inflow hydrograph for the TVA model "by routing the PMF with the HEC-1 and assuming infinite dam heights." This approach is wrong. The correct method is to
' determine the headwa ter eleva tion hydrograph by allowing. water to overtop the dam without failing it.
Second, there are coding errors
-in the TVA program (as acknowledged to us by TVA) which, if not adjus ted,' can produce erroneous resul ts.
b)
Page 25 and 26 - Since the Teton Dam failure does not represent an overtopping breach, the fact tha t the Teton. Dam failed rapidly is of no relevance where overtopping breaching is concerned. The phenomena which lead to the failure are completely different and, hence,'should not be used as a comparison. This entire discussion of Teton Dam should be deleted.
7.
Section 4.5, Recurrence Intervals of' Natural Phenomena Page 26 and 27 - We do not understand why. the staff's opinion and judgment is used in lieu of a rigorous statistical analysis of available
. rainfall data.- The actual ' rainfall data reflect the meteorological and physical variables that. cause these rainfalls to occur.. Therefore, the
~
staff's statement that we did not include rainfalls at Westfield, Kinsman Notch, and Springfield in our -statistical analysis is not appropriate because these locations are outside the drainage ~ basin of the Deerfici River..Rainf all records ' that are applicable for a probabilis tic evaluation rainfall should reflect the same rainfall producing meteorological and: physical variables that exist at the Rove site.
i
, i It is interes ting to note that our estimate of the 24-hour, 100 year point. rainfall of about 6 inches, presented in Appendix A of our final report, is approximately the same as that presented by the National Wea ther Service (then the Wea ther Bureau) in Technical Paper 40, not the 7 to 8 inches the staff has " judged" the 100 year rainfall to be.
In addition, our analysis shows the 24-hour,1000 year point rainfall to be about 8 inches. The logic the staff used to obtain a "500 to 1000 year" rainfall of 13 inches is not clear.
8.
Sec tion 5, ' Recul t -:
Based on the comments above, we contend that the results presented in the dra f t safety evalua tion report (Reference b) are incorrect.
Since Yankee's study is m2ch more comprehensive and quantitative than the staff's evaluation, and since it is site specific, Yankee strongly
. recommends that it be adopted by NRC as the Final Safety Evaluation (FSE) of SEP Topics II-3.A and II-3.B.
9.
Section 6, Conclusions and Recommendations We believe tha t the conclusions were drawn from deficient evidence and
' incorrect. application of licensing criteria. As a result, the recommendations offered should.be withdrawn and, as stated-in item 8, the staf f should adopt the Yankee FSE as a -technically de fensible report on the flooding potential of the Yankee site.
.Chr report includes up-to-date information on design basis rainfall, and
~
' spillway l characteristics, as Lwell as expanding upon a number of other l considerations. ' Consequently, there are added conservatisms beyond those contained in our dra ft report: submittal. These include an increase in the
. design basis rainfall, a decrease of the initial infiltration, a reduction to spillway capacity, and an increase. in the per'centage of rainfall distribution
'in 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
.