ML20107K459

Forwards Responses to Questions 5,6,7,8,20 & 24 Re Potential Riverborne Missiles.Detailed Probability Analyses Required to Respond to Question 3 Still in Progress & Will Be Submitted by 850320

ML20107K459
Person / Time
Site:	Hope Creek
Issue date:	02/22/1985
From:	Mittl R Public Service Enterprise Group
To:	Schwencer A Office of Nuclear Reactor Regulation
References
NUDOCS 8502280285
Download: ML20107K459 (9)

Text

.

. , _ _ -~

Putgic Sersce O PS G Company Electnc and Gas MAILING ADDRESS / P.O. Box 570, Newark, NJ 07101 80 Park Plaza, Newark, NJ 07101/ 201430-8217 Robert L. Mitti General Manager Nuclear Assurance and Regulation February 22, 1985 Director of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission 7920 Norfolk Avenue Bethesda, MD 20814 ATTN: Mr. Albert Schwencer, Chief Licensing Branch 2 Division of Licensing Gentlemen:

HOPE CREEK GENERATING STATION DOCKET NO. 50-354 RIVERBORNE MISSILES SUPPLEMENTAL INFORMATION On January 31, 1985, we transmitted responses to questions raised in your January 9, 1985 letter on information PSE&G submitted on September 17, 1984 on potential riverborne missiles. We indicated in our January 31, 1985 submittal that some analyses would be forwarded at a later date.

Attached are our responses to questions 5, 6, 7, 8, 20 and 24. We are still completing some detailed probability analyses required to respond to question 3. We anticipate submitting a response to question 3 on or about March 20, 1985.

Very truly yours, Attachments .

[f C [D Z Hg Wagner 7 (w/5Tsets 7 of 'attacht ) : ;

, USNBC , Pro ject ; Licensing ., Manager;]

A. R. Blough (w/ attach.)

USNRC Senior Resident Inspector 4 bohh$54 odk U

PDR The Energy People '\

98>491214M) 7 83

k g: +-

HOPE CREEK GENERATING STATION PUBLIC . SERVICE ELECTRIC - AND GAS COMPANY RESPONSES TO QUESTIONS IN JANUARY 9,1985, LETTER -

RIVERBORNE MISSILES.

- Questions:5, 6, 7, 8 5.. Provide the uncertainty associated with the extrapola-tion of an . ll-year data base to ' represent return periods i

of up to 200 years, particularly in the development of a relationship between extreme wind speed and wind direction.

6. Provide further just'ification of the use of .the ' Fisher-Tippett Type 2' dJstribution when other analyses of extreme _ winds (e.g., NUREG/CR-2639, ' His torical - Extreme

- Winds for the United ~ States-Atlantic and Gulf of Mexico

- Coastlines')'suggest use of a' mixed distribution, with all Type I distribution fitting non-tropical storms and the Weibull distribution fitting tropical sto rm s .

7. The statement is made on1page 2.that ' the particularly open exposure of this site is not adequately duplicated at any of .the National Weather Service (NWS) stations in the region,' implying that-extreme wind speeds at the site may be higher- than at the- NWS. stations. Provide a

- discussion of.the exposure of NWS stations in the region, particularly - at .Wilmington~, DL, and provide com-parable estimates of extreme winds at the NWS stations '

-for return periods- of 20, 50, 100, and 200 years as in 4 -Table 2.

. 8._. Typically, extreme winds are represented by the fastest mile windspeed . Provide' a comparison 'of fastest mile wind speeds for return periods of 20, 50, 100, and 200 f~ years between ' regional NWS stations and the site.

- Response to Questions 5, 6, 7, 8 t -

l ' These four questions are interrelated , all bearing either on the representativeness of .the onsite data or on the methods used in :the processing and extrapolation of the results to

'very long return periods by Meteorological Evaluation Serv-ices, Inc. (MES) in their reports relating meteorological conditions to site flooding. The reports dated July 6,

. - 1984, and August 1984.were transmitted, via R. L. Mitti to A.'Schwencer letters dated July 27, 1984, and September 17, 2

1984, respectively.  ;

I 4

M P85 42/09 1-cag Y y--v,---s we e 9 -r,4--ie-,cr-- ,---w ,-& a .-nc-,,--,,m,-w-c-a,e -w w+4

• we %vm% % -n ,-em v e-ww ww w avn<e--w-w,.,'av -> v g-wg--- w -w--vwv v vo y w--,-s--r- vy e

Response to Questions 5, 6, 7, 8 (Cont'd)

Rather than attempting to re-explain the reasons for believ-ing that the analysis performed earlier was reliable, Mr. Michael Changery of NOAA's National Climatic Data Center was contacted directly by MES for both for his opinions and his assistance with the handling of such data. Mr. Changery is the author of " Historical Extreme Winds for the United States - Atlantic and Gulf of Mexico Coastlines." The.

center then processed the most pertinent fastest-mile wind data, _from Cape May and the Delaware Breakwater. These two stations have records extending over 44 and 59 years, respectively.

On Mr. Changery's advice, the data were subjected to a Fisher-Tippet (PT) Type 1 analys even though he prefers a hybridFT-1andWeibullapproach{s, The use of the FT-1 method alone permits a reasonable evaluation of the uncer-tainty of the results, whereas estimation of probable error or sigma values is difficult with the joint distributions.

Based on discussion with NRC staff members and their consul-tant, held on December 18, 1984, in Bethesda, MES first suggested an analysis encompassing the entire set of

' fastest-mile data at each station, regardless of the direc-tion from which the observations came. This approach pro-vides the most reliable estimation of the variability of the data at long return periods because all of the data points are used, and presumably it should also yeild the highest wind speed. estimates possible from the data sets.

The second step in the procedure, which was discussed and verbally agreed to at the December meeting, was to apply the ratios that were obtained f rom the site data to reduce the fastest-mile estimates to one-hour and six-hour time periods.

The results of this new analysis are summarized in two tables and one . figure. Table 1 shows Mr. Changery's fastest-mile estimates together with the standard deviations of the sampling errors. The formula used to obtain the sigma values is also shown. The estimated values for lengthy return periods are lower than those which were de-rived in the MES August, 1984 report. At the 20,000 year return period, for example, the comparison is:

i M P85 42/09 2-cag

Response to Questions 5, 6, 7, 8 (Cont'd)

MES NCDC Estimates Estimate Delaware Cape Hope Creek Breakwater May 130-150 mph 110 mph 123 mph In order to reduce these fastest-mile estimates to the one-

-hour and six-hour periods needed for the major storm surge study, we' returned to the fastest-mile / average ratios shown in Table 1 and Figure 1 of the MES August 1984 report 2, These data showed an unmistakable tendency for the ratios to increase as the fastest-mile figures increased, and MES therefore obtained formulas for the relationship based on a linear least-squares fitting process. The data seem quite linear over the range of observed fastest-mile data (22-59 mph), but it seemed unwise to assume that this linearity would maintain at much higher speeds. Therefore the ratios shown for the one-hour and .six-hour adjustments that remained constant at 2.0 and 2.25, respectively were used, rather than to extrapolate the values as suggested by the formulas. Figure 1 summarizes all of the ratio data and depicts as solid lines the ratios that MES considered reliable.

Table 2 summarizes all of the data from Artificial Island, Delaware Bay and Cape May, presenting estimates of f astest-mile one-hour and six-hour maximum winds from the dif ferent sources. From this comparison, it seems evident that MES's prior estimate of fastest-mile winds of 130-150 mph, at a 20,000 year return period was conservative. Similarly, it seems.most unlikely that one-hour and six-hour sustained winds would exceed 75 and 65 mph respectively at this same return period.

Questions 20 and 24

20. In any given storm, wind speeds over water are higher than wind speeds over ground. It appears this was not taken into account in the estimate of annual probabili-ties of extreme 6-hour wind speeds at 33 ft. eleva-tion. Is this the case . If so, why?

M P85 42/09 3-cag

.., .n

-Questions 20 and 24 (Cont'd)

24. Page 7 of the August 1984 MES report reproduces esti-mates of directional fastest-mile wind. speeds from the NBS BSS-124 Report. However, the estimates of NBS

.BSS-124 pertains to winds blowing from a 360*/16 =

2:2. 5

• sector , rather than from the 79 *-170

• sector.

., Using . data on which the NBS BSS-124 Report is based (which are available on tape at NBS), the fas tes t-mile wind speeds at 21 feet over ground is estimated as 36 mph,173 mph, and 85 mph for .the 10, 50, and 100 year means recurrence intervals, respectively, rather than 24 mph, 57 mph, and 70 mph, as indicated on page 7 of the August 1984 MES Report."

Response to Questions 20 and 24 The-over-water winds would exceed those over land by a small 1 '

fraction, but with the extremely flat terrain at Artificial Island, it is not expected that the difference would exceed a few percent. The use of the NBS data mentioned in Question 24 may have been inappropriate, but the new study reported as a response to . Questions 5-8 supersedes any of the earlier work, either on the part of MES or on that of the National Climatic-Data Center.

It is important to note that even a 10% increase in the estimates of the maximum-winds over six-hour periods would not produce surges capable of developing significant flood-ing on a 20,000 year return period.

a-M P85 42/09 4-cag

. .+ -

r:

REFERENCES 1 Chengery, M: Personal Communication, Feb.-7, 1985 2 Smith,,M.E.i&'Castelli F. P.: Hope. Creek Generating Station' Extreme Event Site Flooding, Supplemental Report ~-on Meteorology,. August 31, 1984.

4 i

h !_ h ',

M_P85 42/09 S-cag

Table 1

~,

DELAWARE BAY ENTRANCE FASTEST - MILE WIND SPEED ESTIMATES

• Delaware Breakwater Cape May Standard Deviation Standard Deviation Return Period Wind Speed of Sampling Error Wind Speed of Sampling Error s (years) (mph) (mph) 50 77 2.9 74 4.8 100 81 3.4 80 5.7 200 85 3.9 85 6.6 500 90 4.6 93 7.7 1000 94 5.2- 99 8.6 10,000 107 6.9 117 11.5 20,000 110 8.3 123 13.9 .o f.

• From Michael Changery, Acting chief, Applied Climatology Branch, National Climatic Data Center, Asheville, North Carolina, February 1985.

1/2 2

SD = n 2 + 1.396(y - .5772) + 1.1(y .5772) o o

ja }n Where: SD = Standard deviation of the sampling error y = -In (-In ( 1-1/n ) )

N = Return period n = Sample size o = Estimated value of scale parameter METEOROLOGICAL EVALUATION SERVICES,INC.

.t..

Table 2 ESTIMATES OF MAXIMUM WINDS (MPH)

FASTEST MILE HIGHEST 1-HOUR HIGHEST-6-HOUR Return Period MES DBCM MES DBCM MES DBCM (years) 50 55 77 28 39 24 34 100 .67 81 34 41 30 36 200 74 85 37 43 33 38 500 90 93 45 47 40 41 1000 101 99 51 50 45 44 10,000 132 117 66 59 59 52 20,000 145 123 73 62 64 55 MES - Estimates derived from Fig. 4 plot, 8/84. report.

DBCM - Delaware Breakwater-or Cape May, whichever was higher (Changery, 2/85).

METEOROLOGICAL EVALUATK)N SERVICES,INC.

Figure 1 RATIOS OF FASTEST-MILE TO LONG-TERM AVERAGES HOPE CREEK

/ /

/ / (The solid curves RANGE OF / / were used for reducing OBSERVED DATA / / FM to longer- term overages.)

f p l /

/

2 -

/ ,

l '

6-Hour l

R - 1-Hour R = AVG

) _ (1 - Hour) R = .022 FM + .654 (6 -Hour ) R = .025 FM + .711 20 40 60 80 10 0 120 140 l

FM (mph) l

.