ML20031B697
| ML20031B697 | |
| Person / Time | |
|---|---|
| Site: | Midland |
| Issue date: | 09/29/1981 |
| From: | Steptoe P CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.), ISHAM, LINCOLN & BEALE |
| To: | NRC COMMISSION (OCM) |
| Shared Package | |
| ML20031B698 | List: |
| References | |
| ISSUANCES-OL, ISSUANCES-OM, NUDOCS 8110050438 | |
| Download: ML20031B697 (14) | |
Text
-
4 e
[
6 9
9/29/81 OCT 11gggi H
3
-9
- ' O ?" '3]
UNITED STATES OF AMERICA mumu Colum3POSDEW 7,
t ;i -
"'~
NUCLEAR REGULATORY COMMISSION L-ATOMIC SAFETY AND LICENSING BOARD g
x In the Matter of
)
Docket Nos. 50-329 OM
)
50-330 OM CONSUMERS POWER COHPANY
)
Docket Nos. 50-329 D
(Midland Plant, Units 1 and 2) 50-33 L
01 g
~1 v
APPLICANT'S BRIEF ON COMPATIBILIT3 EI OF SITE SPECIFIC RESPONSE SPECTRAgT S
-8 2
APPROACH WITH 10 CFR PART 100 APPENDlX A
- /
3 h
/
.y 1981 this Licensing Board issued a DI/l'
'\\'
In August 18, Memorandum which stated, in part:
It is our understanding that the Applicant will elect one of the two alternatives for selection of the Safe Shutdown Earthquake (SSE) and associated ground acceleration suggested in the letter dated October 14, 1980 from Robert L.
Tedesco (NRC) to J. W. Cook of Consumers Power Co.,
titled "Jeismological Input for the Midland Site."
The first alternative is apparently based on a deterministic application of 10 CFR Part 100, Appendix A (utilizing the standarized response spectra of Regulatory Guide 1.60 ar. chored at 0.199).
The second appears to be a probabilistic approach which utilizes site specific response spectra.
If the Applicant should elect to rely upon the site specific response spectra approach, we request that both the Applicant and Staff file trial briefs discussing the compatibility of this approach with the requirements of 10 CFR Part 100. Appendix A.
See, in particular, paragraphs V (a) (1) (ii) and (iv) of Appendix A.
The Staff's brief should, inter alia, provide information concerning the Commission's approval of or other action, if any, with respect to the site specific response spectra approach.
This brief first clarifies the use to which the Midland Site Specific Response Spectra developed by Applicant at the Staff's request will be put.
It then discusses the limited use 9503 8110050438 810929 y /
\\
a L
. __=.
_n,
- - =
t 1
1 i
made of probabilistic techniques in determining Site Specific Response Spectra, and demonstrates that such limited use of probabilistic technique is entirely consistent with 10 CFR Part 100, Appendix A.
This brief also serves as an Introduction to the testimony of Applicant's expert witness, Richard J. Holt.
4 i
4 l
i i
I i
l I
(
i l
I
(
I 1
l
I 3-I.
THE DESIGN BASIS OF THE MIDLAND PLANT IS NOT BEING CHANGED BY THE SELECTION OF A NEW " SAFE SHUTDOWN EARTHQUAKE" The Site Specific Response Spctra ("SSRS") presented in Mr. Holt's testimony will be used by Applicant for the re.nedial (underpinning) work proposed for the Midland Plant.
The SSRS will also be used in a " seismic margin review" to assess the adequacy of Midland I structures which have already been designed and built to the FSAR design spectra.
The distinction between " design basis" and " margin review" is necessary because it is not practicable literally to substitute a new design basis for the old design basis after structures have already been built.
The adequacy of the design basis reflected in existing structures can, however, be assessed in a " seismic margin review" using a new postulated earthquake, and this is what the NRC Staff has done in other recent operating license reviews, including its review of TVA's Sequoyah plant.1/
Therefore the use of the term, " Safe Shutdown Earthquake" in the October 14, 1980 Tedesco letter (Holt Exhibit 3), in this brief, and in the testimony of Applicant's witness Mr. Holt does not imply that the seismic design basis of the Midland plant, is being changed.
1/ See U.S. NRC " Safety Evaluation Report related to operation of Sequoyah Nuclear Plant Units 1 and 2, Tennessee Valley Authority, Supplement No.
1" (February 1980), NRC Docket Nos. 50-327, 50-328.
e
. II. THE SITE SPECIFIC RESPONSE SPECTRA APPROACH USED IN THIS CASE IS PRIMARILY DETERMINISTIC - NOT PROBABALISTIC The Applicant has indeed chosen the Site Specific Response Spectra approach suggested in October 14, 1980 Tedesco letter (Holt Exhibit 3) to establish the ground acceleration associated with the Safe Shutdown Earthquage as defined by the Staff in that letter.
The SSRS developed by Applicant are explained in the testimony of Applicant's expert witness, Richard J. Holt, and shown in Exhibits 1 and 2 attached to his testimony However, as Mr. Holt explnins, the approach taken in developing these SSRS was primarily deterministic, not probabilistic, in nature.
10 CFR Part 100, Appendix A describes procedures by which the vibratory ground motion at a site due to earthquakes shall be determined.
These procedures involve two basic decisions:
first, the selection of a Safe Shutdown Earthquake ("SSE"); and second, the construction of response spectra representing the vibratory ground mot
.. p cduced by the SSE at the site.
See Testimony by Richard J. Holt at page 2.S!
3/ Appendix A defines the term Safe Shutdown Earthquake" in two different ways.
In Sections III(c) and V(a) the term refers to the earthquake itself; in the last sentence of Section V(a) (1) (iv) the statement is made that "the Safe Shutdown Earthquake shall be defined by response spectra corresponding to the maximum vibratory accelerations...."
For purposes of this brief and Applicant's testimony the term
" Safe Shutdown Earthquake" or "SSE" refers to the earthquake itself, not the response spectra describing the resulting ground motion.
n-.
c.-
. A.
THE SAFE SHUTDOWN EARTHQUAKE FOR MIDLAND WAS ESTABLISHED DETERMINISTICALLY, IN ACCORDANCE WITH 10 CFR PART 100, APPENDIX A Appendix A is fairly explicit on how the SSE for a nuclear power plant should be determined.
In the absence of capable faults or tectonic structures in the vicinity of the site, Appendix A states that a tectonic province should be defined based on "a reasonable consistency of the geologic structural feaures contained therein" 10 CFR Part 100, Appendix A, Sections III(h), V(a).
Appendix A then requires that, as a minimum, the maximum historic earthquake intensity experienced within this tectonic province shall be assumed to occur at the site, 10 CFR Part 100, Appendix A, Sections V(a) (1) (ii) and V(a) (1) (iv).
Because this procedure calls for judgments which do not formally take into account uncertainties, it often is referred to as " deterministic" rather than "probabilistic".
The deterministic Appendix A procedures outlined above have been followed in defining the Safe Shutdown Earthquake for Midland.
From the testimony of Applicant's witness Mr. Holt i
(pp. 11-15) it is apparent that the Statt has chosen an j
unreasonable tectonic province which has led to the identification of an unreasonably large SSE for the Midland s
1 site.
The Site Specific Response Spectra presented in Holt Exhibits 1 and 2 are based on the Staft's deterministically defined, unreasonably large SSE.1!
By emphasizing that the Safe Shutdown Earthquake developed for Midland is deterministic, Applicant does not mean to imply that probabilistic considerations may not be taken into account in determining the SSE under 10 CFR Part 100, Appendix A.
For example, it is quite clear that seismicity - that is, the 1/ The Licensing Board's Memorandum evidences some understandable confusion on this point.
It is possible to develop " site specific response spectra" using probabilistic methods, without first defining " tectonic provinces" or " Safe Shutdown Earthquakes" as required by Appendix A.
Indeed, in developing " site specific spectra" for use in its Systematic Evaluation Program (which includes the Lacrosse Boiling Water Reactor), the NRC Staff used a complex synthesis of aeterministic judgments and probabilistic modelling, which did not, at least explicitly, follow the deterministic procedure outlined in Appendix A.
However, the SEP (Lacrosse) approach to developing " site specific spectra" is totally different from the site specific response spectra approach used in Midland.
Confusion also may have been caused by the fact that Applicant itself prepared and submitted on March 2, 1981 to the NRC Staff a report entitled, " Site Specific Response Spectra, Midland Plant Units 1 and 2, Part III, Seismic Hazard Analysis" in which probabilistic response spectra are developed for the Midland site.
These probabilistic spectra were submitted for comparison purposes only.
The Part III report is not being submitted into evidence with Mr. Holt's testimony.
Instead, a summary of Applicant's probabilistic analysis is included as Holt Exhibit 10.
However, a quick reference to the docketed Part III report will show that the probabilistic spectra contained in Figure 20 of the Part III report are not the same curves as the deterministic SSRS shown in Exhibits 1 and 2 attached to Mr. Holt's testimony.
relative frequency of earthquakes in a particular region -
which is a probabilistic consideration, may be considered under Appendix A.
Section V(a) of Appendix A states:
The design basis for the maximum vibratory ground motion should be determined through evaluation of the seismology, geology and the seismic and geologic history of the site and the surrounding region.
(emphasis added).
Moreover, in 1977 the Commission amended Appendix A to clarify that the maximum historical earthquake shall be used for the SSE unless the assumption of the possible occurrence of higher intensity earthquakes is warranted by " geological and seismological data."
One of the examples of the types of
" seismological data" which might warrant selection of a larger earthquake was:
(2) where seismicity in the immediate site vicinity is significantly higher than that generally existing in the tectonic province as a whole. (emphasis added).
Last year, in Public Services Company of New Hampshire (Seabrook Station, Units 1 and 2), CLI-00-33, 12 NRC 295, the Nuclear Regulatory Commission held that the proposed use ot a particular probabilistic methodology to determine the Safe l
Shutdown Earthquake at the Seabrook site was not barred by Appendix A.
On remand the Staff testified before the Appeal Board that it customarily does make limited use of probabilistic evaluations in assessing earthquake hazard, although the results are used a " qualitative" or " relative" m
--v-
1.
manner rather than in the quantitative, absolute manner suggested by the Intervenor in Seabrook.
In short, Appendix A contemplates a deterministic " cookbook" approach to establishing the SSE which involves defining tectonic provinces and maxinium potential earthquakes, but Appendix A certainly does not bar the NRC's receipt and use of any seismological information, including seismicity and probabilistic analyses of seismicity, in making the deterministic judgments about 3
tectonic province boundaries and maximum potential earthquakes I
within such tectonic provinces.
This distinction is important in this case because the NRC Staff and Applicant are in disagreement over the proper tectonic province which should be used to establish the Safe Shutdcwn Earthquake for Midland.
As the testimony of Applicant's witness, Richard J. Holt, demonstr,
- , the NRC Staff's " deterministic" selection of an unreasonably large region of the North American continent called the Central Stable Region as the tectonic province for Midland is unwarranted and inappropriate, not only on the basis of
-geology, but also when seismicity, and formal ptobabilistic l
analysis of seismicity are considered.
The appropriate Safe Shutdowr Earthquake for Midland is undoubtedly smaller than SSE dictated by the Staff in the October 14, 1980 Tedesco letter l
l and used by Applicant in constructing the Site Specific Response Spectra shown in Holt Exhibits 1 and 2.
~
l i
l l
\\
B.
THE SITE SPECIFIC RESPONSE SPECTRA APPROACH USED IN THE MIDLAND CASE TO REPRESENT VIBRATORY GROUND MOTION DUE TO THE POSTULATED OCCURRENCE OF THE NRC STAFF'S SSE AT THE MIDLAND SITE IS MORE CONSISTENT WITH 10 CFR PART 100, APPENDIX A THAN IS USE O.' REGULATORY GUIDE 1.60 The second decision required by Appendix A following selection of a Safe Shutdown Earthquake is the determination of a response spectrum which repte 'ents the ground motion due to the occurrence of an SSE at the site.
]
Appendix A,Section VI(a) (1) states:
- u view of the limited data available on vibratory ground motions of strong earthquakes, it usually will be appropriate that the response spectra be smoothed design spectra developed from a series of response spectra related to the vibratory motions caused by more than one earthquake.
Thus, constructing response spectra is itself a two-step process:
first, a series of representative earthquake records mest be collected; and second, these records must be combined into "smcothed design spectra" in some way.
One method acceptable to the Staff for representing l
vibratory ground motion at the site due to the SSE is use of the spectral shape contained in Regulatory Guide 1.60, scaled to an appropriate assumed zero period acceleration value.
(See l
l Holt testimony at pp. 5-6).
Use of Regulatory Guide 1.60 is however not required by 10 CPR Part 100, Appendix A.
In fact, l
in accordance with long-standing NRC practice Regulatory Guide 1.60 itself states:
Regulatory Guides are issued to decribe and make available l
to the public methods acceptable to the AEC Regulatory l
staff of implementing specif ic parts of the Commission's 1
I
.- ~,., _., - _. - -,..,,. - ~.
,m.
, _ ~ _ _. -...
_...c
10 -
regulations, to delineate techniques used by the staff in evaluating specific problems or postulated accidents, cr to provide guidance to applicants.
Regulatory Guides are i
not substitutes for regulations and compliance with them is not required.
Methods nd solutions different from those set out in the guides will be acceptable if they provide a basis for the fincings requisite to the issuance or continuance of a permit or license by the Commission.
See also Petition for Emergeacy and Remedial Action, CLI 78-6, 7 NRC 400, 406-7 (1978).
Not only is use of Regulatory Guide 1.60 not required, its use is arguably inconsistent with Appendix A.
This is because the Regulatory Guide 1.60 shape is based on a hodgepodge of strong motion recordings (accelerograms, : sken on widely varying local site geologic conditions, froh. rock to soft soil.
Thus, the Regulatory Guide 1.60 spectral shape is site independent.
(See Holt testimony at pp. 5-6).
In contrast, Appendix A contemplates that the response spectra will be site specific.
Section IV(a) states in pertinent part:
l (a)
Required Investigations for Vibratory Ground Motion.
The purpose of the investigations required by this paragraph is to obtain information needed to describe the vibratory ground motion produced by the Safe Shutdown Earthquake.
The investigations shall include the following:
(1)
Determination of the lithologic, stratigraphic, hydrologic, and structural geologic conditions of the site and the region surrounding the site; (2)
(3)
Evaluation of physical evidence concernint
a behavior during prior earthquakes of the surficia2 geologic materials at the substrata underlying the site from the lithologic, stratigraphic, and structural geologic studies; l
=
1 11 -
(4)
Determination of the static and dynamic engineering prcperties of the materials underlying the site.
Included should be properties needed to determine the behavior of the underlying material during earthquakes and the characteristics of the underlying material in transmitting earthquake-induced motions to the foundations of the plant, such as seismic wave velocities, density, water content, porosity, and strength.
(5)
...The comparative characteristics of the material underlying the epicentral location of region of highest intensity and of the material underlying the site in transmitting earthquake vibratory motion shall be considered.
See also 10 CFR Part 100, Appendix A, Section V(a) (1) (iv)
(third sentence).
The site specific response spectra methodology used for Midland closely conforms to the provisions of 10 CFR Part 100 Appendix A quoted above.
As Mr. Holt explains, the method involves searching out earthquake ground motion records which are similar to those which would be produced by the occurrence of the SSE at the Midland site.
One of the important criteria used in making this selection is the similarity of geologic and foundation conditions at the earthquake recording stations to those at the Midland site.
(See Holt Testimony at pp. 6-8).
No attempt is made to match earthquake records to site conditions when the Regulatory Guide 1.60 spectral shape is used, and therefore the site specific response spectra approach employed in this case is more in compliance with 10 cFR Part 100, Appendix A than is Regulatory Guide 1.60.
i l
l l
^
] l The final step in constructing a response spectrum is the combination of the selected series of earthquake records into a single " smoothed" response spectrum.
10 CFR Part 100, Appendix A, does not specify how this is to be done.
It does not require the use of deterministic methods.
Moreover, it is clear that the statistical process of combining earthquake records results in a probabilistic statement.
That is, the only meaningful interpretation of a response spectrum constructed by combining many different individual earthquake th records at, for example, the 84 percentile is the following:
assuming the postulated earthquake occurs at the site, the resulting vibratory ground motion is expected to fall within this response spectrum 84% of the time.
This is the only step in the Site Specific Response Spectra approach used in tnis case which is explicitly probabilistic.
It is worth I
noting that it is no more probabilistic than use of the Regulatory Guide 1.60 spectral shape, which was constructed in i
a similar way.
The Staff's suggestion in the October 14, 1980 Tedesco th letter (Holt Exhibit 3) that the 84 percentile spectral I
level be used has no basis in 10 CFR Part 100, Appendix A and is essentially arbitrary.
In 1980 the NBC issued operating l
licenses to Tennessee Valley Authority in Sequoyah Nuclear Plant, Units 1 and 2, even though the design spectrum fell th th betw3en the 50 percentile and the 84 percentile site i
~
specific response spectra constructed for Sequoyah using a method similar to that used for Midland.S/
J As Applicant's witness Mr. Holt suggests (pp. 17-18), it is impossible to make an informed judgement about what spectral level should be used for a response spectrum without taking into account the overall probability that the level of ground motion represented by that response spectrum will occur at the site.
Without taking into account overall probabilities, it is impossible even to say whether a particular choice is
" conservative."
Neither 10 CFR Part 100, Appendix A nor statistics provides a " deterministic" procedure for fixing the spectral level.
As described in the testimony of Mr. Holt (pp. 17-18), the relatively low seismicity of the Michigan Basin, and the results of probabilistic analysis of that seismicity, demonstrate that there is relatively less seismic hazard at the Midland site than at other sites outside the Michigan basin.
Such probabilistic considerations would justify the use of a lower spectral level for the Midland Site Specific Response i
l Spectra than for sites in areas of greater seismic hazard.
l III. CONCLUSION l
The testimony of Richard J. Holt demonstrates that the l
f Midland Site Specific Response Spectra have been appropriately l
developed in accordance with lu CFR Part 100, Appendix A and i
l 4/
See U.S. NRC, " Safety Evaluation Report related to operation of Sequoyah Nuclear Plant Units 1 and 2, Tennessee Valley Authority, Supplement No. 1" (Fe br uary, 1980), Docket l
Nos. 50-327, 50-328.
_--.______-___.__._.___-.-__.__m 1
1 c.
l f
i.
l i
with the NRC Staff's dictates as expressed in the October 14, t
j 1980 Tedesco letter (Holt Exhibit 3).
Although these SSRS are extremely conservative, they will be used for remedial l
underpinning work and in assessing existing seismic margin for i
the Midland Plant.
i i
i Respectfully submi,tted, 1
,??R okg' a>
One of the Attornqys for Consumers Power Company Isham, Lincoln & Beale I
Suite 4200 One First National Plaza Chicago, Illinois 60603 (312) 558-7500 I
i 1
1 3
i 5
)
1
_ _ _ _. - _,.. _ - _ -, - _.. - - - - _.., - _ ~. _ _ _ _ _,... -, _ _ - - _ - _. _ _ _., _. _, _ _, _ _ _.., _ _ _, -,., _ _. _ _... _