ML20136E423
Text
-
er ,'
- D D Q
!@ MC o, UNITED STATES
! \* ' ,o NUCLEAR REGULATORY COMMISSION
.- A(/ ,E WASHINGTON, D. C. 20555
\[* 0 /
pp 4 1983 1.DVISORY COMMITTEE 0N REACTO3 SMEGUFRDS, U.S.N.R.C.
APR 4 1983 MEMORANDUM FOR: Commissioner Ahearpe au h8J,1011,12,12 i 3,4[
FROM: William J. Dircks Executive Director for Operations A
SUBJECT:
ACRS COMMENTS ON SEISMIC DESIGN MARGINS O
M WC-This is in response to your memorandum of January 24, 1983 regarding ACRS concerns related to seismic design margins. As used in our gg discussions with ACRS and throughout this memorandum the terms seismic design margin or seismic margin refer to the inherent additional capacity of structures, systems and components above that required to withstand the earthquake specified for design purposes.
In their letter of January 11, 1983 the ACRS recommended a number of [Mh7 fp programs that address both the development of an improved capability for f estimating the severity of earthquake induced ground motion which may have a return frequency in the range of 10-4 to 10-5 per year at reactor sites and the likely capability of structures, systems and components as currently designed to withstand these ground motions. Significant activity has taken place, or is currently underway, in each of the areas recommended by the committee. We view the ACRS letter as a call for consolidation of these efforts into a systematic appraisal of seismic ~
risk. We concur that such an effort is timely. The committee anticipates that the programs constituting their recommendations could be conducted during the next 2 to 4 years. We believe that this is an appropriate time frame.
The specific committee recommendation with regard to geology, seismicity and geophysics calls for continued strong support of ongoing programs.
We concur in this recommendation and, as discussed in the attached remarks, we believe that we have in place a consolidated program in the geosciences that fulfills in large measure the committee's recommendations.
The committee recommendations with regard to generic studies of seismic risk, seismic review of specific plants, and seismic safety research, represent a group of acnivities that cut across every engineering discipline involved in seismic design and involve considerable industry participation. We concur in principle with the committee's RHERENCEDPDCi?.1EHI-RE!!"UD RE E pycg a d!E CUZTIEE-dei.13. L 'II,a.::Z.T Rt O Rc. h ht A_ rand Ext. 277:
Ph p w yx y w-a b .. xjb-s Vg QB Ua a """ - 0mc ok P -
Qodemoveem;@4pge,_ - %_
I '
l
( Comissinrer Ahaarne recomendations. Powever, the recomendations cre necassarily ceneral and extensive discussion with the appropriate ACPS subccanittees will be necessary to define the specific prograns thet are feasible considering tho availability of data and rascurces.
EPR will take the lead in plannitig and coordinating activities relr.ted to seisnic naroin study prograns. A neetina with the ACES Subccmittee en Extrene External Pheronenr.n is currently scheduled for aid April, 1983. The staff and industr.v representatives will be prepared to
! discuss the censcience prc9rens now contenplated or underway. He shall also be nrnoered to discuss the scope of prepnsed prenrons to address other aspects of the ACRS concerns releted to seisnic narcin studies.
Additional staff coments on the ACRS recomendations for scisnic margin l studies are attachrd.
@WHRas !.Dhtk5 Millian .1. Dircks Executive Pirector for Operations
Enclosure:
As stated cc w/ancl:
Ch?.irran Pelladino Comissionar Gilinsky Comissioner Poherts Comissiorer Asselstine OPE OGC SECY DISTRIBUTION:
WDircks JKnight JRoe LShao TRehm WAnderson GCunnir.gham JRichardson ACRS RJackson Htenton MGarver (ED0-12702) & (12663)
EC.ase EGoodwin Central File HBerkson DE:AD/CSE Rdg. File PPAS
-DEisenhut TSpeis RMattson *DE:AD/CSE HThompson JPKnight PCheck 3/2s/83 *SEE PREVIOUS CONCURRENCE aseyer cmen .R,y,o,1,1,m,ey, ,,,,,, ,*,g,S,,,,,,,,,,,,,,,, , ,,* D E,,,,,,,,,, ,,,,, ,
,
- N ,R R,, ,,, ,,, ,,,,,,, ,,N.RR .. .... .. ....... . ..li R R. ............ ... E 00. .... ...... .......
.um- > ....................... .. .L,S,h,ag,,,,,,,,,,,, ,,,R,H,yp,],],m,e r,,,, ,,,,
Ggg dw1,n,.... ..$.CU.9........... ..liRQRntan...... WDir.cks........
em, ........................ .342s/,8,3
,,,,,3L2 sg 3,,,,,, ,,,,,,3,/,,2,s,/,s,3.,,,, ,,3t ,g,3 ,,,,,,,, ,,3,f ...g,3...,,,,, ,,3 f ..,,f 8,3..,,,,...
cr ee:.. ....n.-== nun . nFFiCIAI RECORD COPY w.m.ai w-
. I i-o Additional Staff Comments r
on ACRS Recommendations for Seismic Margin Studies i
The need for identification and quantification of seismc design margins was recognized by NRR in 1976. At that time new seismological data indicated that the ground motion employed to characterize earthquakes specified for some plant designs may have been underestimated. The Office of Nuclear Reactor Research (RES) initiated the Seismic Safety 1
Margins Research Program (SSMRP) with the two fold purpose of developing '
methodology for producing quantitative seismic risk estimates and, as a by-product of the risk studies, producing more quantified estimates of 1
margins in some plant structures, systems and components than were l previously available. The SSMRP was initiated at the beginning of FY 1979 and is scheduled for completion in FY 1984. Total costs for the program will be approximately 12 million dollars. -
The methodology developed in SSMRP and subsequent industry seismic risk studies (most of which adopt or adapt some elements from SSMRP) constitute the state-of-the-art in seismic risk analyses for nuclear power plants. The results of these efforts have enabled the scientific and engineering communities to gain significantly better insights regarding seismic risk at nuclear power plants and provided the bases for systematic appraisal of the seismic design process through probabilistic risk assessments (PRA). Lack of data in the areas of structural and component fragility, i.e., actual failure modes and failure levels, and the need to extrapolate a small seisaological data base to very long recurrence intervals, have resulted in large
4
_2_
uncertainties in absolute risk estinates in seismic PRA's conducted to date. The_immediate value of the PRA's performed to date is for comparison of relative seismic risk values at various sites and identification of possible large contributors to the total seismic risk at a given site. Numerical results of these analyses alone are not presently considered sufficiently reliable to be compared with the 4
results of internal event PRAs nor to d'etermine if the sources of relatively higher risk should be upgraded. Present SSMRP methodology requires simplification in order to be useful for broad scale application. This effort is underway as one of the final program activities scheduled for completion in FY 1984.
With respect to seismic hazard, the basic problem is that there is not a long' historical record of large earthquakes. There is only one recorded earthquake in the eastern U.S. of magnitude greater than 6, the range of most damaging earthquakes. There are a few recorded earthquakes in the
. central U.S. of magnitude greater than 6. Most instrumental information we have for eastern and central U.S. seismic events is at relatively lower magnitudes, intensities, and accelerations.
Consequently, the critical parameters for larger earthquakes in the eastern U.S., including the probability of their occurrence, cannot be estimated at this time without wide bands of uncertainty. In the western U.S., where the earthquake activity is higher, there is a more substantial accumulation of data and this includes large magnitude earthquakes.
3
i '
- Examination of seismological studies done over the past several years indicate _that past practice, including application of Appendix A to 10CFR100,'has resulted in safe shutdown earthquakes whose peak acceleration have probabilities of exceedance on the order of 10-3 or 10-4 per' yea r. Our current seismological programs are directed towards making better estimates of these probabilities of exceedance, as well as determining the existence of outliers, if any. Additional studies are focused towards determining the validity of numerical estimates at low probability levels which require extrapolation orders of magnitude beyond the existing data base. The current state-of-the-art does not permit us to make' meaningful estimates of ground motion exceedance at probabilities less than about 10~4 per year; this is due primarily to the small data. base discussed above and to limited understanding of the causal mechanisms of earthquakes in the eastern U.S.
NRR, in conjunction with RES, has undertaken a probabilistic project entitled, " Seismic Hazard Characterization for Nuclear Power Plants in the Eastern U.S." which should increase our ability to estimate the seismic hazard for nuclear facilities. This project is designed to produce seismic hazard curves for all eastern and central U.S. plant sites. These seismic hazard curves will express the probability of exceedance of the Safe Shutdown Earthquake.for each site. Completion of this effort is presently scheduled for FY 1984.
Geologic and geophysical studies designed to better understand the
'causes of large earthquakes, such as the Charleston, South Carolina,
~
( e l l
7 earthquake in the easterr seaboard region, are also underway. To date no generally accepted association between eastern seismicity and crustal structure exists. The approach is to emphasize studies in areas of relatively higher. seismicity in the eastern seaboard at the depths of earthquake occurrence in order to determine if tectonic features and processes responsible for seismicity can be identified and correlated.
The areas are: Charleston, South Carolina; Giles County, Virginia; Central Virginia; Ramapo Fault Area (New Jersey and New York); and New England. The initial results of tFese studies are scheduled to be avail'able in FY 1985.
The seismic margins that result from application of current design and construction practices have been evaluated for various features of most plants licensed over the last several years. One of the most --
compreher..ive studies to date is being performed by consultants to Consumers Power Company for the Midland plant, the subject of a June 3, 1982 ACRS letter. These seismic margin studies have generally demonstrated by analysis substantial margins in all major plant structures, systems and components. Margin studies for electrical and mechanical equipment have been based in part on plant qualification test data and on available data. developed in military and other non-nuclear programs. These equipment margin studies generally support the judgement
't that significant seismic margins can be expected in nuclear plant 9 equipment.
Studies of actual earthquake experiences in power plants and industrial facilities are being proposed under the sponsorship of the Seismic
. Qualification Utility Group (SOUG). The prinary purpose of these studies is to provide an alternative to specific seismic qualification test' programs for safety related equipment in operating plants; that is qualification by reference to past performance. A SQUG pilot study has ':
added significantly to the confidence level of the judgement that substantial margins are present but is of limited use from the standpoint of providing fragility da.ta because few, if any, failures of typical industrial equipment have occurred under. actual earthquake conditions.
The pilot program for the SQllG historical performance studies was l
completed in October 1982. Staff comments on the pilot program were issued in December 1982. A meeting between the SQUG and the staff is
~
scheduled for late March 1983 to discuss the most effective means of i
extending this program and the possible role in overall seismic margin studies.
As noted in the January 11, 1983 ACRS letter the NRR staff and the ACRS have been discussing the general issue of seismic safety margins in the j context of specific case reviews. ACRS recommendations for study of various aspects of seismic ~ margins have been made during a number of l
these licensing reviews and have been included as license conditions.
In the closing paragraph of the January 11, 1983 letter, the ACRS notes that a generic study such as recommended, given sufficient priority, could be,an appropriate substitute ior case specific studies of seismic design margins. The staff also believes that most aspects of seismic
, ( (
, , : := -
. i contribution to risk, including seismic design margins, are most l
appropriately studied on a generic basis. We plan-to take action, wherever feasible, to make.the staff and industry efforts related to case specific studies an integral part of a more generic program.
s kb g
,* ,;..... ( +
- UNITED STATES 3 1 NUCLEAR REGULATORY COMMISSION j $ WASHINGTON, D. C 20$55
. e
\g / APR 4@
MEMORANDUM FOR: Commissioner Ahearne FROM: William J. Dircks -
Executive Director for Operations
SUBJECT:
ACRS COMENTS ON SEISMIC DESIGN MARGINS This is in response to your memorandum of January 24, 1983 regarding ACRS concerns related to seismic design margins. As used in our discussions with ACRS.and throughout this memorandum the terms seismic design margin or seismic margin refer to the -inherent additional capacity of structures, systems and components above that required to withstand the earthquake specified for design purposes.
In their letter of January 11, 1983 the ACRS recommended a number of programs that address both the development of an improved capability for estimating the severity of earthquake induced ground motion which may have a return frequency in the range of 10-4 to 10-5 per year at reactoi-sites and the likely capability of structures, systems and components as currently designed to withstand these ground motions. Significant activity has taken place, or is currently underway, in each of the areas ,
recommended by the committee. We view the ACRS letter as a call for consolidation of these efforts into a systematic ^ appraisal of seismic -,
risk. We concur that such an effort is timely. The committee anticipates that the programs constituting their recommendations could be conducted during the next 2 to 4 years. We believe that this is an appropriate time frame, The specific comittee recomendation with regard to geology, seismicity and geophysics calls for continued strong support of ongoing programs.
We concur in this recommendation and, as discussed in the attached
. remarks, we believe that we have in place a consolidated program in the geosciences that fulfills in large measure the committee's recomendations.
The committee recomendations with regard to generic studies of seismic risk, seismic review 'of specific plants, and seismic safety research, represent a group of activit
- that cut across every engineering discipline involved in seism.. _ sign and involve considerable industry participation. We concur in prnciple with the committee's I
CONTACT:
J. P. Knight, NRR -
Ext. 27733
~ - - - - - . __
( (
Comissiorer Ahearre
-2 recomendations. However, the reccmendations are necessarily ceneral and extensive discussion with the appropriate ACRS subccanittees vill be necessary to define the specific programs thet ara feasible considering the availability of d,,ta anc' resources.
f:PR will take the lead in n1anning and coordinating activities related to seisnic naroin study prograns. A reetino with the ACRS Subccrnittee en Extrere External Phercrenen is currently scheduled for aid April, 1983. The staff and industrv representatives will be prepared to discuss the geoscience programs now contemplated or underway. We shall also be prepared to discuss the scope of prepnsed prograns to address other aspects of the ACRS concerns related to seismic margin studies.
Additional staff ccments on the ACRS recorrendations for seismic margin studies are attached.
O:4 Y41P:2L5 c
i W '. nircks l
E.v c - - T ctor #cr Orcr tions
Enclosure:
As stated cc w/anc1: ;
Chairran Felladino f Cemissinrar Gilinsky i l
Comie niomr Poher" Comit:iorer Assc OPE OGC ,
SECY DISTRIBUT .
WDirc JKnight l JRoe LShao TRe" WAnderson GM ; ham JRichardson A^RS RJackson l HDenton MGareer (EDO ' '?; & (12663)
ECase i I
EGoodwin Centr:1 File HBerkson DE:AD/CSE Rdg. F: e PPAS DEisenhut TSpeis RMattson *DE: AD/CSE HTho= an JPKnight PChec 3/25/83 *SEE PREVMUS COElRENCE "5"yt
,, , , c , ,
RVo11n - *RES *DE *f m NRR _ . ..NRR. ... . E D 0. .............. .
summe ) .... . . . . . . . . .
L5ho RHVollr r EEcedwt . .J.G.C.4.51_ . ..HRQe.ntan.. .. WJDir.cks.... ..
.3/25483. .... 342 sr 3,,... .... 3,gs/13 3L.,Las 3L ./.9.1. . 2 L ./.8.a.._.
on> . . . . . . . . . . ...
W_ m.o Q . m u.co OFFiri AI. RECORD COPY .
uan= =*-==
( (
Additional Staff Comments on ACRS Recommendations for Seismic fiargin Studies The need for identification and quantification of seismc design margins
.was recognized by NRR in 1976. At that time new seismological data indicated that the ground motion employed to characterize earthquakes specified for some plant designs may have been underestimated. The Office of Nuclear Reactor Research (RES) initiated the Seismic Safety Margins Research Program (SSMRP) with the two fold purpose of developing methodology for producing quantitative seismic risk estimates and, as a by-product of the risk studies, producing more quantified estimates of '
margins in some plant structures, systems and components than were previously available. The SSMRP was initiated at the beginning of FY 1979 and is scheduled for completion in FY 1984. Total costs for the prc; ram will be approximately 12 million dollars. --
The methodology developed :in SSMRP and subsequent industry seismic risk studies (most of which adopt Ol' edapt some elements from SSMRP) constitute the state-of-the-art in seismic risk analyses for nuclear power plants. The results of these efforts have enabled the scientific and engineering communities to gain significantly better insights regarding seismic risk at nuclear power plants and provided the bases
~
for systematic appraisal of the seismic design process through' probabilistic risk assessments (PRA). Lack of data in the areas of structural and component fragility, i.e., actual failure modes and failure levels, and the need to extrapolate a small seismological data base to very long recurrence intervals, have resulted in large 4
_. .: . . ; - . .. .. -- - - -- - --- - - - - - - - - - -- - - - - - - ~
. ,- . t (
uncertainties in absolute risk estinates in seismic PRA's conducted to date. The immediate value of the PRA's performed to date is for comparison of relative seismic risk values at various sites and identification of possible large contributors to the total seismic risk at a given site. Numerical results of these analyses alone are not presently considered sufficiently reliable to be compared with the.
results of internal event PRAs nor to determine if the sources of relatively higher risk should be upgraded. Present SSMRP methodology requires simplification in order to be useful for broad scale application. This effort is underway as one of the final program activities scheduled for completion in FY 1984.
With respect to seismic hazard, the basic problem is that there is not a long historical record of large earthquakes. ~ There is only one recorded earthquake in the eastern U.S. of magnitude greater than 6, the range of most damaging earthquakes. There are a few recorded earthquakes in the central U.S. of magnitude greater than 6. Most instrumental information we have for eastern and central U.S. seismic events is at relatively lower magnitudes, intensities, and accelerations.
Consequently, the critical parameters for larger earthquakes in the eastern U.S., including the probability of their occurrence, cannot be estimated at this time without wide bands of uncertainty. Ir. the western U.S., where the earthquake activity is higher, there is a more substantial accumulatic of data and this includes large magnitude earthquakes..
- . - . - . - . . - - . = _ - . . - . - . . - .
, .,.. s
( (
Examination of seismological studies done over the past several years indicate that past practice, including application of Appendix A to 10CFR100, has resulted in safe shutdown earthquakes whose peak acceleration have probabilities of exceedance on the order of 10-3 or 10-4 per year. Our current seismological programs are directed towards making better estimates of these probabilities of exceedance, as well as determining the existence of outliers, if any. Additional studies are focused towards determining the validity of numerical estimates at low probabil_ity levels which require extrapolation orders of magnitude beyond the existing data base. The current state-of-the-art does not permit us to make' meaningful estimates of ground motion exceedance at probabilities less than about 10-4 per year; this is due primarily to the small data base discussed above and to limited understanding of the causal mechanisms of earthquakes in the eastern U.S. "
NRR, in conjunction with RES, has undertaken a probabilistic project entitled, " Seismic Hazard Characterization for Nuclear Power Plants in the Eastern U.S." which should increase our ability to estimate the seismic hazard for nuclear facilities. This project is designed to produce seismic hazard curves for all eastern and central U.S. plant sites. These seismic hazard curves will express the probability of exceedance of the Safe Shutdown Earthquake for each site. Completion of this effort is presently scheduled for FY 1984.
Geologic and gecphysical studies designed to better understand the ,
causes of large earthquakes, such as the Charleston, South Carolina,
- .-.- - . . e . . . . - - . . . . --p.- - . . . .
. - . . . . ~ . . - - .--..._..............-..r.. . .
.:.l-' , :. (
(
earthquake in the eastern seaboard region, are also underway. To date no generally-accepted association between eastern seismicity and crustal structure exists. The approach is to emphasize studies in areas of relatively higher seismicity in the eastern seaboard at the depths of earthquake occurrence in order to determine if. tectonic features and e processes responsible for seismicity can'be identified and correlated.
The areas are: Charleston, South Carolina; Giles County, Virginia; Central Virginia; Ramapo Fault Area (New Jersey and New York); and New England. The initial results of these studies are scheduled to.be available in FY 1985.
The seismic margins that result from application of current design and construction practices have been evaluated for various features of most plants licensed over the last several years. One of the most- --
comprehensive studies to date is being performed by consultants to Consumers Power Company for the Midland plant, the subject of'a June 8, 198'2 ACRS letter. These seismic margin studies have generally demonstrated by analysis substantial margins in all major plant structures, systems and components. Margin studies for electrical and mechanical equipment have been based in part on plant qualification test data and on available data developed in military and other non-nuclear programs. 'These equipment margin studies generally support the fjudgement that significant seismic margins can be expected in nuclear plant equipment.
Studies of actual earthquake experiences in power plants and industrial facilities are being proposed under the sponsorship of the Seismic
. .-1
( (
9
- 5 Qualification Utility Group (SQUG). The primary purpose of these studies is to provide an alternative to specific seismic qualification test programs for safety related equipment in operating plants; that is qualification by reference to past performance. A SQUG pilot study has added significantly to the confidence level of the judgement that substantial margins are present but is of limited use from the standpoint of.providing fragility data because few, if any, failures of typical industrial equipment have occurred under actual earthquake conditions.
The pilot program for the SQllG historical performance studies was completed in October 1982. Staff comments on the pilot program were issued in December 1982. A meeting between the SQUG and the staff is scheduled for late March 1983. to discuss the most effective means of extending this program and the possible role in overall seismic margin studies.
As noted in the January 11, 1983 ACRS letter the NRR staff and the ACRS have been discussing the general issue of seismic safety niargins in the context of specific case reviews. ACRS recommendations for study of
- various aspects of seismic margins have been made during a number of these licensing reviews and have been included as license conditions.
In the closing paragraph of the January 11, 1983 letter, the ACRS notes that a generic study such as recommended, given sufficient priority, could be an appropriate substitute for case specific studies of seismic , ,,
design margins. The staff also believes that most aspects of seismic
...--.w..~, .__..g,,. . . . . . . . . ---.---v
___ . . . . .- ...e ~_...mn.- . .
- ;, . .' k
.~
(-
(
contribution to risk, including seismic design margins, are most appropriately studied on a generic basis. We plan to take action, wherever feasible, to make the staff and industry efforts related to case specific studies an integral part of a more generic program.
1 9
b e
- W e--.-, a