ML19254D678
| ML19254D678 | |
| Person / Time | |
|---|---|
| Site: | Vallecitos File:GEH Hitachi icon.png |
| Issue date: | 08/09/1979 |
| From: | Slemmons D AFFILIATION NOT ASSIGNED |
| To: | Rolonda Jackson Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19254D659 | List: |
| References | |
| 790927, NUDOCS 7910290334 | |
| Download: ML19254D678 (4) | |
Text
.
Appendix C DAVID B. SLEMMONS CCNSULTING GECLCG;ST 2995 GCLOC*t VALLEY ACAO. RENO. NEVACA 395C6 CC2) 972 8877 August 9, 1979 Robert E. Jackson, Chief Geosciences Branch h
Division of Site Safety and QU Environ = ental Analysis
@Q U
U. S. Nuclear Regulatory Co= mission
{U Washington, D.C.
20666
Dear Bob:
I have finished my review of the report, " Probability Analysis of Surface Rupture Offset beneath Reactor Building General Electric Test Reac ce," as prepared by EDAC, Inc.
This letter su=marizes my assessment of the methodology and basic data used for this report.
The EDAC report fo11cws the probabilistic approach of Cornell.
I believe that the basic seismologic cycle that appears to occur on many faults involves pregressive strain, earthquake-microearthquake build-up of activity, at times with smaller faulting events or creep, a main shock, and a decay pericd of decreasing activity.
Earthquake sequences on adjoining or connecting faults may vary the cycle.
This is a deterministic process and the probabilistic apprcach used for this report may not adequately represent the risk.
In addition, the numerical values used for the probabilistic analysis, although they are listed with three significant figures, are based mainly on assumed ages of older soils, en scattered radiccetric ages of the modern solum, and on li=ited field data.
The overall effect is that the results are presented with an aura of scientific accuracy and precision that =ay be misleading.
My cc=ments are reported under the three tcpics:
basic assu=p-tions, temporal relations, and displacement data.
3asic Assu=etions:
1.
The analysis assumes that there is adequate data for a probabilistic analysis using a Poisson =cdel.
The Poissen
=cdel assu=es spatial and temporal independence of all earth-quakes, and that an adequate, statistically valid data base is available.
Possible difficulties with both of these assu= ciens are not discussed.
Many faults exhibit nonrandem patterns' of activity in terms of locatien, si:e, and frequency of ccourrence, and result from determinis:1: processes cf gradual or sporadic strain accumula:icn, fell:wed by ce= plex episcdes of energy release.
The faul:ing may be in main episedes only, may have smaller f:re-and af ter-events, :: =ay be interspersed w;:n 1345 330
? age 1 of 4 7 gi.m agnt n :t'1'
,79102903I53
Robert E. Jackson Page 2 aseismic creep ~.
In complex =cnes the displacement may be triggered by or be coupled with displacement of other faults.
The earthquake size and amount of displacement may vary from event to event.
The appropriateness of a probabilistic approach with a Poisson distribution is not discussed.
2.
The basic parameters of ti=e since last displacement and/or recurrence interval and the a= cunt of that displacement must be adequately known for success of the probabilistic approach.
I do not believe that the present data base permits this type of analysis.
3.
It is assumed from temporal relations that the paleo-sols which were developed locally can be correlated directly with the sea level fluctuations of a worldwide nature, and this, in turn, is accurately, determined by marine oxygen isocese data of Shackleton and Opdyke (1973).
These corre-lacions and their error bands are not adequately established for this analysis.
4.
The analysis appears to assu=e that the three faults, B-2, B-1, 3-3 and H have separate histories.
The possibility of simultaneous distributed displacements on two or more fault strands connecting at depth or a single cumulative displacement on one strand has not been evaluated, nor have the analytical problems associated with possible randem-nonrandem patterns of cccurrence, or their nonranden effects on maximum possible displacement.
Ace of Solum and Paleosols:
1.
The mest important factor in the analysis is time.
Table 4.I su=marized, without discussion.cf the limitatiens of the detailed supporting data, the cbserved offsets in fee:
for Shears 3-2, E and B-1/3-3 for various assumed or inferred ages.
These values are most accurate for the modern solum, which has some radiccetric dates.
The age relatiens cannot be directly deter =ined, however, and the discussica of the age of this soil is given in terms of phrases dealing with mean residence ages, extent of centamination by "mcdern" carbon and with the understanding that the soil developed during a peried in which the cli= ate varied.
Samples frem apparently similar hori: ens have ages that differ by facters of almes: 2X.
The age of the paleoscis, listed as ranging frem 17,000 to 20,000, frem 70,000 to 125,000 and frem 123,000 cc 195,000 are determined by correlatica rather than by absciute dating techniques.
2.
The work of Roger Merrisen, presented orally a: the 1975 Geolcgical Scciety of America Annual Meeting in Salt lake Ci:y, suggested tha: there are de: ens of buried scils in the J
@S')[
1345 531 D
I D
- g
]-
m e c.
Robert E. Jackson Page 3 Lake Bonneville sedimentary sequence.
The full and detailed stratigraphic / climatic record with =any periods of moderate to strong soil formation suggest difficulties in making correla-tions with the =arine record, unless absolute dates are avail-able.
The correlation used in the probability analysis makes use of dates listed with up to three significant figures.
The assumed dates =ay be valid, but no quantitative verification is possible at this time and the errors =ay be large.
Displacement Data:
1.
The data of Table A-1 suggest that the most recent of fset on the B-2 shear, which underlies the GETR site and has a icw-angle reverse-slip =echanism, has an average displacement of 1.8 ft. and a maximum displacement of 3.0 ft.
The known tendency of reverse faults to rupture with new, short-cut paths (Sle= mons, 1977, p. A33 to A37) suggests the potential for future faulting in the 1,320 f t.
interval between B-2 and B-1/B-3 faults.
The 72 fcot width of the reactor building and the 80+ ft. displacement of the paleosols suggest a reccurrence interval measured in thousands of years and a higher risk than proposed by the analysis.
The data presented by Shie=on in the Earth Sciences Associates 1979 report shows that the earlier events could have displacements larger than that shcwn by the youngest offset, although = ore frequent, but smaller, displace-ments are not precluded.
The potential for rupturing at the GETR is considered in formula 5-14 for ti=e equal to 128,000 to 195,000 years, with an annual probability of an offset beneath tye reactor building that is less than 4.5 X 10~' and 3.0 X 10
, respectively.
The greater than 80 f t. displace =ent of the B-2 fault with its icw dip under the site suggests the pcssibility of rupturing by shcrt-cut paths through the GETR site.
The average displacement shcwn by the B-2 trenches appears to be about 1.8 ft.
Sc==ary:
1.
I believe that the deterministic =echanism that applies to fault parameters and associated earthquake si:es is co==only accompanied by nonrandem patterns of activity in terms of location, si:e, frequency of occurrence and time histories of activity.
I believe that a prebabilistic approach that assumes a
poisson distribution =ay not be appropriate for risk analysis of the potential for surface rupturing.
2.
The use of a probabilistic approach requires adequate and accurate nu=erical data and a basic understanding of the character Of the problem.
The present analysis is deficient in several importan: aspects:
Robert E. Jackson Page 4 (1) There are no definitive dates, or dates that are expressed with an accurate appraisal of the errers involved.
No accurate dates are available for the paleosols; the dates used are asst =ed and may be in error in the first signifi-cant figure.
(2) The only exact displacement for the faults is the = cst recent event, with separations of up to 3.0 f t.
The number of offsets and amount of displacement for all of the paleosols is not known.
Larger offsets are possible.
(3) The cumulative separations of 3-2 and 3-1/3-3 shears are minimal valurs and the true offsets are greater than 80 and 40 ft.,
.espectively.
These displacements, if caused by 3 ft. df placements during each event, L= ply much shorter recurrence intervals and much higher risks than the results of the analysis suggest.
(4) The =ede of origin of the shears has not been resolved and the geometry of the faults, their relations at depth and their relationship to the major faults of the region are all uncertain.
These relations affect the possibility of whether or not "short-cut" faults =ay eventually rup-ture the GETR site, and determine the type of probabilistic
=cdelling assumptions that should be made to assess the potential for surface faulting at the GETR site.
In su==ary, I do not believe that adequate data is available for the use of the probabilistic approach used in the ICAC report.
Sincerely yours, h'd 0 Nusurd z
David 3. Sle==ons Consulting Geologist 1345 333
Appendix 0 staff 07 CAUFCtNIA *NE tt3CU9C25 ACENCF
' E2muMO G. IscwN Jt "-
OfPAffkiNT CP CONSERVAf1CN CIVISION OF MINES AND G~dCLCGY civisicN Hr.AccuAnas f Ale aetNTH STRiff, ACCM f 343 BC2Am&NTO. f.A 95414 3
% etwo n August 16 1979 Mr. Racer: Jackson Chief, Geosciences 3 ranch E-b U.S. Nuclear Regulatory Cecnission
'.f ~
Wasnington, D.C. 20555 1
~
p Cear Bob:
The staff of the California Division of Mines and Geology (CONG) has reviewed all of ne submitted rescr:s regarding the General Electric Tes: Reacter (GETR) at ce Vallecites.'fuelear Center near Livernere. Enclosed is a copy of Freprint Special Publication 56, Geologic Evaluation of tne General Electric Test Reac:ce 3ite, Vallecitos, Alameda County, California, which reccr:s our findings. As you knew, staff have also been present in a nuscer of technical discu:sions regarding the geology and seismicity of the site and have examined trenches wnich have been excavated by the licensee at the request of ue U.S..'luclear Regulatory C ::nission (NRC). We nave the folicwing ceservattens:
1.
~he thrust dis:lacement alcng faults near the GETR facility can be inter;reted as tectonic off-set er landslide. It is the staff's finding that while neitner interpretation can be definitively confirnec, landsliding is a = referred c:nclu-sien. Please note that in this letter and in the recort, dis lacements are called faults, regardless of the interpretatien of ne mechanism of their crigin.
2.
Rece:itive displacement is evident along the faults which have been ex;:csed in ce trenenes. *he cumulative effect of these dis lacements is significant.
Displacement is considered to have cc:urred before and during Holocene time.
3.
- is a staff c:nclusien that ac;roximately nree feet of surface dis:Tacement at One reacter site recresents a c:nservative judg=en as to the ground ructure wnica mignt be associated with either the landslice interpretation or tne tec::nic =ccels of Dr. 0.G. Merc.
a.
~he ;eak ground ac:eleraticn asscciated at ne site has :een generally
- nsicertd :y CD*G s:sff, assuming tota an ear acuake oc:urmnce at tne reac::r along 2e 10 kilcme:ar Verona faul: ::cstula ed by 3.3. Herd anc a 7.5 magni ude event alcng ne Calaveras fault, wnica is ac rcximately two miles dis: ant fecm Oe GE-~4 1ccatien. In our Opinien, a 0.3g peak ac:elerati:n shculd be acecuate for plant cesign.
- f /cu 7 ave any cuestions regarcing :ne ::ntan: of ne re:cr. Or my emarxs
- nce Ming it, please Oc net nesi:a:e :o c:ntac: te.
/*
Si ncerely',
U gO
,h&m h\\
. l Lees : :w s
/
ita a Iac':gis:
1 5 3M isessoo M / o n
a
PREPRI?:T SPECIAL PUELICATI0tl 56 GEOLCGIC EVALUATICil CF THE GGERAL ELECTRIC TEST RUCTOR SITE VALLECITOS, ALAt4 EDA COUtiTI, CALIFCR!ilA Sy l
Salem Rica 2
~
Elgar Stephens 3
Charles Real August,1979 CALIFOR?lIA DIVISICtl CF tiI!!ES AtiD GEGLOGY Resources Suilding, Rec:a 1241 1416 tiinth Street, Sacrs=en:c 95314
+
1.
Gec1cgist, San Fr:ncisca Distric: Offica, 014G
~
2.
Seclegis:, Sacrsmento istric: Offic-ev 3.
Ieis=clegis:, Iacramento Distric: Of
.m DUPLICATE DOCUMENT j'
zzq Entire document previously entered into system under:
ANO r
u
- r No. of pages:
_