ML20005B192
| ML20005B192 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 05/08/1981 |
| From: | Dignan T PUBLIC SERVICE CO. OF NEW HAMPSHIRE, ROPES & GRAY |
| To: | NRC ATOMIC SAFETY & LICENSING APPEAL PANEL (ASLAP) |
| Shared Package | |
| ML20005B186 | List: |
| References | |
| NUDOCS 8107060448 | |
| Download: ML20005B192 (23) | |
Text
,
y UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION before the ATOMIC SAFETY AND LICENSING APPEAL BOARD
)
In the Matter of
)
)
PUBLIC SERVICE COMPANY OF NEW
)
Docket Nos. 50-443 HAMPSHIRE _et _al.
)
50-444
)
(Seabrook Station, Units 1 and 2)
)
)
CDN ex
.g ON REMAND PROM THE DOCWED g
COMMISSION WITH RESPECT th USNRC TO SEISMIC ISSUES MAY 111981,
'I G
Office of the Secretar; Dxketing & Senice BRIEF AND PROPOSED Bach p
FINDINGS AND RULINGS c>
U 0F PERMITTEES INTRODUCTION This submission of the permittees in the above-entitled matter is in two parts.
The first constitutes an argumenta-tive brief.with' respect to the issues.
This brief provides the analytis which forms the basis for the proposed findings and-rulings which the permittees request the Appeal Board to include in its decision.
The specific proposed findings and rulings are set out as Appendix A to the brief.
l 8107060448'81'05 PDR ADOCK 05000 3
0
_R
]
v
STATEMENT OF PRIOR PROCEEDINGS In its initial decision authorizing _ssuance of the Seabrook construction permits, handed down on June 29, 1976, LBP-76-26, 3 NRC 857, the Atomic Safety and Licensing Board held the seis-mic design criteria of Seabrook to be adequate.
3 NRC at 868-71, 919-22.
In so doing, the Licensing Board rejected the thesis of Michael A. Chinnery, Ph.D., that Seabrook's seismic design criteria should account for the possibility that an earthquake with an epicentral intensity of MMI IX in the " seismic zone" containing the Seabrook site had a probability of approximately 10-3 per year or greater.
3 NRC at 920.
The Licensing Board also rejected the contention of the intervenor NECNP that, even assuming that the proper SSE to be chosen for Seabrook was of an epicentral intensity of MMI VIII, the appropriate zero period acceleration to be used as the design criterion was 0.4g as opposed to the 0.25g actually selected by the NRC Staff.
3 NRC at 871, 921-22.
On July 26, 1977, this Appeal Board affirmed those findings and rulings' of the Licensing Board, ALAB-422, 6 NRC 33, 57-60, 62-64, over the dissent of then Member Farrar, 6 NRC at 106, 111-13, who reserved the right to set forth at a later date "the full reasoning underlying [his] position", id. at 106.
In affirming the Licensing Board, this Appeal Board rejected Dr. Chinnery's "probabilistic theory" as "both technically deficient and inconsistent with (10 CFR 100] Appendix A."
6 NRC at 60.
NECNP's 0.4g contention was rejected and the J s
0.25g criterion accepted as reasonable on the basis of the testimony of the applicant and Staff witnesses, 6 NRC at 62-63, and as being buttressed by data presented by Dr. Mihailo Trifunac, 6 NRC at 63-64.
NECMP petitioned the Commission for review of both of these holdings; the Commission deferred review pending receipt of the promised " full reasoning" of Member Farrar.
That " full reason-ing" was. issued two years later on August 3, 1979, ALAB-561, 10 NRC 410-36,1 and failed to persuade the majority, see 10 NRC 436-a--h.
One more year later, on September 25, 1980, the Commission issued'its order with respect to the NECNP petition for review.
CLI-80-33, 12 NRC 295 The Commission "[found]", 12 NRC at 297, the Chinnery " methodology" not to be inconsistent with Appen-dix A.
Next,-the Commission found that a greater exploration on the record as to the " factual validity of Dr. Chinnery's hypothesis" was required and directed this Appeal Board to
" reopen the record to take additional evidence on Dr. Chinnery's methodology and reconsider its opinion on this matter."
Id.
l.It is ironic to note that Mr. Farrar originally based his view that Seabrook's SSE should be MMI IX on the convergence of three lines of evidence:
(1) the Chinnery theory, (2) the Smith catalog assignment of MMI IX to the 1755 Cape Ann event, and (3) the supposed similarity of the Montreal geology to New
. England coupled with the MMI IX Montreal event.
6 NRC at 112-13 Since that time it is conceded by everyone with expertise i
that Smith overrated the 1755 event; and the Montreal event has been' officially downgraded by Canadian authorities to MMI VIII.
It is not. clear that absent the con".csence of the three lines, Member Farrar would have dissenten. J
o In addition, this Appeal Board was directed to " reopen the
. record to take more evidence on the consistency of Appendix A and staff's methodology for correlating vibratory ground motion
~
with the-SSE," and to-" reconsider its opinion on this matter".
12 NRC at 298.
With respect to this second matter the Commis-sion stated that:
"In particular, the parties should provide a discussion of the relation between the mean of the maximum ground accelerations and maximum effective ground acceleration."
Id.
In due course, a hearing was convened and held on April 6-9, 1981, at which the evidence summarized below was adduced.
SUMMARY
OF EVIDENCE A.
AS TO THE ISSUE OF THE FACTUAL VALIDITY OF DR. CHINNERY'S THEORY Dr. Chinnery again defended his theory that one can predict the return-times of earthquakes which are larger than any his-torically recorded in a given area by extrapolating from the return times of smaller earthquakes for a given time period in that area.
Chinnery, passim.
To do this he begins by deter-mining a return time for all but the largest earthquakes in his area of interest.
This is done by dividing the numoer of all earthquakes for a given intensity or higher into the number of years over which they occurred; this number is then plotted on semilog paper with the return periods on the abscissa and _________-_____D
intensities on the ordinate.
Tr. 313-16.2 He then fits a line with a slope of 0.57 through the data points, Chinnery 11.
In the case of the " seismic zone" which Dr. Chinnery selected for Seabrook this line fits two of the data points which he terms the "most reliable".
Id.
This line is then extended out to intensities which have never occurred historically and a conclu-sion is reached that the probability of an earthquake in this seismic zone having an epicentral intensity of MMI IX or greater is approximately 10-3 Chinnery 16.
The Chinnery theory or methodology is based upon four assumptions, viz:
(1) there exists a linear relationship between frequency and intensity; (2) that in all areas the appropriate slope of the line to be fitted to the data is 0.57; (3) that there exists no upper bound to earthquake size in any area; and (4) that it is valid to extrapolate frequency-intensity d_ta linearly to predict larger than historical earthquakes.
Chinnery 8-13; Chinnery Ex. 2 at 757; Chinnery Rebuttal 5-14; Reiter 3 Dr. Chinnery admits that in order for his results to be accurate, all four of these assumptions must be correct Tr. 90.
assumptions.
r On the other hand, as Dr. Reiter testified and Dr. Chinnery
^
admits, Reiter 10; Chinnery Ex. 2 at 769, there is a great deal of uncertainty associated with each of Dr. Chinnery's four 2 The largest. earthquakes historically are excluded from this exercise by the device of selecting a time period which excludes them from consideration.
Chinnery Ex. 2 at 761; Tr. 59, 60. - - -
~.
assumptions. LDr. Chinnery himself admits that there is no way to make a realistic assessment of the uncertainties and hence error involved.
Tr. 100 ASLB Tr. 4055-56; Chinnery Ex. 2 at 769 See also Reiter 10.
Finally, Dr. Chinnery finds support for his conclusions in the so-called TERA report by virtue of the fact that one-half of'the experts polled gave MMI > X as their high estimate of the upper bound for earthquakes in the region containing the Seabrook site.
In response to-Dr. Chinnery's testimony, evidence was adduced which contradicted the various assumptions.
It appears that in actuality exercises similar to Dr. Chinnery's in various areas have produced a wide range of slope values ranging as broadly as from 0.24 to 0.76.
Reiter 6.3 Permittees' witness Holt and two. Staff witnesses challenged the. validity of the "no upper bound" assumption saying that on a geological basis the existence of an upper bound is a reasonable assumption.
Jackson 3-9; Reiter 6-8; Holt 4.
In addition, Mr. Holt noted that in Sec' and England where the geology is similar to New England there are 1000 years of historical data; that no earthquake of epicentral intensity > MMI VIII has occurred; and that if the Chinnery method were applied to England the return time predicted for an event of epicentral intensity MMI > VIII would be a good deal less than 1000 years.
Tr. 401, 443-44 3 Dr. Chinnery's selection of a.57 slope was, as he testified, somewhat arbitrary.
See Appendix B hereto. -
J
~
1 Mr.(Holt-strongly contested the validity of extrapolation of frequency-intensity. data without reference to the specific-geology of the site involved.
Holt 1.
Mr. Holt testified that various' data showed;that1the probability of the occurrence of a given. intensity on the Seabrook site rock as compared-to soil is lower..by a' factor of 4 to a-factor of 15.
Holt 1, Figs. 1-2.
- He pointed out that historical accounts of major earthquakes
~
demonstrate the'correctitude of the assertion that intensities on rock are much lower for a given event than on soil, Holt 1, Age. 1; indeed,.the intensities on rock versus soil are any-where fromil to 4 intensity units less, Tr. 349-50.
These views of Mr. Holt are supported by the lack of any strong motion recordings associated with MMI VIII on rock _in the United' States.
N Reiter 23 Dr. Holt also demonstrated that by changing the temporal period..over which Dr. Chinnery's analysis is done for a given area, absurdly =short return times will be predicted for major events.
Holt 2-4.
Indeed, when pressed, Dr. Chinnery essentially admitted this would be the case for New England had he chosen-a time period which encompassed all of the reported
- large events in the-catalog he used.
See Tr. 60-61, 96-98.
Dr. Reiter, questioned the propriety of extrapolating in this case from only two data points, showing that if two points are to-be: considered valid for this prupose, one could, by
- utilizing Mr. Holt's data set, obtain significantly longer
- returnntimes for large ea."thquakes.
Reiter Rebuttal 4.
Dr. Reiter..also challenged Ihs. Chinnery's use of the TERA.
d -
J
r reports saying that the "best estimates" of the experts should be used as opposed to treating the report as though only the "high estimates" were~ relevant as Dr. Chinnery did.
Reiter Rebuttal 2-3 The "best estimates" of the majority of the experts were the equivalent of MMI VIII or less.
Chinnery 13 Similarly, a majority thought that MMI < VIII was the highest epicentral intensity probable for the next 150 years.
TERA Q. 2-5, Page II-23 Indeed, Dr. Chinnery agreed that it was just as valid for this Board to consider the "best estimates" and 150-year estimates as it was to consider the "high estimates" in its analysis of this case.
Tr. 70-74.
Mr. Holt, for the permittees, and each of the three Staff witnesses testified that in their judgment the uncertainty under-lying the assumptions and subjectivity involved made the Chinnery methodology " unacceptable" as opposed to "less preferable" for use in selecting nuclear power plant design criteria.
Tr. 319-401 (Holt); Tr. 573 (Jackson); Tr. 574-75 (Reiter); Tr. 579 (Knight).
Finally, both the Staff and. permittee witnesses pointed out that even if the Chinnery methodology was assumed valid, the present Seabrook design should still be deemed adequate and acceptable because of the numerous conservatisms that have already operated within the design process.
Holt 5-7; Jackson 9-12; Reiter 11-16.
_8_
J
r
~
B.
AS TO THE ISSUE OF THE STAFF'S METHODOLOGY FOR CORRELATING VIBRATORY GROUND MOTION WITH THE SSE-The Staff explained that it correlated vibratory ground motion with the selected SSE by utilizing, in the Seabrook case and for the last five years generally, the trend of the means relationship of Trifunac and Brady.
Reiter 17.4 The Staff also explained in detail why in the context of the entire design process (which was explained in detail in the record, Knight 3-8; Holt 5-7), use of the Trifunac and Brady relationship was both conservative and entirely in accord with Appendix A to 10 CFR 100.
Jackson 13-18; Reiter 16-22.
The Staff testified to the large degree of conservatism in the design, Knight 9-15; Reiter 25, as did Mr. Holt, Holt 5-6.
Indeed, it was the Staff's view that'the safety margin inherent in the Seabrook design was gre9 tar than 2.
Tr. 676, 720.
Seabrook's seismic design of the Reg.LGuide 1.60 spectral shape anchored to 0.25g is as high as; any seismic design specified for any nuclear power plant east of the Rockies.
Jackson 15.
Although admitting that, assuming an event which resulted in MMI VIII at the Seabrook site, there 4 This is not the same as saying that'the Staff approves a value equal to the mean of the observed peak accelerations observed in connection with F.MI VIII events.
In fact, the mean value of such observed events, as reported by Trifunac and Brady, is.169g.
See 6 NRC at 63-64.
A value of.255 (or 245 cm/sec2) equals the mean plus 0.935 standard devia-tions, a value which, on a normal distribution curve, captures more than 80% of all expected events (82.51%), and thus satisfies Dr. Trifunac's standard for "a reasonable upper bound.".J
F.
could be isolated acceleration peaks in excess of 0.25g;-the Staff explained why these were without engineering signifi-cance.
Knight 8; Tr. 719.
The Staff's overall conclusion as to the acceptability of
'the_Seabrook seismic design is uncontradicted.
Dr. Trifunac concluded that insofar as he had analyzed the problem, the design "may be acceptable"; the reason for the "may be" was that he needed to do additional work (which he did not do).
Trifunac 10.5 Dr. Chinnery adopted the Trifunac conclusion.
Tr. 46.
Dr. Trifunac characteri-zed the Staff methodology, assuming the correct SSE is MMI VIII, to be " conservative".
Tr. 794-95 He said if the SSE were a MMI X, it still would be conservative "in a crude fashion".
Tg. 796.
Finally, he deemed it " acceptable" under Appendix A.
Tr. 762, 797-98.
Thus, the evidentiary record is bereft of any legally signifi-cant conflict en this issue.
STATEMENT OF ISSUES 1.
Is there factual validity to the hypothesis espoused by Dr. Chinnery underlying his conclusion that the probability 5 Dr. Trifunac's analysis showed that, given Dr. Chinnery's hypothesis but limiting it to a maximum intensity of VIII, the probability that the Seabrook design criteria (i.e.,
the Reg. Guide 1.60 spectral shape anchored at 0.258)-'would be exceeded is substantially less than.05 Trifunac_, Fig. 3 Indeed, the probability of such an occurrence is.05 or less evan if Dr. Chinnery's. hypothesis is accepted together with his hypothesis of no upper bound on earthquake intensity.
Id.
The probability of exceeding the design criteria in the range of relevant frequencies (1-15 cps; period =.07 -
1.0 sec.) is even lower. -.
a
r of an-earthquake at the Seabrook site of MMI IX is 10-3/ year?
2.
Whether the Staff's methodology for correlating vibratory ground motion (acceleration) is consistent with the terms and requirements of 10 CFR.100, App. A?
ARGUMENT I.
.THERE IS NO REASON FOR THIS BOARD TO DEPART FROM ITS PRIOR RULINGS WITH RESPECT TO THE CONSISTENCY OF_THE' STAFF'S METHODOLOGY WITH APPENDIX A We address the second issue first because, in light of the evidentiary record, it is the most easily disposed of.
As noted above in the summary of evidence, there is no evidentiary conflict with respect tc tha question of whether the Staff's method of-correlating vibratory ground motion with the chosen SSE is consistent with 10 CFR 100, Appendix A.
No witness says the Seabrook design in this regard is not acceptable.
The Staff witnesses affirmatively say its conservative; Dr. Trifunac appears to endorse this conclusion.
A possible legal problem has been suggested by emphasizing that Appendix A talks in terms of " maximum vibratory accelera-tion" and admittedly the Staff methodology ignores peak accelera-
.tions on the basis that they are without engineering signifi-cance.- Such action is entirely consistent with Appendix A which requires only assurance that the SSE selected is one where the
" maximum vibratory acceleration at the sine throughout
.3 frequency range of interest is included".
10 CFR 100, Arr. A,,_.
I S V(a)(1)(iv).
The evidence is uniform _that peak accelera-tions are not within "the frequency range of interest".
II.
THERE IS NO REASON FOR THIS BOARD TO DEPART-FROM ITS PRIOR OPINION AS TO THE VALIDITY, FACTUAL OR'OTHERWISE, OF THE CHINNERY HYPOTHESIS As appears in the summary of evidence above, the witnesses for the Staff and the permittees are uniform in their view that Ih?..Chinnery's methodology is unacceptable.
We submit this view is correct.
Its validity is contingent upon the correcti-tude of four assumptions, each of which has a great deal of uncertainty associated with it.
Indeed, Dr. Chinnery himself admits that there is no way to perform a realistic assessment of the unc..
- minty involved.
In addition to this basic flaw, the record reveals other good and sufficient reasons for the Board to reject Dr. Chinnery's views.
Dr. Chinnery admits to having no idea whether the basic data from which he worked (the Smith catalog) is ar.y good.
Tr. 55 Nor does he have any idea how much, if any, of the Smith dita with respect to larger' events (MMI > V) involves damage on rock as opposed to soil.
Tr. 20.
When his thesis is tested by checking his predictions as to return times for small events (MMI II-IV), the predictions are in error by factors ranging from three to six when-compared with actual data for the years 1970-80.
Tr. 316-20, 327 In addition, there is frankly the question of the witness' general credibility.
In his direct testimony, he stated "of I t--
a
i what scientific literature there is, the vast bulk assumes that frequency-intensity relationships are linear."
Chinnery 13.
On cross-examination he admitted that these other writers were engaged in analyzing problems much different than that with which Dr. Chinnery wa: dealing, and the acceptance of'the assump-tion was not as critical to their analysis as it was to Dr. Chinnery's.
Tr. 68-69.
Dr. Chinnery testified that "a good scientist will never quote an observation without also quoting his best estimate of the error associated with his piece of data."
Chinnery Rebutt'al 2.
And yet We defy anyone to find Dr. Chinnery's compliance in his testimony with his own admoni-tion.-
A review of his cross-examination reveals an evasive and nonresponsive witness.
Perhaps the most revealing exchange in the record appears at Tr. 622:
"Q
_All right, sir.
Whether or not it was under oath was that a true statement so far as you're concerned?
A I cannot give a real definitive answer to that, Mr. Dignan.
Q Wait a minute, Dr. Chinnery.
The question is very simple:
Whether or not that state-ment was, under oath was it true?
Do you have tr'ouble with the word 'true', sir?
A In the context in which you ask it, yes, I de have trouble with that word." ~
r Finally, despite the' fact that he apparently convinced the Commission to.the contrary, the utilization of Dr. Chinnery's
. theory, by his own admission requires changes in 10 CFR 100, App. A.
Tr,. 15-18; 29-31; Chinnery 1-2.
Indeed, advocacy of such a change in Appendix A is his whole purpose in pressing his views.
Tr. 625.
CONCLUSION For the foregoing reasons the Appeal Board should fully
(
' adhere to its prior opinion in:this matter and in its decision doing so should include the findings and rulings set forth in Appendix A hereto.
Respect _fu bmitte c
W f c)
Thomas G.
Dignan, Jr.
R. K. Gad III Ropes & Gray 225 Franklin Street Boston, MA 02110 (617) 423-6100 Dated:
May 8, 1981
r APPENDIX A The permittees respectfully request that the following proposed findings and rulings be included in the Appeal Board's decision-in this matter.
I.
AS TO THE ISSUE OF THE FACTUAL VALIDITY OF DR. CHINNERY'S THEORY 1.
The Chinnery theory or hypothesis depends for its validity on the acceptance of four assumptions viz:
(1) that there exists a log-linear relationship between cumulative frequency and epicentral intensity of earthquakes, (2) that the slope of the line to be fitted to the empirical data as to return times in all areas of the world is 0.57; (3) that there is no upper bound to earthquake size in any area; and (4) that in any given area it is valid to extrapolate frequency-intensity curves to predict return times for earthquakes greater than those recorded historically in the area.
Chinnery 8-13; Chinnery Ex. 2 at 757; Chinnery Rebuttal 5-14; Reiter 3 2.
If any of these assumptions are not accepted as accu-rate, the Chinnery theory and any results therefrom must be disregarded.
T.
90.
3 There is a great deal of uncertainty associated with each of'these assumptions, as Dr. Chinnery himself admits.
Reiter 10; Chinnery Ex. 2 at 767; Tr. 100; ASLB Tr. 4055-56.
4.
There exists no basis for accepting the uniform slope
' assumption in light of studies of a similar nature which have produced widely varying slopes.
Reiter 6.
I=.
5.
The "no upper bound" assumption ignores geology and, indeed, the assumption of the existence of upper bounds is reasonable in geological terms and supported by 1000 years of data from Scotland and England.
Jackson 3-9; Reiter 6-8; Hoya 4; Tr. 401, 443-44.
6.
The log-linear extrapolation of cumulative frequency-intensity data without any-reference to the geology'of the particular site is highly questionable; it is a fact that in any given event the damage on soil far exceeds that on rock; indeed, the probability of damage of a given 4.ntensity occurring on rock as opposed to soil is lower by a factor of 4 to'15.
Holt 1, Figs. 1-2, Ayp_. 1; T_r,. 349-50.
7 There have been no recordings of strong motion assoc 1-ated with MMI VIII on rock in the United States.
Reiter 23 8.
The results to be gleaned from a Chinnery-type analysis for a given area may be radically affected by the choice of temporal period; a choice'which is wholly subjective.
Holt 2-4; Tr. 60-61, 96-98.
9.
The Chir.necy analysis does not check out when its predictions of small events are compared with historical data.
Tr..?l6-20.
~
10.
Dr. Chinnery.has never checked his data base; and has no idea as to its validity especially with respect to higher epicentral incensity events.
Tr. 55 See also Tr. 20. -
7_ _
11.
The Board finds the Chinnery methodology to be unacceptable for use in designing nuclear power plants.
Tr.,319-401, 573-75, 579 12.
The Board finds Dr. Chinnery not to be a credible witness.
13.
Dr. Chinnery admits that his methodology is not consistent with 10 CFR 100, Appendix A.
II.
AS TO THE ISSUE OF THE STAFF'S METHODOLOGY FOR CORRELATING VIBRATORY GROUHD MOTION 1.
The Staff testified without contradiction that the methodology it employs to correlate vibratory ground motion is consistent with Appendix A, is acceptable, and is conserva-tive, and the Board so finds.
See Summary of Evidence, 5B supra.
2.
The Board rules that 10 CFR 100, Appendix A, does not require that the seismic design criteria for a nuclear power plant account for.1solated peak accelerations which have been observed for earthquakes of various intensities because such isolated peak accelerations are not within the " frequency range of interest."
10 CFR 100, App. A, 6 V(a)(1)(iv).
III.
CONCLUSIONS 1.
The Appeal Board having reconsidered its prior opinion in light of the evidence adduced adheres to the findings, rulings i
and views expressed therein.
"5
, L
_.,,._-.,_.----3
- .+,
2..
'Even'if the Chinnery hypothesis were'to be accepted,
._there'is no basis for concluding.that the.Seabrook seismic
- design is other than conservative'or requires modification.
.t I
8.-
F'
APPENDIX B 4
Using the data in Table 4 of Chinnery, 1979 (Ex. 2 at 764), one gets the frequencies and cumulative frequencies shown in Table B-1.
Note that Dr. Chinnery has shortened the observation period for the lower magnitude events.
This is done to compensate for the hypothesized lower sensitivity of older records to lesser intensity events, and doing so substantially undercuts the argument that the record (and the frequency computations derived therefrom) of these lesser intensity events is unreliable.
Regressing the data in Table B-1 over various ranges of MMI intensity, on a log-linear basis, gives the results shown
.in Table B-2.
Note that the slopes range from -0.37 to -0.52 (and the latter is L regression of only three data points).
The difference in coefficients of correlation is insignificant.
There is, therefore, no mathematical or statistical warrant for selecting a slope of -0.57, except, of course, unless one has already concluded that the slope of log-linear regressions of cumulative frequency versus intensity curves is uniform from region to region and that uniform slope is -0.57.
How-ever, or.e -does not generally prove a point by assuming it.
One might ask why Dr. Chinnery would select a slope that is steeper than he might have, and which (because it predicts a lower frequency for earthquakes of MMI > VIII than would a shallower slope) is more favorable to the existing Seabrook,
r-5 o
design.
The answer lies in the fact that Dr. Chinnery must somehow account for the lack of any earthquakes of Inten-sity VIII or greater in the Boston-New Hampshire region in the last'350 years.
Dr. Chinnery essays a plausible theory for his
-0.57 slope.
Chinnery, Ex. 2 at 767-70 and Table 5.
- Note, however, that for the slopes actually implied by the data he records, the probability that an MMI IX or greater would have occurred in the last 350 years is between.5 and.75.
That no such event occurred tends to imply that there exists a physical upper bound on earthquake intensity in the Boston-New Hampshire zone in the geologic "present" for which Dr. Chinnery has failed to account.
More generally, this analysis confirms the truism that inference of physical phenomena from statistical analysis is a trap for the unwary unless it is coupled with and governed by an understanding of the underlying physical relationships at work._ Dr. Chinnery admits to a lack of any a priori basis for his extrapolatable log-linear cumulative frequency versus intensity with no upper bound hypothesis.
It is instructive that one could hypothesize a linear (not log-linear), non.umu-lative' frequency versus intensity relationship, which, given
.7744
.1258I)
Dr. Chinnery's data, would yield a curve (N
=
of equally good fit (R2=.9861) and which predicts an upper
' bound at approximately MMI = 7 Table-B-1 I
MMI No/ Events'(N)
Period (Yr)
Frequency (N/Yr)
Cum. Freq. (N,,/Yr)
Log Cum. Freq.
II 16.0 32 0.500 1 375 0.138 III 13.5 32 0.422 0.8749
-0.058 IV 17 5 60 0.292 0.4529
-0.344 V
12.0 100 0.120 0.1609
-0 7934 VI 35 160 0.0219 0.0409
-1.3885 VII 3.0 160 0.019 0.019
-1 721.
i I
Table B-2 2
~
Points R
Equation Predicted IX Freq.
Prob. IX Occ. (350 yri 2-7 9744-1.0716
.3925 3 5 x 10-3 706.
2-6 9557 1.0262
.3788 4.1 x 10-3 762 3-6 9763 1 3524
.4441 2 3 x 10-3 553 4-6 9936 1 7693
.5223 1.2 x 10-3 343 4-7 9896 1.5376
.4726 1 9 x 10-3
.486 k
- 4
l
~
CERTIFICATE OF SERVICE I, Thomas G. Dignan, Jr., one of the attorneys for the applicants herein, hereby certify that on May 8, 1981, I made service of the within document by mailing copies thereof, postage prepaid, first class or airmail, to:
Alan S. Rosenthal, Chairman Ellyn R. Weiss, Esquire Atomic Safety and Licensing Earmon & Weiss Appeal Board Suite 506 U.S. Nuclear Regulatory Commission 1725 I Street, N.W.
Washington, D.C.
20555 Washington, D.C.
20006 Dr. John H. Buck Robert A. Backus, Esquire Atomic Safety and Licensing O'Neill Backus Spielman Appeal Board 116 Lowell Street U.S. Nuclear Regulatory Commission Manchester, New Hampshire 03105 Washington, D.C.
20555 Stuart K Becker, Esquire Dr. W. Reed Johnson Maxine I. Lipeles, Esquire Atomic Safety and Licensing Assistant Attorneys General Appeal Board Environmental Protection Division U.S. Nuclear Regulatory Commission Department of the Attorney General Washington, D.C.
20555 One Ashburton Place, 19th Floor Boston, Massachusetts 02103 Ms. Elizabeth H. Weinhold 3 Godfrey Avenue Roy P. Lessy, Jr., Esquire Hampton, New Hampshire 03842 Office of the Executive Legal Director Atomic Safety and Licensing U.S. Nuclear Regulatory Commission Soard Panel Washington, D.C.
20555 U.S. Nuclear Regulatory Commission Washington, D.C.
20555 Office of the Attorney General 208 State House Annex Concord, New Hampshire 03301
/2 Thomas G. Digna # Tr.