ML20141H696
Text
pr:
DRAFI MEMORANDUM FOR: Leon Reiter', Acting Chief Geosciences Branch Division of Engineering FROM:
Phyllis Sobel, Geophysicist Seismology Section, Geosciences Branch Division of Engineering
SUBJECT:
SUMMARY
0F NRC-EPRI MEETING 0F JULY 9,1985 On July 9,1985, representatives of NRC, EPRI, LLNL and the USGS met in Bethesda to compare seismic hazard methodologies. The agenda, a meeting summary, the list of attendees and viewgraphs are enclosed. At this meeting EPRI presented an explanation of their seismic hazard methodology, a summary of their preliminary results and a description of their sensitivity studies.
LLNL presented some comparisons of the EPRI and LLNL hazard results and discussed their planned comparison studies. The USGS reviewed generic problems in ground motion estimation.
In the meeting NRC and EPRI presented their sensitivity and comparison programs. One important difference in the methodologies is the minimum magnitude considered (3.75 for LLNL versus 5.0 for EPRI); the effect of the choice of minimum magnitude on hazard estimates is being investigated by LLNL.
Perhaps the largest difference in the two methodologies is in the ground motion models.
EPRI's ground motion attenuation curves are lower than the LLNL curves. Some of this difference may be due to EPRI's consideration of Canadian data. The differences between the LLNL and EPRI ground motion models may be the most significant issue in the comparison program and this topic should be vigorously addressed in the next few months.
LLNL is comparing the EPRI results to the LLNL program using the same minimum magnitude (5.0) and the same ground motion models (EPRI's).
For five of the sites the results overlapped; the 50th percentile curves for peak ground acceleration were similar, but uncertainties were greater for the LLNL resul ts. For the other four sites, the EPRI results are significantly lower.
AtthosesitesthehazardYdiminatedbytheeffectsofthebackgroundzone.
8601170381 351205 ppR FOIA BELLSS-535 PDR
r Leon Reiter One of the LLNL comparison studies will address the'effect of the background zone. The LLNL comparison studies should focus on those sites where the differences between the EPRI and LLNL hazard results are the greatest.
It is encouraging that in spite of the different inputs and methodologies, the two studies are producing similar results at some sites. An additional comparison study LLNL should attempt is a test using the same input paraneters into the two hazard programs to assess any differences in the hazard calculations.
The next NRC-EPRI meeting on comparison studies is scheduled for November.19 in Bethesda. Both groups plan to complete their sensitivity and comparison studies by that data. NRC needs to begin a program to assess methods of comparing'the hazard results at the EUS sites and procedures for identifying outliers.
I would also recommend that EPRI examine this problem.
Phyllis Sobel, Geophysicist Seismology Section Geosciences Branch Division of Engineering
Enclosures:
- 1. Meeting Agenda
- 2. Meeting Summary
- 3. List of Attendees
- 4. Viewgraphs cc:
J. Knight C. Ong R. Bosnak J. Chen G. Lear A. K. Ibrahim GSB N. Chokshi L. Beratan D. Perkins, USGS E. Zurflueh K. Campbell, USGS M. Blackford B. Bender, USGS T. Schmitt T. Algermissen, USGS A. Murphy C. Stepp, EPRI L. Abramson J. King, EPRI C. P. Tan R. McGuire, Risk Engineering, Inc.
J. Savy, LLNL D. Bernreuter, LLNL D. Mensing, LLNL 9
r Int (cktt-f, I
DRAFT AGENDA NRC-EPRI Meeting on Comparisons of Seismic Hazard Methodologies Bethesda, Maryland July 9, 1985 8:30 am Introductory remarks (Leon Reiter &
J.C. Stepp) 8:40 Tectonic issues addressed in the development of the EPRI methodology (W. Arabasz) 9:00 Methodology for assessing seismo-genic potential of tectonic features and interpretation of seismic source zones and maximum magnitudes (K. Coppersmith) 9:45 Seismicity-parameter methodology and application (R. McGuire) 10:15 Break 10:30 Ground motion models (R. McGuire) 10:45 Hazard calculation--simplifying assumptions and explanation of results (R. McGuire) 11:15 LLNL Presentation--Comparative studies, to date, and planned studies 12:15 pm Lunch 1:15 LLNL--continued 1:45 USGS Presentation--Completed and planned studies 2:45 Break 3:00 EPRI plans for parametric and sensitivity studies (J. King) 3:15 Discussion of planned studies 5:00 Adjourn i
r DRAFT NRC-EPRI MEETING ON COMPARIS0NS OF SEISMIC HAZARD METH000LOGIES MEETING
SUMMARY
EPRI Seismic Hazard Metnodology EPRI presented a sumary of their seismic hazard rcethodology.
In his introductory remarks Carl Stepp of EPRI cited some of the improvements to the methodology still in progress - estimation of maximum magnitude, reevaluation of instrumental magnitudes, improvements to the earthquake catalog and ground motion studies. The EPRI methodology is discussed in the EPRI report,
" Seismic Hazard Methodology for Nuclear Facilities in the Eastern United States" (EPR4/50G-Draf t 85-1).
Walter Arabasz of the University of Utah sumarized the tectonic issues addressed in the development of the EPRI' methodology.
In 1984 and 1985 six expert teams and other scientific experts met in workshops and seminars to discuss tectonic models for the Eastern United States (EUS). The teams largely concluded that the principal sources of stress in the EUS were plate boundary forces (ridge push generated at the Mid-Atlant.ic Ridge) and stresses associated with lithospheric flexure due to sediment loading in the Coastal Plain. The most important observations related to the source of stress were the direction and magnitude of stress (from drill hole data., geologic observations, and earthquake fault plane solutions) and the locations of seismicity.
Kevin Coppersmith of Geomatrix discussed the EPRI methodology for assessing the seismogenic potential of tectonic features.
Each team mapped tectonic features and listed their characteristics.
For each tectonic feature an assessment was made of its potential for activity. _(See the example in the attached viewgraphs). Each team chose their own criteria for assessing
. potential activity. Then the geometry of seismic source zones were defined
r ORAFT based on the tectonic framework interpretations.
Each team defined the maximum earthquake magnitudes for each source zone based on historical seismicity and physical constraints (for example, the dimensions of tectonic features). The documentation of the basis for each teams' source zones is an important aspect of the EPRI program.
Robin McGuire of Risk Engineering discussed the seismicity parameters estimated by each team for their source zones.
The methodology team developed an earthquake catalog which most of the teams used.
Relationships were established among the magnitude scales and a consistent size measure (body-wave magnitude) was established for the catalog. Aftershocks (about 20 percent of the events) were removed from the earthquake catalog.
Some of the teams used their own methodologies to select seismicity parameters (a and b values), but most of the teams used a new methodology developed by Veneziano and Van Dyck. This new methodology for calculating a and b values allows (1) frequency of occurrence versus magnitude data to be weighted as a function of magnitude, (2) variation of a and b values within source zones, (3) prior weighting of b values, and (4) estimation of maximum magnitude.
This methodology uses historical earthquake data and accounts for the lack of reporting due to sparse population distributions.
McGuire then discussed the EPRI ground motion models.
Four ground motion models are used in the hazard calculations (Nuttli's model, two intensity-based models and a stochastic model); all the models are weighted equally. These ground motion models were compared to existing EUS strong motion data and EPRI found them to be typical for rock and stiff soil recording sites.
The EPRI ground motion curves have more uncertainty at larger epicentral distances, perhaps due to the consideration of Canadian data in the stochastic model.
Sone sensitivity studies have been performed using hazard estimates for spectral velocity at a frequency of 10 hertz.
The uncertainties in the ground
, motion models and differences in the team's inputs have sensitivities of about an order of magnitude each. These are judged to be the major sources of uncertainty. The hazard calculations show little sensitivity to smoothing of seismicity parameters, choice of maximum magnitude, or source combinations for any one team. The overall uncertainty in the hazard calculations is estimated at about one order of magnitude. These results are based on only a few runs for each parameter; more examples will be generated as part of the EPRI sensitivity program.
Jerry King of EPRI noted the following differences between the EPRI and LLNL methodologies and results.
EPRI's ground motion attenuation curves are lower than LLNL's. LLNL integrates down to a magnitude of 3.75; EPRI integrates down to a magnitude of 5.0.
EPRI has not developed local site condition corrections but LLNL has.
EPRI's preliminary hazard results overlap the LLNL results. For some of the sites the 50th percentile peak ground acceleration curves are close, but the LLNL 15th and 85th percentile curves show greater uncertainty.
For the other cases the EPRI results are lower than the LLhl results.
LLNL Comparison Studies Don Bernreuter sumarized the differences between the LLNL and EPRI methodologies and the planned comparieon studies.
(See viewgraphs). Jean Savy compared the EPRI results to LLNL estimates using the different input parameters and methodology, but starting the integration at the same magnitude (5.0), and using the same (Nutt11 1984) ground motion mndel, and not using site corrections, for Millstone, 3
Limerick, Braidwood and Wolf Creek, the LLNL 50th percentile peak ground acceleration curves are similar to EPRI.
for Maine Yankee EPRI's results are slightly lower.
For Vogtle, Watts Bar, Shearon Harris and River Bend, the EPRI results are significantly up4tWe--lower than LLNL.
In all cases the LLNL range of uncertainty (differences between the 15 and 85th percentiles) were greater than EPRI. The maxin.um dif ferences are at sites where the hazard is due to either sources at least a hundred Km f rom the site or the background zone where the site is located. Differences in the ground motion estimates at larter distances or upper magnitude cutof fs in the background zones could be key factors.
%MT LLNL also showed an example where the lower bound magnitude was set an magnitude 3.75 or 5.0; the effect for peak ground acceleration was mostly at the higher probabilities of exceedance (lower g-values).
LLNL will consider examining this effect on snectra also.
USGS Presentation Dave Perkins outlined his concerns with the EPRI and LLNL programs.
For example, it is possible that all tectonic models have not been represented by the LLNL experts and EPRI teams. Also the LLNL and EPRI ground motion models should agree with observed EUS intensity data.
Perkins presented a sensitivity study showing the effect of introducing uncertainty in the location of earthquakes assigned to a source zone.
By varying the earthquake locations by 0, 10, 20, 30 and 40 Km, the ground motion contours are smoothed at the source zone boundary. At high uncertainties (40km) the ground motion highs do not center on the source zones, due to the effects of other source zones.
For a given site, this effect could change the ground ration prediction by 10 to 25 percent, which is less than some of the differences between the EPRI and LLNL results, and he concluded, is thus not very significant.
Perkins looked at the ef fect of different zonation nodels in the Eastern Seaboard. The ratio of highest predicted ground motion to lowest predicted ground motion varied from not significant to significant (ratio greater than 1.4), which occurred largely in areas of historical seismicity.
The highest ratio (about 2) was near Charleston.
The offact of changes in the ground motion attenuation function was observed along an E-W profile; the variations in attenuation function effect the absolute level of hazard more than the relative hazard between sites.
. Ken Campbell reviewed generic problems in estinuting ground motion models in the EUS.
For example, ground riotion estimates may vary depending on the type of stress regime (compressional versus extensional). What are the effects of differences in EUS versus WUS attenuation? Does source spectral scaling in the EUS differ from the WUS? The limited EUS strong motion data base and site conditions at the recording sites may..can the data base is not appropriate.
EPRI Sensitivity Studies Jerry King discussed EPRI's plans f or parametric and sensitivity studies.
EPRI will assess the uncertainties in the data base used to define tectonic features.
Tectonic interpretations near the nine EPRI test sites will be refined.
Source-zone interdependencies will be examined.
The EPRI earthquake catalog will be revised and the offects of the catalog on the hazard estimates will be tested, including the effect of magnitude uncertainties for the larger events. One study will examine whether old epicentral intensities in New England were higher or lower than recent seismicity. A workshop on the earthquake catalog will be held October 2, 3 and 4 in Denver.
EPRI will examine the assumptions involved in the probability of detecting seismicity including use of the Poisson model to describe earthquake occurrence.
The form of the upper bound magnitude distribution and trethods f or estituting maximum magnitude will be examined.
(EPRI uses 3 points to define upper bound magnitude.)
The EPRI hatard results will be compared to historical carthquake hatard calculations.
(LLNL did this for their ten test sites.) EPRI has no plans for EPRl/LLNL comparisons, but is cooperating with the LLNL comparison studies.
EPRI is planning more ground motion workshops and several workshops with their peer review panel.
Discussion of Planned Studies There was some discussion of the EPRI team approach versus LLNL's individual experts, but there are no plans to assess the ef fect of these dif ferent approaches en the hazard estimates.
It would be useful to have a test in which for the same inputs, the differences between the results are assessed.
For example, some dif ferences between the inethodologies such as the form of upper magr.itude truncation may lead to dif ferences in the hazard results.
EPRI's ground ration model study will not be resolved this year and no site corrections are planned in the near future; however, NRC suggested more ef fort was needed sooner.
Dave Perkins suggested looking at hazard along a traverse where large changes are expected.
Leon Reiter stated that we will eventually identify sites that are outliers and assess significant dif f erences between all the EUS sites. We should concentrate on those sites where there are the greatest differences between the EPRI and LLNL results.
The recting gave EPRI and NRC a chance to reassess their planned cortparison and sensitivity studies.
The next NRC-EPRI recting on comparison studies is tentatively scheduled for November 19 in Bethesda. 110th groups plan to complete their sensitivity and corparison studies by that date.
DRAFI.. Enclosure 3 List of Attendees July 9, 1985 NRC-EPRI Meeting on Corparisons of Seismic Hazard Methodologies Nace Organization Phyllis Sobel NRC-GSR Ernst Zurflueh NRC-ESB Michael Blackford NRC-WMGT Tom Schmitt NRC-RES Stephan Brocoum NRC-GSB John T. Chen NRC-SCEB Harold E. Lefevre NRC-GSB Sob Rothnan NRC-GSil A. K. Ibrahim NRC-hMSS Dave Perkins USGS-Golden Kenneth Campbell USGS Golden Cole itcClure EPRI Adv. Panel Carl Stepp EPRI Gus Gieseroch NRC-GSB Lee Abramson NRC-RES Robin McGuire Risk Engineering, Inc.
Lean Reiter NRC-GSB Jerry King EPRI Kevin Coppersmith Geomatrix Andrew Murphy RES/NRC Jeffrey Kimbell Roy F. Weston, Inc.
Richard Holt Weston Geophysical Corp.
DRAFT List of Attendees (continued)
Name Organization Jean Savy LLNL DickMe) sing LLNL Don L. Bernre*jer LLNL C. P. Tan NRC/DE Tom O'Hara Yankee Atomic John P. Jacobson Yankee Atomic John J. Owyer Law Engineering Noel Barstow Rondout Assoc.
J b McWhorter Dawes and Moore Cookic Ong NRC/0PE Tom Cardone NRC/GSB Joe Litchiser Bechtel Walker Arabasz University of Utah loni IYh d* ' d 24
6n c losurt
'I NRC/EPRI C0 ORDINATION MEETING JULY 9, 1985 e
PURPOSE
-- PRESENT EPRI SEISMIC HAZARD METHODOLOGY
-- DISCUSS SPECIFIC SENSITIVITY STUDIES, SELF-CONSISTENCY TESTS AND COMPUTATIONS TO BENCHMARK THE METHODOLOGY
-- LEARN OF NRC PROGRAM 1
e RESULTS
-- FEEDBACK ON PLANNED SENSITIVITY STUDIES
-- IDENTIFY ADDITIONAL SENSITIVITY ANALYSES
e
_sh
.a
,b LA:
.-_L_.
j
....~
)
i i
EPRI i
i SEISMIC HAZARDS RESEARCH PROGRAM 1
f i
1 l
e PROGRAM OBJECTIVES i
r e
PHASE 1 PROGRAM ELEMENTS AND APPROACH 1
1 o
PHASE 2 PROGRAM ELEENTS I
1 i
i i
i j
i I
t 1
t 4
4 i
i i
i e
i
3 EPRI i
l SEISMIC HAZARDS RESEARCH PROGRAM OBJECTIVES l
e DEVELOP STATE-0F-THE-ART PROBABILISTIC SEISMIC HAZARD 4
i METHODOLOGY e
DEVELOP EARTHQUAKE SOURCE ZONE INTERPRETATIONS THAT ACCOUNT FOR COMPETING HYPOTHESES OF EARTHOUAKE CAUSES e
BUILD SCIENTIFIC CONSENSUS e
BENCHMARK METHODOLOGY FOR GENERAL APPLICATION i
e IDENTIFY SCIENTIFIC IMPROVEMENTS I
I l
L l
1
?
l-EPRI SEISMIC HAZARDS RESEARCH PROGRAM PHASE 1 PROGRAM ELEMENTS i
1 e
LATA COMPILATION AND ANALYSIS
)
e METHODOLOGY DEVELOPMENT e
TECTONIC INTERPRETATIONS 1
e SEISMIC WAVE ATTENUATION i
~
l I
l-l I
1 T
i
EPRI SEISMIC HAZARDS RESEARCH PROGRAM PHASE 2 PROGRAM ELEMENTS e
ASSESSMENT OF PHASE 1 RESULTS
-- PARAMETRIC ANALYSES AND BENCHMARKING 0F. METHODOLOGY
-- SCIENTIFIC PEER REVIEW
-- QUALITY ASSURANCE OF THE COMPUTATIONAL SOFTWARE e
SCIENTIFIC IMPROVEMENTS
-- SEISMIC WAVE ATTENUATION
-- SEISMICITY OCCURRENCE MODELS
-- MAXIMUM MAGNITUDE ASSESSMENT
-- IMPROVEMENTS TO THE EARTHQUAKE CATALOG s
l
- -., - = - -.,
., - -, = -,,,,,
ELECTRIC POWER RESEARCH INSTITUTE I
Program Directors:
J. Carl Stepp l
Jerry L. King h
e s
y SENIOR REVIEW PANEL METHODOLOGY TEAM EARTH SCIENCE TEAMS DATA BASE MANAGER Shelton S. Alexander, Technical Director: Robin K. McGuire Bechtel Group Inc.
Project Manager: Dennis Smith Chairman Dames & Moore Clarence R. Allen Earth Science and Seismictty Parameters Law Engineering Inc.
C. Allin Cornell Probability facilitation and Hazard Analysis Rondout Assoc., Inc.
Cole R. McClure Weston Geop. Corp.
l Otto W. NuttIf Group Leader: Ashok Group Leader: Robin K.
Woodward-Clyde Cons.
M. Naff Toksor Patwardhan McGuire Wooerard-Clyde Consultants Dames & Moore g
g l
K. Coppersmith A. Eckert ro l
R. Kulkarni G. Toro J
l'
[--*
R. Youngs Yankee Atomic Electric' Co, l
R. Keeney T. O'Hara R. Winkler D. Veneziano W. Arabasz Contractual Relationship Information Flow Figure 2-1.
Project Organization and Participants 4
Table 2-1 EARTH SCIENCE TEAMS Team Members Bechtel Group, Inc.
Dr. Thomas Buschbach Dr. Robert D. Hatcher, Jr.
Dr. Joseph Litehiser*
Dr. Rolfe Stanley Dr. Isidore Zietz Dames & Moore Prof. Charles Fairhurst Prof. Robert Herrmann Prof. Lyle McGinnis Mr. James McWhorter*
Dr. Rene Rodriguez Law Engineering Testing Company Prof. Robert Butler Dr. Martin Chapman Dr.' John Dwyer Prof. Arch Johnston Prof. Timothy Long Mr. Malcolm Schaeffer Mr. William Seay
-Dr. Robert White
- Rondout Associates, Inc.
Ms. Noel Barstow*
Prof. William Hinze Prof. Pradeep Talwani Prof. Barry Voight Weston Geophysical Corporation Mr. Richard Holt Dr. George Klinkiewicz*
Dr. Gabriel LeBlanc Prof. Donald Wise Woodward-Clyde Consultants Dr. Terry Engelder Dr. John Kelleher-Dr. Richard Quittmeyer Mr. Thomas Statton*
Dr. Thomas Turcotte
- Team Leader 2-6
[
TECTONIC FRAMEWORK / SEISMIC SOURCE DEFINITION l
GOAL To make explicit use of the current state-of-knowledge of tectonics in the eastern United States to define and characterize seismic sources for seismic hazard analysis.
OBJECTIVES e identify tectonic features that may generate moderate-to-large earthquakes (mb 25.0) e Evaluate criteria for assessing the seismogenic potential (activity) of tectonic features e Assess the likelihood of activity of each tectonic feature i
e Define seismic sources for seismic hazard analysis that incorporate knowledge of tectonic features and seismicity data
L I
me s v. ~ < ny
/
f.:
' ~
TECTONIC STRESS REGIME n
e Interpretations of Operative Processes for Gererating Crustal Stresses in the Eastem United States o'
e inwrpretations of Stres Orientation and Maritu&
k 1 r 1 r g
fe
->...us o 4x s
~ '
4 o.
9..
TECTONIC FRAMEWORK
[
i: '
e Identification of PowntiaHy L _.+ c Tectonec h5 Features
~
e Definitionof Criteria for Assessang Activity
?
e Assesment of Probability of Actkity for Each
[
Tactonic Feature lT2 Figure 3-1.
Key Components and Products of Tectonics Stress Regime'and Tectonic framework Assessments uS 3-3 l
'4-
- - - - +,, _ -,,., _ _ _,, _
EPRI WORKSHOPS WORKSHOP TOPIC
]
Workshop # 1 Data Needs Workshop # 2 Tectonic Processes of intraplate Stress Generation Jnkraeh've Afetf&
Workshop #3 Tectonic Stress Regime in the Eastern United States / Potential Stress Concentrators
[n k ro c h N # te h e)
Workshop #4 Approaches to developing Tectonic Frameworks and Jahrnehk #h*k)
Workshop #5 Presentation of Tectonic Frameworks and Seismic Sources Workshop #6 Approaches to Assessing Seismicity Parameters Workshop #7 Presentation of Seismicity Parameters
H l
Scientific uncertainty 8 Uncertainty in the scientific understanding of physical, observable characteristics of features that are diagnostic for assessing activity Informational uncertainty
- Uncertainty because of incomplete information on the physical characteristics of particular tectonic features l
1
s Tectonic feature e A tectonic feature is a large-scale gEdlogic structure l
.or element of the earth's crust 4nanifested a' a geophysical anomaly.
o w" d '/
x e A tectonic feature has a distinct geographical I
location, but may also be a member of a group of features that have similar characteristics, I
genetic history, etc.
e Examples of tectonic features are Mid-Continent l
Geophysical Anomaly and the Brevard fault zone.
t i
I l
t, Activity S Activity is defined as the capability to generate moderate-to-large(mb.2 5? earthquakes within the present stress.aregime.
(
9 Activity is stated as a probability that a particular l
feature is cap'able of generating moderate-to-large l
j e The probability of activity is independent of time l
as long as the stress regime is assumed to stay the same.
Active tectonic feature e An active tectonic feature is a feature that can i
generate moderate-to-large magnitude i
earthquakes within the present tectonic stress I
regime.
l l
8 The particular criteria for assessing whether or i
not a tectonic feature is active is left to the Earth ScienceTeams.
1
~
l l
7 Tectonic framework
- A tectonic framework is a geographical depiction and interpretation of the tectonic features that are potential sources of moderate-to-large magnitude earthquakes under the comtemporary stress regime.
e The-tectonic framework includes an evaluation of the probability of activity of each tectonic feature and an evaluation of the criteria for estimating activity.
cf CRITERIA FOR ASSESSING ACTIVITY e Spatial association with seismicity e Geometry and most recent sense of slip relative l
to present stress regime e Crustal scale expression (expressed in the deep crust) e Proximity to structural barriers or intersections e Persistence of deformation through a significant-period of geologic time e Recent regional strain, including that recognized j
from geodetic measurements or from geomorphic i
indicators e Feature-specific brittle slip in mode consistent with the present stress regime j
j e Feature lies within a region of higl. seismic flux j
e Evidence of local stress amplification e Paleoseismicity (prehistoric earthquake evidence)
e a--u--+a
-nu a
.a-a n-
-u-,
/o ij r
j i
I 8
e F
l u
.i
?
m.
1 k*
k 1
e I
~
s l}e N
t m
w r
l ll He n
r
- / a.C 9
i
//
4 i
I 3-39 h
/
~ ~. -
A,
- hy
T
'vg Mwg
- w,
~ll Probability that the SLven Featurw Eshtb1Le e Steen e
I Level of Bach Charesterkette t
Featurw Posture Posture Physical Characteristic 8
1.
Seattel Association with Beimatetty:
Focal Mechanisme Consistant erith Feature Geometry 1.
Moderate to-Large Earthgmehoe 2.
'Sas11 Earthquakes enly 3.
so setsmicity 3el.e 1.0 1.0 2.
Geometry of feature Whign to stress ori " =a w
^'er ^ --
of 811e l
1.
Feverene ry/Somoe of sito 2.
aufswordle seamstryIsamos of
- sti, 1:1.0 1.0 1.0 l
3.
Brittle alls en a fantare 1.
Fleistetene Bolotene SlbF 2.
eroteseeue-Teethery Slip 3.
Pre-Cretaceene Skip er Bo l
Brittle Slip x=1.0 1.0 1.8 Figure 3-7.
Form for Assessment of Feature Characteristics 3-42 l
PRODUCT DOCuneENTATK)N I
\\
\\
e me for m e.,-
e Defaruten of each esctonac foeture Map of Tactonec Features e Defwntion of each criterion x
.._f.__
(ese Fy 34 for esample)
Meina of Physcal Cherectorstics ca:Tr aiA enosasittry.
. Deenensrersem een C
ines Fspare 3-7 ter enamotel
=
=
Feuevre Assessmesw forme Figure 3-3.
Products and Documentation for Tectonic Framework Assessment e
3-19 l
l
\\
\\
/
i
\\*\\
.f
'rt
\\
/
s
\\, t\\
Epeanne r
p
[
\\
\\
\\
\\
l 9
.\\
-e-
\\
l
,-. c /. \\i
\\
te;~:,
)
g,u \\
R :/
G ~.,',,'
\\
l
,7
+
\\.
\\
X
\\\\
N.\\
\\
/
\\
\\.
p
-.A
\\
,/
\\
\\
A' A
/y,,,,,,,a
./-
h T/ w
' < w.r +. ;,. c c..:. -..,....... -
~' ' $ '.' h P .' u
.. -s~. ?.i
. '= L3. ~.Gii;1 c.?M.
~
- e..: U sh.'k.Y'N' ~'~,:,,c..'.E.:t.x?::
n :... -... '.r&.
- .'.: *~ : * ::;.
=G;*:;.
.w s n
?.u:n, v:..-:-
-: e:. c.:.
n.
Figure 3-8.
Hypothetical Map of Significant Tectonic Features and Earthquake Epicenters e
t 3 45 1
/-/
sV
'N.
had iem nskan thus.
En kummmmmy e
Ns
- g A
~
g, is
.a
.s a7 as me e,,
u
.4
=
u u
ai pg u
i k'%
i
- s. %
\\.~
=
l
.i one-s=
es e.:
Figure 3-9.
Retrix f Physical Cherecteristics f r Example Ev.1mett.n 3-48
lR Probability that the Given Feature lhahibits a Siven Level of Bach Chorectaclette I
Featurw Feature Feature Physical Characteristic
- A 95 SC 1.
seattal ammeetatian erith amt " eLiv-Fecal Mechant
-^
-L_=* -^ erkik Feature enametry 1.
Wederste-to-tarte Bart W 0.3 C.I 9.2 2
2.
hall L l'; ' - enly 0.2 0.5 0.1 3.
We senseletty 0.5 0.4 0.7 Ia1.0 1.0 1.0 2.
enametry of Feature Balative h Strees Orientation and/or Sense of Riin 1.
Favorable Geometry / Sense of 0.3 0.3 0.1 81Lp 2.
Unfevorable Geometry / Sense of 0.T 0.T 0.9 Elip Is1.0 1.0 1.0 3.
Brittle Elle en a Feeture 1.
Fleistocene-Helocene Slip 0
0.1 2.
Cretaceous-TertLary slip 9.1 0.9 3.
Pre-Cretaceous slip or so 0.9 0
Brittle slip I.1.0 1.s Figure 3-10. Example Feature Characteristics Assessment en 3 50 i
4 3
/G SEISMIC SOURCE A seismic source is a region of the Earth's crust that has a single probability of being active and a single distribution of maximum magnitude
F7 SEISMIC SOURCE DEFINITION Definition of the geometry of each seismic 1.
source,ie. locations of potential future i
earthquakes, using the tectonic framewprk interpretations.
2.
Specification of the probability of activity of each source and interdependencies of activity l
l among several sources.
i Evaluation of the adequacy with which the 3.
l primary seismic sources identified from the tectonic framework fully represent the t
potential for future earthquake occurrence in a region. Additional default and/or background sources may be added to supplement the primary sources.
i
.-..,-.__,___..--...--__--..-_-____,.-_._..,,.-,,.-.,_.._,..,__.,.,,-___,.,__m_.,
_,...,.. - _,. _,.. _.._.,.. _ _ _, - - ~,
.~
_- _ _ _ =
I 1
8)
Lene5ewc,5 l
Anoe Sm ce i
b)
I
/
Ceanecien et L,ie 5eueces g)
/
/
f' Ceanecte et La Saueces
'" " * **"'" Z***
~
d)
% /
)
~
N
\\
/
^
Figure 4 1.
Es M les of Seismic Source Conftgurations 2
i 43
JO X = f(A,8)
Parameter A parameter S Dessees Deerensemen of X mi si.b l t
i bi P(h le,)
pg.,). pa.,).ptb ssil i
i ag plail b2 N
p4 tei) 2 I
i bi
/
pdh le2) i b2
- 2 p(b 1a2) p(e2) 2 a ster.bs)
\\
b3 p(n5 p(e l.p(b Ia )
p(b 1 s i 1
r 3 z 3 a Figure 4-3.
Sample Logic Tree I
e 4-8
N Line source
\\
doeth 8-30 km
~
\\
k k\\
./
\\.\\
\\.
f
., k
\\
\\
..-\\. 4' g
/
\\
\\4,,
a i
u A
. gi ",h';[h s4 1
y'ti;*'/_
ji3 L
Q
?R,,'} YD r
chtj N.!k i[s g9,( n e,- *
/
ke lr~,: ~.'~ ~,
r..
.c,-
, - ~,.
\\
@,5
. i'-
xj m
9,s
- s,'l, A'
\\f
\\ ~ ~.,,
M
% y, w
- k
~
?
1-
-4 i
/
A w\\
- r.,
'g3 wa a
..a
_./
hS 'h Y. _ _ [A E.
a Aree source depth 8-20 km i
a ut h.,am e i
sowce i
s.,c, I
Poteni.W I
Coa 8.p.e se.en i C7wecteessation
/
x p..,
m, pWM)
L '"* $ "'"
Feen,ea x
2, p40 721 3
Figure 4-20.
Representation of Feature 8 as a Seismic Source 4-41
W I
I I
I I
I I
I I
I I
I I
I I
I I
I I
g.
.g 1
s l
k l-ij l
-3 l-I'il I
i, 1
P I
j.! s i f
I' J-l 1
I,j s ei l-
- l'It ilii l
e i.
t c, \\
E l'
- l j
Q j
l l-
,r l
.x 9
.t '.
i.
R
...s l
g I-
.I p
e..
l.
j
.g i
I 8
l'
- l
.e E
l-
-l 4
l E
I-1 3.
g s
s g
l 1
l I
l l-
-l I
I I
I I
I I
I I
I I
I I
I I
I I
I I
i 4-54
MAXIMUM EARTHQUAKE MAGNITUDE Historical Seismicity Constraints e Past large earthquakes e Historical plus increment 2
e Recurrence (1,000 yr per 100,000 km i
Physical Constraints o Spectral scaling relations I
j e General consideration of " dimensions" i
e Geologic characteristics and' analogy to locations of historical events o Crustal strain volume e
I i
,,.,.__,,___..,_-._.-,_....,,,.__n,-
27 EXAMPLE MAXIMUM MAGNITUDE ASSESSMENT Maximum Characteristics Magnitude (mb)
7.4 Rifts
well developed, relatively young, open-ended (developed in non-cratonic crust such as New Madrid, St. Lawrence)
6.8 Rifts
poorly defined, relatively old, surrounded by cratonic crust 6.8 Major crustal failure: entire brittle crust, limited source length (e.g., Charleston)
Shallow crystalline crust (New Brunswick) j 5.8 l
l l
5.5-4.9 Background (away from significant features) i i
i I
t
I i
i CHARACTERIZATION OF STRESS REGIME IN EUS?
o Multiple stress data (hydrofrac, overcoring, well break-out data, earthquake fault-plane solutions, geologic indicators) o Zoback and Zoback compilations (1980,1983) supplemented by data from EPRI efforts and revisions resulting from first-hand interactions with Zobacks Remarkably uniform orientation (NE-SW) of o
. maximum principal stress throughout EUS consistent with boundary forces acting on the North American plate and due to ridge push o
Distinct Atlantic Coast stress province doubtful o
Possibility of locally induced stresses in Atlantic coastal region due to ocean-continent boundry (but subordinate in magnitude to ridge-push stresses); regional stress field along Gulf Coast may similarly be effected locally by sediment loading and plate flexure ftG IS
' TECTONIC ISSUES ADDRESSED IN THE DEVELOPMENT OF THE EPRI METHODOLOGY" Motivation:
What scientific information can be used to identify zones of earthquake potential in the EUS--in addition to historical seismicity--as part of a seismic hazard methodology?
EPRI ACHIEYEMENTS:
o Explicit incorporation into a hazard methodology of up-to-date, scientific understanding of tectonics, stress, and earthquake generation in the EUS.
o Unprecedented attention to fundamental tectonic issues relating to earthquake generation in the EUS--and engaging of broad scientific expertise within the earth science community to address those issues in a structured way.
4~t G. 4
GENERAL TECTONIC ISSUES What tectonic processes are the principal causes of o
stress in intraplate regions?
o What are the diagnostic manifes.tations of those processes?
o What tectonic processes are operative in the EUS?
l How can the contemporary tectonic stress regime o
(i.e., the 3-D orientation and magnitude of crustal stresses) be characterized in the EUS?
o How do geologic structures interact with a tectonic stress regime? (Given some stress regime, what factors lead to potential for brittle failure and earthquakes as a result of (1) stress concentration, (2) strain localization, and/or (3) weakening of
)
strength?)
o What are the implications of stresses, structures, i
and other factors for assessing the seismic potential of individual tectonic features in the EUS--and what are the uncertainties in all of these?
~
i
-u..._
_a_..
.r___
h I.
lll\\b!!hihh.$. h. kkNhh.i:5?Cllk.........'?:::\\lNI\\\\$%!!?:b;..::P
...: killlllilllu?ih,.
\\!!!!
sr:'
TECTONIC STRESS REGlME fl[j!;j!!!
i i;iiiij
o Interpretation of Operative Processes for 1.
Generating Crustal Stresses in the EUS
!!Ujill!i
)
e o interpretation of Stress Orientation and Magnitude
[!N!P m.
i 4
II-s -
s r
!!f! j!h! hhl i!hh '
lI!ih!!!!!!
. I]
!!lj!]fl!!.0,i[If$!!I!hh!!f! f
]
ti!
y!!jij TECTONIC FRAMEWORK inhijbj o identification of Potentially
]lllj[
Seismogenic Tectonic Features ijg lgh.
!! - ]!0'
- i..yjj o Definition of Criteria for Assessing Activity
.g jj-1 J!f o Assessment of Activity for Each Tectonic Feature 19r 7
F/q.3
I HOW TECTONIC ISSUES WERE ADDRESSED (Workshops Academic Seminars & Interactive Meetings--involving 6 expert teams, eminent academic researchers, and methodology team)
JAN 84 Workshop 81 (Initial consideration of
" tectonic models")
MAR 84 Seminar
" Defining Tectonic Mechanisms i
Causing Earthquakes in the EUS" MAY 84 Workshop 82
" Tectonic Processes" JUN 84 Seminar
" Stress Concentration, Conditions
)
of Failure, and EQ Potential" 1
JUN 84
--interactive Mtg. 81--
JUL84 Workshop 83
" Crustal Stresses" AUG 84 Workshop 84 "Geomechanical Processes of i
Failure" AUG/SEP 84
--interactive Mtg. 82--
l OCT 84 Workshop 85
" Tectonic Framework and Seismic Sources" 5/G. d -
I i
h EVALUATION OF TECTONIC PROCESSES (Participating Experts)
Prof. Donald Turcotte
- Cornell University i
l Prof. Randall Richardson
- University of Arizona Prof. Marcia McNutt*
MIT Prof. Nafi Toksoz*
NIT Prof. Richard Sibson Univ of CA, Santa Barbara Prof. Bradford Hager
- Caltech Dr. Paul Pomeroy Rondout Assoc.
Prof. Gill Bollinger Virginia Tech Dr. Art McGarr U.S. Geological Survey.
Prof. Robert Smith University of Utah-
- Participation both in academic seminar and academic tutorial as part of Workshop 82 1
RG. 5~
4 I
4 l
STRESS REGIME, STRESS CONCENTRATION, CONDITIONS OF FAILURE, AND EARTHQUAKE POTENTIAL (Participating Experts)
Prof. Shelton Alexander
- Penn State University Dr. Mark Zoback*
U.S. Geological Survey l
Dr. Mary Lou Zoback*
U.S. Geological Survey Prof. Barry Yoight*
Penn State University l
Dr. Robert Reilinger Air Force Geophysics Lab j
Dr. Nicolas Retcliffe U.S. Geological Survey l
Dr. Stephen Kitty
.U.S. Geological Survey 4
- Participation both in academic seminar and in academic tutorial as part of Workshop 84 F/G. 6
_. -. _ _ _ _ _ _ ~.. - _ _ _. _ _ _ _ _ _., _ _ _ _ _ -.. _. _ _. _ _ _ _ _. _ _.. _ _ _..., _. _, _. _ _ _
~
Table 1 INITI AL LIST OF " TECTONIC MODELS" 1.
Reactivation of Failed Rifts /Aulscogens 2.
Isostasy 3.
-Reactivation of Decollement Structures 4
Reactivation of Mesozoic Rift Structures 5.
Epeirogenic Structures 6.
Deep-Seated Structural Boundaries 7.
Onshore Extensions of Oceanic Fracture Zones 8.
Block Tectonics 9.
Intrusive Bodies 10.
Thermal Expansion / Contraction 11.
Structural Intersections 12.
Induced Seismicity 13.
Growth Faults 14 Eastern Piedmont Ductile Shear Zones 15.
Cenozoic Reverse Faults 16.
Areas of Intensive Jointing / Fracturing 17.
None of the Above 18.
Random 19.
Initiation of New Faults 20.
Meteorite Impacts
Intraplate Tectonic Manifestations /
Processes Expected Observations A
Operative Tectonic Tectonic
+
Stress y
Processes Regime Actual Observations a
in Eastern United States Figure 3-2.
KeyEleEentsandInteractionsinAssessing Tectonic Processes and Tectonic Stress Regime FIGURE /f
h l
l KEY DEFINITIONS 4
I l
TECTONIC PROCESS t
1 A mechanism for generating and concentrating stress in the lithosphere or crust. (Tectonic
}
l processes may act over large regions, such as plate-driving mechanisms, or may be more i
localized, such as small-scale mantle upwelling.)
i TECTONIC STRESS REGif1E l
The three-dimensional orientation and magnitude l
of stress within large parts of the earth's l
lithosphere or crust.
i r
I l
l 5tG. 9 r--
,,.m._,_,-_-
m,.-.-...-.,.__.-___.,,.-_..__..-,-...,__,._.__-_.,,,,,,_-._--_-..._,m-._,.--
- ~~,%..--...,,,
_.,-.+_,-,---.-_---r..---
-- =
i l
I PRINCIPAL SOURCES OF INTRAPLATE STRESS l
1.
Plate-boundary forces (mechanisms related to plate motion such as ridge push or slab pull)
I
{
2.
Traction forces (acting at the base of the lithosphere) i
)
3.
Thickness variations of the crust and
)
lithosphere l
4.
Variations in topography and crustal density j
(e.g., lateral density contrasts) i 5.
Flexure of the lithosphere due to loading (e.g.,
i glacial, erosion / deposition, and sub-or intra-crustal loads) i 6.
Small-scale mantle convection (e.g., mantle l
upwelling and other thermal processes) j 7.
Membrane tectonics (due to earth's ellipticity) e F/G. Io e
DR/V/NG t/ECH4N/St/S
/.
Nor/zonte/ ~ Dens /ty Contrasts (A) Ocean /c L /thosphere F, m AN&
N.....~
I l
)F i
R I
I F, a a few x 10' Nm
~I N (geoid) oc
( AW)2 (water depth) oc 1
I t
i I
F (slab pull) sp i
i Af F
s a few x 10 Nm 'I sp s
--based only on A jo
_)
FIGURE 17.
flG. lf
a SOURCES OFSTRES&
Renewdle hea_-
Replanal Lac 2/
Renewdle I. plate-boundary forces /
/
/
I
/
/
- 2. Tract /on forces J Thickness var /ations of E
/
/
crust & //thosphere 4 Vorlsflons In topogralohy
& crusts / dens /ty
/
/
- 5. Flexure of //thosphere 2
due to load /ng
/
/
- 6. Therms / I
/
/
7 Membrane (Earth's ell /pt/ city) I
/
/
- 8. Eros /on & sed /montst/m I
/
/
Superscr/ pts
/,2,3 Ind/cate, respect /vely, first,
second,
and th/rd-order Importace.
FIGURE 10.
~
fic. It
I i
i MANIFESTATIONS / EXPECTED OBSERVATIONS RELATED TO TECTONIC PROCESSES l
(including Specification of Spatial and Temporal Scales) i i
- 1. Stress Orientation and Magnitude l
)
- 2. Seismicity (Size, Spatial Distribution, Nechanism) i j
- 3. Crustal Deformation (Type, Rate) 4.
Geophysical Anomalies (Type, Amplitude)
--isostatic gravity
- geoid
--heat flow
--variations of seismic velocity within mantle
~
l FIG.
15 1
i f
WHAT ARE THE IMPORTANT SOURCES OF i
INTRAPLATE STRESS IN THE EUS?
i o
No active tectonics in the EUS (in the sense that there are no plate boundaries, no incipient I
rif ting, and no hot spots) 3 I
{
o The two important sources of stress are:
l.
Ridge-push (dominant)--generated at the i
Mid-Atlantic ridge and transmitted through the lithosphere acting as a stress j
guide. (Stresses induced by basal driving drag due to convection represent another j
possibility, but ridge-push forces are preferred from global models.)
2.
Stresses associated with lithospheric flexure. (Likely to be important along the Atlantic and Gulf coasts.)
1 1
o Renewable sources of stress are dominant, and l
residual stresses can be neglected.
l o
Erosion /redeposition and glacial rebound are the j
most likely active processes redistributi,ng mass and affecting the stress field; but very low strain i
rates and stresses are involved (100's of bars).
I FIG. IV
8 13 I
I L
i 7
's 1
........t.
... j j
/.
lf
- y.
f
,4 40' p
MID - CONTINENT 4 4=,
~,
W 5
8 O
y.
\\ye'
~
4 p
\\
e#
"" % 4 Y
, 30'
.C O A ST
/
/
V sq
~
~
~
~
~ " "
g l
y,s,
~.
i i
i 1
~ r.-
100*
90' 80' Figure 1.
Generalized stress map of Central and Eastern United States (modified after Zoback and Zoback,1980). Relatively uniform northeast-southwest compression seems to persist through the mid-continent and Southeastern U.S.
FIGURE 14 l~lG l(o
Tentatively Planned Sensitivity Studies and Consistency Checks Concern /Need Test / Test Element Assess sensitivity to key elements e
Scientific uncertainty (generic) matrix elements of the EPRI methodology e
Data uncertainty (feature-specific) matrix elements Test sensitivity to and elucidate e
Test sensitivity to assumed magnitude uncertainties statistical analyses of earth-quake catalog e
Probability-of-detection functions -- examine assumptions and test sensitivities Assess sensitivity to other e
Site-vicinity vs. regional interpretations aspects of the EPRI methodology Form of upper-bound magnitude probability e
distributions Source-zone interdependencies e
Assess sensitivity to earth-e Revise current earthquake catalog and test effect quake catalog deficiencies on hazard calculations e
Evaluate historical intensities and felt areas in New England; estimate magnitudes, compare with calculated " uniform" magnitudes; assess impact on hazard Check results for consistency e
Compare result with EQHIST calculations with historical method EPRI 7/9/85
'O A_nSSP ksf i
l 4
OVERVIEW OF LLNL'S ONG0ING COMPARATIVE / SENSITIVITY STUDIES l
D. BERNREUTER R. MENSING J. SAVY i
NCR-EPRI MEETING ON i
COMPARISONS OF SEISMIC HAZARD METHODOLOGIES JULY 9, 1985 I
gnsR__
LS THE MAIN PURPOSE OF OUR PRESENTATION TODAY IS TO 9
DISCUSS ONG0ING COMPARATIVE TASKS.
8 PROVIDE INITIAL RESULTS.
O ANSWER QUESTIONS.
FOR THOSE NOT FAMILIAR WITH OUR STUDY, WE FIRST GIVE 8
A BRIEF OVERVIEW 0F THE LLNL STUDY.
O A
SUMMARY
OF THE MAJOR DIFFERENCES BETWEEN THE LLNL, EPRI, AND USGS STUDIES.
EJ m
Z
~
Z O
U H
1 Z
O h
H Z
M M
O
~
g J
M D
k U
W Je
=
M wo O
J M
ao H
A O
gg A
E W
a M
'M O
M W
O b
ZZ m
H O
W 40 0
Z Z
M O
E
~
b<
O O
y O
W g
Om Z
W e
O ed b
S OeO W
M O>
W O
WrW A
4 OW e
J O
ZWZ J
x w
W O
MEM M
HU M
M Z
"EJ W
J J-a I
wp o
O W
A*M C
kbk E(
Z W
e mm a
M W
mp D
W MEM Om J
O H"H O
y W
J O
J J
wM Z
O M
=1D O
o 4
cm am MOM W
M a
OW EW W
W g.
ZH
-H m3M H
J ZU e
H-W O
We a
WM EM E-C M
EZ O
E
">W Z
W a
a MWW W
om WW WM WM4 M
M JO m
OW OW G
W U
H ZMA W
Z
>H 3
ZW H
"WD WE W
OJ ZZ W
c4
-W WW
@ AM H
D w
WM EW M
O J
W A
- W -@
R T
J m
v1 OW O
-O JZ k
O H
g J
W&
'M E
Z
-O W
A3 JZJ O
H ZW W
aW WO
-W" W
W
-y a
OW M&M E
W AZA N
mW m>
W M
My M
MM MO D
mW m
G GO O
-m
-J
-J
~
i e M i
i l
I i
l I
e O
a$
NZ O
-~
NZ MI se M
MW ma m
mm QW G,
gg
- a.
-H
-E M
glll:gs is,
mm OUR APPROACH IS TO QUANTIFY THE UNCERTAINTY IN THE ESTIMATES OF THE SEISMIC HAZARD IN THE EUS BY USE OF EXPERT JUDGEMENT TO SUPPLEMENT THE PAUCITY OF DATA AND THEORIES 8
MANY EXPERTS TO ENSURE AN ACCEPTABLE SAMPLE OF THE UNCERTAINTY IN THE FORM 0F DIVERSE VIEWS OF THE SEISM 0-TECTONIC PROCESS OF THE EUS TWO PANELS OF EXPERTS FOR INPUT DATA:
PANEL FOR ZONATION AND SEISMICITY PANEL FOR GROUND MOTION MODELING 8
INTERACTIVE MULTI-STEP ELICITATION AND FEEDBACK PROCESS TO OBTAIN BEST ESTIMATE AND UNCERTAINTY FOR EACH OF THE INPUT PARAMETERS 0
DEVELOPED AN ANALYTIC, METHODOLOGY T0 INCLUDE AND PROPAGATE UNCERTAINTIES QUANTIFIED BY THE EXPERTS 8
USE OF A PANEL TO HELP ASSURE THAT THE ELICITATION / ANALYSIS PROCESS IS REASONABLE AND SUGGEST CHANGES WHICH WOULD IMPROVE THE POST FEEDBACK RESULTS EG-85-069 i
lnSSP Q
WE ARE EXECUTING A NUMBER OF TASKS TO ASSIST NRC IN THEIR ASSESSMENT OF THE l
SIGNIFICANCE OF THE DIFFERENCES BETWEEN THE LLNL AND EPRI STUDIES.
1.
DEVELOPMENT OF HAZARD CURVES TO DIRECTLY COMPARE TO EPRI'S RESULTS.
e SAME LOWER B0UND OF INTEGRATION e
SAME GROUND MOTION MODEL(S) 2.
SIMPLE ASSESSMENT OF THE DIFFERENCES BETWEEN LLNL'S, EPRI'S, AND USGS'S HISTORICAL CATALOGS.
e IDENTIFY MAJOR REGIONAL DIFFERENCES BETWEEN CATALOGS.
e EXAMINE IMPACT OF CATALOGS ON THE HISTORICAL ANALYSIS FOR EACH OF THE TEN SITES.
e IN CONJUNCTION WITH TASKS (3) AND (4), EXAMINE THE EFFECT OF USING VARIOUS CATALOGS ON THE EQ. RECURRENCE MODEL AT FOUR SELECTED SITES.
n m wm -
WE ARE EXECUTING A NUMBER OF TASKS TO ASSIST NRC IN THEIR ASSESSMENT OF THE SIGNIFICANCE OF THE DIFFERENCES BETWEEN THE LLNL AND EPRI STUDIES. (Corn. )
3.
ASSESSMENT OF THE IMPACT OF USING A UNIFORM APPROACH FOR THE DEVELOPMENT OF THE EARTHQUAKE RECURRENCE PARANETERS FOR THE SEISMICITY PANEL MEMBER'S Z0 NATIONS.
e APPLY UNIFORM CORRECTIONS FOR INCOMPLETENESS, AFTERSH0CKS, AND RELATIONS BETWEEN MAGNITUDE AND INTENSITY SCALES.
e APPLY UNIFORM METHOD TO OBTAIN ESTIMATES OF THE A AND B VALUES FOR ZONES WHICH CONTRIBUTE MOST SIGNIFICANTLY TO THE HAZARD AT FOUR REPRESENTATIVE TEST SITES.
e RECOMPUTE THE HAZARD AT FOUR REPRESENTATIVE TEST SITES USING THE NEW A AND B VALUES.
4.
ASSESSMENT OF EPRI'S APPROACH FOR THE DEVELOPMENT OF EARTHQIJAKE RECURRENCE MODELS.
e OBTAIN SOFTWARE DEVELOPED BY VENEZIAN0 FOR EPRI'S PROGRAM, i
EXAMINE THE IMPACT THAT DIFFERENT OPTIONS SPECIFIED BY EPRI'S TEAMS HAVE e
ON THE A AND B VALUES FOR THE ZONES WHICH CONTRIBUTE MOST TO THE HAZARD AT FOUR REPRESENTATIVE TEST SITES.
{
RECOMPUTE THE HAZARD AT FOUR REPRESENTATIVE TEST SITES.
e
o Anssa
.m-smemw -
WE ARE EXECUTING A NUMBER OF TASKS TO ASSIST NRC IN THEIR ASSESSMENT OF TH SIGNIFICANCE OF THE DIFFERENCES BETWEEN THE LLNL AND EPRI STUDIES.
(CONT.)
l S.
SENSITIVITY STUDIES l
e IMPACT OF SITE CORRECTION.
IMPORTANCE OF THE CONFIDENCE LEVEL ASSIGNED BY THE EXPERTS TO TH e
CONTRIBUTION OF THE COMPLEMENTARY ZONE.
e CONTRIBUTIONS OF DIFFERENT DISTANCE / MAGNITUDES TO THE HAZARD.
e 6.
ASSESSMENT OF THE SIGNIFICANCE OF THE DIFFERENCES BETWEEN EPRI'S LOGIC TR METHODOLOGY AND LLNL'S MONTE CARL 0 SIMULATION METHODOLOGY.
LLNL AND EPRI JOINTLY DEFINE A TEST CASE (S).
e LLNL USES ITS METHODOLOGY AND EPRI USES iTS METHODOLOGY FOR TEST e
CASE (S) TO OBTAIN COM?ARITIVE RESULTS.
D NW O
D
&M tr) f./)
H ct:
M A
LAJ LAJ A
TN
-c
_2 0
=2" ct::
M D
A
=
C3 C1.
=2",
J Z
J 1""
J C3 af Z
w Z
H ta.J i J LAJ LAJ
'. 2 I
Z H
H
>==
LAJ CQ M
M LAJ LAJ M
N N
w M
M
- t et
=
- u)
T T.
C2C LaJ c\\
C1.
<x: t>
a.
E JZ LaJ LaJ
- LAJ 2". Ct:
- =
>=
- La.J c3 C
M LA.
D D
LA.
H LA. -
M C/)
OC M
M cx1 H
~
~
LAJ Z J
2" 2
ti' E =t--
_J A
"3 t'r!
J LAJ ZO 4
LAJ e
S
- ,g "[
O A
LAJ M m $s cr:
LAJ LaJ af VnI ww LAJ
-c G
h >z
1 ut wo,s WHILE THERE ARE A NUMBER OF SIMILARITIES BETWEEN LLNL'S AND EPRI'S STUDIES, i
THERE ARE SOME IMPORTANT DIFFERENCES.
(Cost.)
l THE LLNL STUDY EMPHASIZES THE RETAINING OF THE DIVERSITY OF EXPERT JUDGMENT AND POSSIBLE UNCONVENTIONAL CONCEPTS.
CHOICE OF A LARGE GROUP 0F DIVERSE EXPERTS e
NO ATTEMPT TO PROVIDE A COMMON FRAMEWORK OR CONSENSUS e
DATA SETS UNTIL THE FEEDBACK MEETING DID NOT REQUIRE EXPERTS TO FORMALLY DEFEND OR JUSTIFY e
THEIR MODELS USE OF AN INDEPENDENT PANEL FOR SELECTION OF GROUND e
MOTION MODELS
a w wm m WHILE THERE ARE A NUMBER OF SIMILARITIES BETWEEN LLNL'S AND EPRI'S STUDIES, THERE ARE SOME IMPORTANT DIFFERENCES.
(Cont.)
THE EPRI STUDY EMPHASIZES THE DEVELOPMENT OF CONSENSUS IN DATA SETS, DOCUMENTATION AND DEFENSE OF MODELS.
e ONLY SIX TEAMS FORMED e
CONSIDERABLE EFFORT WAS SPENT IN THE DEVELOPMENT OF COMMON DATA SETS e
CONSIDERABLE EFFORT WAS SPENT IN DEVELOPMENT OF COMMON CONCEPTS AND COMMON UNDERSTANDING.
e DID REQUIRE DOCUMENTATION AND DEFENSE OF MODELS e
GROUND MOTION MODELS AND WEIGHTS DEVELOPED BY ANALYST i
l
WHILE THERE ARE A NUMBER OF SIMILARITIES BETWEEN LLNL'S AND EPRI'S STUDIES, THERE ARE SOME IMPORTANT DIFFERENCES.
(Cont.)
THERE IS A DIFFERENCE IN PHILOSOPHY USED TO DEVELOP THE SEISMIC SOURCE ZONES.
LLNL'S EXPERTS WERE ASKED TO MODEL ALL SEISMIC SOURCE ZONES.
e EPRI'S TEAMS WERE ASKED TO ONLY MODEL IDENTIFIABLE SOURCE e
ZONES WITH POTENTIAL FOR MODERATE TO LARGE EARTHQUAKES.
O lnSSP NUCl(AA SYSTOAS SAF(TV PROGAAM WHILE THERE ARE A NUMBER OF SIMILARITIES BETWEEN LLNL'S AND EPRI'S STUDIES, THERE ARE SOME IMPORTANT DIFFERENCES. (CONT.)
IT APPEARS TO US THAT EPRI'S STUDY INCORPORATES A STRONG RELIANCE ON MAT (E.G., POISSON MODEL) FOR ACCOUNTING FOR INCOMPLETENESS, AFTERSH0CKS, CLUSTERS, AND THE COMPUTATION OF THE A AND B VALUES.
BY CONTRAST, THE LLNL STUDY EMPHASIZES EACH EXPERT'S INDEPENDENT SUBJECTIVE TREATMENT.
gnSSP o
=<= see wm -,
l WHILE THERE ARE A NUMBER OF SIMILARITIES BETWEEN LLNL'S AND EPRI'S STUDIES, THERE ARE SOME IMPORTANT DIFFERENCES.
(Coni.)
8 LLNL METHODOLOGY USES A MONTE CARLO SIMULATION APPROACH FOR THE UNCERTAINTY 9
EPRI METHODOLOGY USES A LOGIC TREE APPROACH FOR THE UNCERTAINTY ANALYSIS.
A955L_
LG SENSITIVITY RUNS 1.
COMPARE RESULTS FOR LLNL Vs EPRI ZONATION & SEISMICITY 2.
ANALYZE EFFECT OF LOWER B0UND OF INTEGRATION (M )
o 3.
ANALYZE EFFECT OF LLNL GM MODELS l
EG-85-070 u
g,nSSP uum smu smw -
Z0 NATION AND' SEISMICITY (CASE 1 RUN) 8 LLNL Z0 NATION & SEISMICITY 0
Mo = 5.0- (SAME AS EPRI) e G.M. MODEL
= NUTTLI, 1984 (SAME AS EPRI PRELIMINARY)
DIFFERENCES BETWEEN LLNL Vs EPRI ARE POSSIBLY DUE T0:
SEISMICITY (A & B, Mg)
Z0 NATION OVERALL METHODOLOGY EG-85-070
l EUS SHC. SENSITIVITY TASK LLNL-EPRI COMPARISON gat =NUTTL I 1984, M0=S.0, NO SITE CORRECTlON PERCENTILES = 15,0.50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
- --e G'P RT
-1 10 t
-2 10
-3 10
-4 10
-5 10
-6 10
-7 10 S
8 8
8 8
8 8
n n
n 4
ACCELERATION CWSEC"2 LIMERICK
'8
.[
,.-n c.
-g a
y g
-.. -,. g:f....
EUS SHC. SENSITIVITY TASK LLNL-EPRI COMPARISON gat =NUTTLI 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10
-3 10
-4 10 r
-5 10 m
-6 10 m
-7 10 S
8 8
8 8
8 8
N N
m n
ACCELERATION C WSEC**2 SHEARON HARRIS 4
4.
g
oes EUS SHC. SENSITlVITY TASK LLNL-EPRI COMPARISON GM=NUTTL I 1984, MO-5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 to
-3 10 m
-4 10 m
-5 10 e
-6 10
-7 10 L
8 8
8 8
8 8
8 S
8 N
N m
m ACCELERATION C WSEC**2 l
RIVER BEND 1
1
~
~
(gy'
.-..n-EUS SHC. SENSlTIViTY TASK LLNL-EPRI COMPARISON N TLI 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10 m
-3 10
-4 10 10
-6 10
~7 10 S
8 8
8 8
8 8
8, 8
a a
n n
ACCELERATION CWSEC"2 BRAIDWOOD
\\
l l
1
.a
. -. ~.
.,,r---
O EUS SHC. SENS1TIVlTY TASK LLNL-EPRI COMPARISON GM=MJTTLI 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10 m
-3 10
-4 10 r
-5 10
-6 10 m
-7 10 S
8 8
8 8
8 8
. *S e
n n
n n
ACCELERATION CM/SEC"2 WOLF CREEK
f 1
en I64 l
EUS SHC. SENSITIVITY TASK LLNL-EPRI COMPARISON GM=NUTTLI 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10
-3 10 m
4 10 m
-5 10
-6 to r
-7 10 S
8 8
8 8
8 8
8 8
N N
M M
4 4
ACCELERATION C WSEC**2 WATTS BAR l
l
EUS SHC. SENSITIVITY TASK LLNL-EPRI COSARISON gat =NUTTLI 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10 10 m
)
-3 10
-4 10 m
-5 10 m
-6 to r
-7 10 S
8 8
8 8
8 8
8 8
e e
n n
n n
ACCELERAT10N CWSEC"2 V0GTLE
l l
EUS SHC. SENSITIVlTY TASK LLNL-EPRI COMPARISON gat =NUTTL 1 1984, M0=5.0, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10
-3 10
-=
-4 to r
-5 10
-s 10
'r
-7 10 S
8 8
8 8
8 8
n n
n 4
ACCELERATION CWSEC"2 MILLSTONE
'1
~
>03 EUS SHC. SENSITIVITY TASK LLNL-EPRI COh4'AF ' EU4 GM=NUTTLl 1984, M0=5.0, NO SITE CORRECTIL*l PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING,.LL EXPERTS
-1 to
-2 10 m
-3 10
-4 10
-5 10
- 10 r
j[
S 8
8 8
8 8
8 S
S n
a n
m ACCELERATION CWSEC"2 MAINE YANKEE.
A9sst_
ts LOWER BOUND Mo 1.
SAME AS CASE 1 RUN, EXCEPT Mo = 3.75 0
IN MOST CASES, NO DIFFERENCE AT HIGH VALUES OF ACCELERATION BRAIDWOOD, V0GTLE, MAINE YANKEE 8
IN SOME CASES FACTOR OF 1.5 TO 2 AT LOW VALUES OF ACCELERATION RIVER BEND, MILLSTONE NOTE:
THESE SENSITIVITY RUNS ARE MADE WITH A LIMITED NUMBER OF SIMULATIONS.
MEDIAN'IS WITHIN APPR0XIMATELY 1%
MAINE YANKEE EG-85-070
-? "? ?
3.g.,;.
EUS SHC. SENSITIVITY TASK LLNL-EPRI CO PARISON G4M4JTTLt 1984, W0=3.75, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 10
-3 10 9
-4 10 9
33 5.0
-5 10 3 '7f
$.0
_s lf
'O S.O
-7 10
-8 10 S
8 8
8 8
8 8
8, n
u n
n ACCELERATION CWSEC"2 BRAIDWOOD l
i
7-; api ju ' ':t.+ %%:
$$gj.Il ih?J l
DJS SHC. SENS1TIV1TY TASK LLNL-EPRI C0hPARISON GAMAJTTL1 1984 N0=3.75, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 to
-3 10
-4
\\ 'N'%
10 e
N.
gg 4
~%,
10 5.o,
16
-6
,p 10 m
N 31f
% N
-7 Nw6.,
10
-6 10 S
8 8
8 8
8 8
8 2
e n
n n
n w
ACCELERATION CWSEC"2 RIVER BEND i
9
~,
~.
ly:+.s ATl' ;.
'th{' Ysif>h;+:b:?']
^
$k?h"f' 5
l EUS SHC. SENSITIV1TY TASK LLNL-EPRI C0bMARISON l
Ge**AJTTLI 1984, WO=3.75, NO SITE CORRECTION ~
PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 10
-3 N
10 N..,
N.
\\
?$
4 S.p 10 m
N.
') 'K 50
-5 10
'N x-.
3,4
- 5.,
-6 10 m
-7 10 r
-6 10 R
8 8
8 8
8 R
R, n
n n
n ACCELERATlON CWSEC"2 VOGTLE
.._.,-._,..,____,_.___..m
. ~...
4 y[fy; N
,g f,.
'~
EUS SHC. SENSITIVITY TASK LLNL--EPRI C0hPARISON Get:NUTTLI 1984, M0=3.75, NO SITE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 to sN N
-3 10 Ns
\\
6.0 10
's.16 -
af 50
-5
~ ~.. _
10 1$
5.p
-6 10
-7 10
-6 10 S
8 8
8 8
8 8
8 8
n n
n n
ACCELERATION CWSEC"2 MILLSTONE i
9 v
7
..-r7
~
i5*
.$ll2
.d ;
~ fi g.
.f }M.
N EUS SHC. SENS1TIVITY TASK LLNL-EPR1 00hPARISON W4JTTLI 1984, M0=3.75, NO SITE CORRECT 10N PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 10
' s..
.6
' s.,N's.
-3 10 6g 76
-4 5.0 ' N 10 r
N..
~., _
-5 10
~~
-6 10 m
-7 to
-4 10 S
8 8
8 8
8 8
8 8
N N
m m
ACCELERAT10N CWSEC"2 MAINE YANKEE l
i l
l r0 lnSSP
~
NUG(AA SYSTOAS SAKTY PROGAAM GROUND MOTION MODELS COMPARES LLNL GROUND MOTION MODELS RESULTS WITH NUTTLI, 1984 ALONE RUNS WITH Mo = 5, LLNL ZONATION & SEISMICITY DIFFERENCES:
ALMOST NONE:
BRAIDWOOD, RIVER BEND SMALL:
SHEAR 0N HARRIS SUBSTANTIAL:
V0GTLE, MAINE YANKEE i
NOTE:
1-2% VARIATION DUE TO LIMITED NUMBER OF SIMULATIONS i
EG-85-070
._t
^
[M
( dJ;.,.
4-j.4'(J :f 3.yp I'
M i
EUS SHC. SD451TIVITY TASK LLNL-EPRI COMPARISON 3
i mat L LLNL DJS, W0=5.0, NO SITE CORRECTION PDtCD4 TILES = 15.0,50.0 AND 85.0 9
HAZARO CURVE USING ALL XPERTS hD Mb Wohaw mo/ils I
A - All L.t, ML gram.
3
,I C0 f Yt C45 rn He s S.b b-k.&i j re n.&
W0% tw v%e)L(LLSIY by
)
-1 a s.o 10 r
1
-3 10 H-D.
i I
~
A.
1 f
I
\\
b
's,
4 10 N
A
's -
-4 to r
i
-7 10
-4 10 R
8 2
8 8
8 8
8 2
e n
n n
n s
ACCELERATION CWSEC"2 L
SHEARON HARRIS j
1 9
.c,.--
.-,..--,,.-...--,..,-...n-.-
~~.--e
. _ = -.. -.-
w.w.;....,.,. r. c..,,..,,,,,1..,.......,4...
,,,..,, g,.
7 4
r._1 s (._4pvq.c '
ra v). f-
. )Q
-.,'.1:.
- g y,,.
i EL*S SNC. SENSITIVITY TASK LLNL-EPRI CohPARISON W LL LLNL EUS. WO=S.0, NO S1TE CORRECTION PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 10 r
-s to 9
(
-4
~'-%..'.
10 m
9 s%
' - ~. '
t s.
~ __
N
'~~
-s 10
~
'N
.s
-4 N.
10
' N -,
i
~. _,
-7 10 1
i
-e l
10 1
2 8
2 8
2 g
g l
a u
f ACCELERATlON Cn4/SEC"2 4
i BRAIDWOOD I
=
~ v v r.., y c e w..:,g,. c y ra,p y n~..a. e w t. % : w. 9 y % : m f Q :
M
,w. s.
Sf ' ;;&wh essust}E* thW9l% %rt.hiaFMafrws~% uw+4t&
' f^:;$5+kTl:
A 4?? 4[ h t' p):f1V *~., :',C
- ~
49;y~!'.m. bvi% *
- '.-:dh. )4'i19.Y. j y,. a.fp e y o. u.s. a..
- ~.
.:,. m.va..;;e,;p
- m. : t-
.,,, a.. -
..s.:.
%. tWS SIC'. SDdSITIVITY TASK LUE.Pf OwAftlStM -
'i er* '" 5' Ge>ALL LUS. DJS. WD=S.0 NO 5iTE C0ftRECTlON,
4 2.,
PDICDdTILES er 15.0,50.0 AND 45.0
~!'
HAZAftD CURVE USING AL( OtPDtTS'
-2 10
-3 10
)
-4 10 m
-6 10 m
N \\,
s 4
10 I
i
-7 to 1
-4 to R
8 8
8 8
8 8
8, 2,
n n
n n
l l
ACCELERATION C$EC"2 3
{
RIVER BEND i
,----------.__-___e.,
.,~,-,_-
__,___.,_,_,___,---,.,,,_,__-,,.__--_,._-_-__,w
l l. - -
i n c.-
f;;-
9'gl,2by.}F 4
. t zv, 1
EUS SHC. 5ENSITIV1TY TASK LLNL-CPRI C0hPARISON i
QM=ALL LLNL EUS, M0uS,O, NO SITE CORRECTION PERCENilLES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS l
-2 10
~
r
_3 N
sN s'N_
10
, ~,
-*9$4, g \\
4 10 s
1Y 10
-7 10 l
4 10 S
8 8
8 8
8 8
8, 8,
n u
n n
ACCELERAT lON CWSEC"2
\\
VOGTLE
_,,_-,____-_-,,y-f.-----_-_r--,
r
+.
.p;; 9i 1p.ggr.p7q9..
7 M. ;sj6'.
4 EUS SE. SD451TIVlTY TASK LLNL-CPR1 00hPARISON MLL LLNL EUS, W.0, NO $1TE CORRECTlON PERCD4 TILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-2 10
-3 i
l 10 g'%,
y 10 N
--...m__
-5 10 s s
~
~sy
-4 10
-7 i
10
-6 10 8
8 8
8 8
g g
g a
n ACCELERATION ChVSEC"2 i
MAINE YANKEE il l
9
gnSSP
- o
.uta se-w -
LOWER BOUND Mo 2.
COMPARES UCID 20421 REPORT RESULTS (Mo= 3.75) WITH SAME RUNS WITH Mo = 5.0 SHOWS INFLUENCE OF Mo FOR LLNL ANALYSIS ALONE l
BRAIDWOOD, RIVERBEND, V0GTLE, MAIN YANKEE DIFFERENCE FROM ALMOST NONE TO FACTOR OF 1-2 EG-85-070
V
,g..
3 g.
EUS SHC. SD45iTIV1TY TASK LLNL-EPRI C0hPARISON UC10 20421, WlTH EPRI SCALES PERCDITILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS
-1 10
-2 10 r
N
-3 10 s'-
-4 10
-\\
i s
-5 10 m
4 W
-4 10
-7 y
10 E
8 8
8 8
8 R
S e
n n
n m
s ACCELERAT10N OVSEC"2 BRAIDWOOD LLNL %Q R o 3.7 8 A L = 5 o gm
-se..,
'O;;[q:.Yr.'
, l7,<T-l l 1
n*
EUS SNC. SENS1TIVITY TASK Lt.NL-EPRI C0heARISON UCID 20421, WITH EPRI SCALES PERCENTILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS a
-1 10
-2 10
-3 10 s
\\ N
-4 10
- \\
1Y
-6 10
-7 l
10 S
8 8
8 E
S S
e e
n n
n ACCELERATION C$EC"2 RIVER BEND J
.;3 g 01.%p.* SPf{fSE.Nf' N' '
- G '.
m DJS SHC. $ENSITIV1TY TASK LLNL-EPRI COPARISON UCID 20421, WITH EPRI SCALES PERCENTILES =: 15.0,50.0 AND 85.0 HAZARO CURVE USING ALL EXPERTS
~1 10
-2 10
-3 10 s
-4 10 w%
-5 10
-4 10 5
-7 to R
8 8
8 8
8 S-8 8
N N
m n
ACCELERATION C$EC"2 VOGTLE
' f[';ijrl(3. WQjpg;c,sppfti,pxi U-riy l
,a 'tptg.yc, ygq.p 4 : -
-. ; -)(_..
j ;pg i
DJS SHC. SD45ITlVITY TASK Lt.NL-EPRI COnFAR1 SON UC10 20421, WITH EPRI SCALES PERCD4 TILES = 15.0,50.0 AND 85.0 HAZARD CURVE USING ALL EXPERTS i
-1 10
-2
' 's 10
-3 10
-4 10
-5 10 r
-4 10 m
-7 to S
8 8
8 8
8 8
8 8
a a
m n
ACCELERATlON ChVSEC"2 MAINE YANKEE 1
_._