United States Nuclear Regulatory Commission - Protecting People and the Environment

478th Advisory Committee on Reactor Safeguards (ACRS) - December 8, 2000

                                                               390
 1                      UNITED STATES OF AMERICA
 2                    NUCLEAR REGULATORY COMMISSION
 3                                 ***
 4                     478TH ADVISORY COMMITTEE ON
 5                      REACTOR SAFEGUARDS (ACRS)
 6
 7
 8                 U.S. Nuclear Regulatory Commission
 9                        11545 Rockville Pike
10                             Room T-2B3
11                         Rockville, Maryland
12
13                      Friday, December 8, 2000
14
15              The above-entitled meeting commenced at 8:30 a.m.,
16    pursuant to notice, the HONORABLE DR. DANA A. POWERS,
17    chairman, presiding.
18
19
20
21
22
23
24
25
.                                                               391
 1    ACRS COMITTEE:
 2              DR. DAN A. POWERS, Chairman
 3              DR. GEORGE APOSTOLAKIS, Vice Chairman
 4              DR. THOMAS S. KRESS, ACRS Member
 5              MR. JOHN D. SIEBER, ACRS Member
 6              DR. GRAHAM B. WALLIS, ACRS Member
 7              DR. ROBERT L. SEALE, ACRS Member
 8              DR. WILLIAM J. SHACK, ACRS Member
 9              DR. ROBERT E. UHRIG, ACRS Member
10              DR. MARIO V. BONACA, ACRS Member
11              DR. LEITSCH, ACRS Member
12              DR. JOHN T. LARKINS, ACRS Executive Director
13              MR. JOSEPH MURPHY
14
15
16
17
18
19
20
21
22
23
24
25
.                                                               392
 1                        P R O C E E D I N G S
 2                                                     [8:30 a.m.]
 3              CHAIRMAN POWERS:  The meeting will now come to
 4    order.  This is the third day of the 478th meeting of the
 5    Advisory Committee on Reactor Safeguards.
 6              During today's meeting, the Committee will
 7    consider proposed modifications to the Commission's Safety
 8    Goal Policy Statement for Reactors.  We'll have an
 9    unfortunately abbreviated discussion of the NRC's safety
10    research program, a discussion of future activities, and a
11    report of the Planning and Procedures Committee,
12    reconciliation of ACRS comments and recommendations.
13              We'll have the election of ACRS officers for
14    calendar year 2001, and we'll discuss some proposed ACRS
15    reports.
16              This meeting is being conducted in accordance with
17    the provisions of the Federal Advisory Committee Act.  Dr.
18    John T. Larkins is the Designated Federal Official for the
19    initial portion of the meeting.  We have received no written
20    comments or requests for time to make oral statements from
21    members of the public regarding today's session.
22              A transcript of portions of the meeting is being
23    kept, and it is requested that speakers use one of the
24    microphones, identify themselves, and speak with sufficient
25    clarity and volume so that they can be readily heard.
.                                                               393
 1              I'll remind members that we're scheduled to have a
 2    picture taken, and that during the lunchtime break, we will
 3    be interviewing possible candidates for membership of the
 4    Committee, and that we are going to celebrate Christmas with
 5    the Staff today.
 6              MR. LARKINS:  Dana, since it's going to be hot
 7    food, they're going to try to start at 11:30 or as close to
 8    11:30 as possible.
 9              CHAIRMAN POWERS:  Okay, we will endeavor to stay
10    closely on schedule, and I don't think that's a problem.
11              Any members have any other opening comments they'd
12    like to make?
13              [No response.]
14              CHAIRMAN POWERS:  Then we'll move to the first
15    item of business, which is proposed modification of the
16    Commission's Safety Goal Policy Statement for Reactors.
17              Professor Apostolakis?
18              DR. APOSTOLAKIS:  Thank you.  I suspect that this
19    whole effort was started with the report that we sent to the
20    Commission in 1997, suggesting or recommending that the core
21    damage frequency goal of ten to the minus four be elevated
22    to the level of a fundamental goal.
23              We also recommended several other things such as
24    using a three-Region approach and so on.  Then a lot
25    happened since then, and the Commission, in a Staff
.                                                               394
 1    Requirements Memorandum this year, approved the
 2    recommendations that the Staff made in SECY 00-0077.
 3              They specifically rejected the elevation of the
 4    CDF goal beyond the fundamental level, among other things,
 5    and the Staff, following that SRM, has developed this policy
 6    statement, which, of course, abides by the Commission's
 7    wishes.
 8              And Joe Murphy, a relative newcomer to this
 9    subject, is here today to explain to us, how life was
10    created several thousand years ago.
11              [Laughter.]
12              DR. APOSTOLAKIS:  Mr. Murphy?
13              DR. KRESS:  Are we expected to have a letter on
14    this subject?
15              DR. APOSTOLAKIS:  I don't know.  The Committee
16    will have to discuss this.  I have mixed feelings about it,
17    but right now I'm inclined to say no, but we'll see, but I'm
18    sure Joe will insist on a letter.
19              Mr. Murphy.
20              MR. MURPHY:  Thank you.  I thought a template that
21    had a dark-colored border would be appropriate for this
22    subject.
23              [Laughter.]
24              MR. MURPHY:  I don't know how many times I've
25    addressed the Committee on this subject, but it's got to be
.                                                               395
 1    pushing ten now.
 2              DR. KRESS:  You have to keep doing it till you get
 3    it right.
 4              MR. MURPHY:  I'm afraid you're right.
 5              MR. LARKINS:  The other didn't fit, Joe.
 6              MR. MURPHY:  I think George already gave most of
 7    the background.  We proposed several modifications to the
 8    paper in SECY 00-0077.  We discussed this with the Committee
 9    back in February.
10              The Committee, in turn, wrote a letter that really
11    didn't address the specifics of what we had proposed, but
12    told the Commission that, in fact, a different kind of
13    policy statement was also needed.
14              I believe you referred to it as a risk-informed
15    policy statement that would include the consideration of the
16    three-Region approach, the concept of risk limits for
17    individual plant applications, to be expressed in terms of
18    CDF and LERF, and guidance on defense-in-depth to address
19    uncertainties in the risk assessments.
20              The Commission, in turn, issued its SRM on June
21    27th.  They approved the modifications that we proposed with
22    two exceptions:  One is, they disagreed with the elevation
23    of the qualitative statement on prevention of severe core
24    damage accidents to a qualitative goal.
25              They disapproved that, but they also left a
.                                                               396
 1    statement in the policy statement.  And they also
 2    disapproved the recommendation to include a statement that
 3    there be no adverse impact on the environment in the Safety
 4    Goal Policy Statement, directly, but did say we could put in
 5    some phraseology to consider the need to minimize adverse
 6    environmental impact in regulatory decisionmaking.
 7              DR. KRESS:  They didn't like that absolute value
 8    of no adverse?
 9              MR. MURPHY:  Yes.  And I think what was there is
10    that particularly Commissioner Merrifield was sensitive to
11    the problems that EPA has run into with similar phraseology,
12    and so they were -- it was a semantic problem, a legal
13    problem.
14              The Commission, as I said, did give us essentially
15    an additional phrase to put in that dealt with that subject
16    without the no, and then directed that the Policy Statement
17    of safety goals or goals not limit, which was also something
18    that we had initially proposed.
19              That comes out of the June 15, 1990 SRM. As you
20    are aware, what this is, you know in a way, is a
21    housekeeping job.  We've taken the information that's in the
22    June 15, 1990 SRM, which is basically un-retrievable until
23    we started this effort.
24              There were only a few paper copies that existed
25    that didn't exist in electronic form.  But it had probably
.                                                               397
 1    the best explanation of what the Commission really meant
 2    with its '86 policy statement.  It's probably the longest
 3    SRM that's been written.
 4              I think it's about 11 or 12 pages long.  We tried
 5    to take the important stuff out of that and put it into the
 6    policy statement.  We also updated the policy statement to
 7    reflect what's been going on in the last ten years of the
 8    PRA implementation plan, Reg Guide 1.174.
 9              In any event, the modifications we made were
10    consistent with the guidance we were given by the
11    Commission.  We reflect the plant-specific use of the safety
12    goals as a definition of how safe is safe enough, as an
13    aspirational goal, something you don't have to meet, but
14    something you aim for, not a regulation.
15              We did maintain the use of the CDF as a subsidiary
16    objective.  That, in fact, came out of the June 15, 1990
17    SRM, as opposed to the original policy statement.
18              We expanded the treatment of uncertainty, taking a
19    portion of what's in Reg Guide 1.174.  We're keeping that at
20    the same high level that the rest of the policy statement
21    was written in.
22              We incorporated the Commission's white paper
23    definition of defense-in-depth; we deleted the reference to
24    a general performance guideline, but did incorporate a new,
25    a subsidiary goal for large early release frequency of ten
.                                                               398
 1    to the minus five per reactor year.
 2              And we incorporated a statement expressing the
 3    Commission's intent to protect the environment, using the
 4    words that they had given us in their SRM.
 5              DR. SEALE:  Joe, you guys are awfully slippery
 6    sometimes, and so I want to see how you're going to slither
 7    with this one.  In Bullet 2 on this thing, you speak of CDF
 8    as a subsidiary objective.
 9              And in Bullet 5, you talk about LERF as a
10    subsidiary goal.  Now, are those the same thing?
11              MR. MURPHY:  Yes, they both should be objectives.
12              DR. SEALE:  Okay, because we've been goal'ing and
13    objective'ing and so forth, back and forth, and I just
14    wanted to make sure we didn't get a higher hierarchy or
15    whatever.
16              MR. MURPHY:  They're the same thing.
17              DR. WALLIS:  How many points do you get for a
18    goal, and how many do you get for an objective?
19              [Laughter.]
20              DR. APOSTOLAKIS:  Where do you define how safe is
21    safe enough?  Is there a sentence somewhere here?
22              MR. MURPHY:  Yes, there should be.  I'd have to go
23    back and look through to find the exact wording, but the
24    attempt was to get the message into the policy statement
25    that the safety goal expresses how safe is safe enough.
.                                                               399
 1              DR. APOSTOLAKIS:  Yes, but there are also some
 2    pretty strong statements that the quantitative health
 3    objectives are not a substitute for existing regulations,
 4    right?
 5              MR. MURPHY:  Right.
 6              DR. APOSTOLAKIS:  You make that very clear?
 7              MR. MURPHY:  Yes.
 8              DR. APOSTOLAKIS:  So meeting these objectives,
 9    you're not safe enough.
10              MR. MURPHY:  No.
11              DR. APOSTOLAKIS:  You still have to go back and
12    read all the regulations; isn't that correct?
13              MR. MURPHY:  No.  Let me go back to the concept of
14    three reasons.  The safety goal expresses a point where
15    you're safe enough; you would not impose additional
16    regulatory requirements.
17              Above that line, you might impose additional
18    regulatory requirements, if they pass the backfit test.  So
19    if they are cost-beneficial as defined by 51.09, you can
20    impose them in that region.  And most of our regulations
21    take us above the safety goal.
22              Above that, you have some area that the Commission
23    has told us we're not ready to define yet, known as adequate
24    protection.  But between adequate protection and the safety
25    goals is this region that it's cost-beneficial in licensing
.                                                               400
 1    actions.
 2              So the safety goals are a goal.
 3              DR. WALLIS:  Yes, that's true.
 4              MR. MURPHY:  That's what you try to reach.  But
 5    they're not a regulation; they're not something you have to
 6    reach; there's not a problem if you don't reach them.
 7              DR. WALLIS:  So what are you --
 8              MR. MURPHY:  They give you a goal as to when to
 9    stop regulating.
10              DR. WALLIS:  That's it, they tell you when to stop
11    regulating?
12              MR. MURPHY:  They tell you how safe --
13              DR. WALLIS:  So they have a negative effect on
14    regulation; they tell you when to stop instead of where to
15    go to?  That's a very strange concept to me.
16              CHAIRMAN POWERS:  It's always been.  The entire
17    idea was to decide when things were safe.
18              DR. WALLIS:  In a game, it's to score a goal, not
19    to stop at the goal line, and that's what I don't
20    understand; I never did.
21              CHAIRMAN POWERS:  The difficulty that you see --
22    and you can see it in the Department of Energy -- is you
23    have an uncapped continuous improvement kind of mentality.
24              And you keep honing away at things, and at
25    tremendous cost, and it incapacitates you.  So you get a
.                                                               401
 1    stagnation of activity.
 2              The Commission really was searching very hard for
 3    something that would say, okay, we have done enough; these
 4    plants are safe enough; they meet the public mores on the
 5    tolerable amount of involuntary risk that they have to
 6    accept from these things.
 7              DR. KRESS:  So if they're above that level, then
 8    they're not safe enough, because that's how safe is safe
 9    enough?
10              CHAIRMAN POWERS:  No, what they're saying is that
11    there is no need to keep looking for ways to make them
12    safer, if they had a probability of failure that was less
13    than the goal.
14              Between the range of adequate protection and the
15    safety goal, then you could look for ways to make them
16    safer, provided you pass the backfit test, the cost-benefit
17    test.
18              Above adequate protection, they were
19    objectionable.
20              DR. KRESS:  Joe used an interesting choice of
21    words when he talked about adequate protection.  He says
22    we're not prepared to define what that is, what adequate
23    protection is, yet.  He put that word, yet, in there.
24              MR. MURPHY:  Yes.  I think we have -- several of
25    us, and, I think, members of the Committee as well, have
.                                                               402
 1    thought long and hard about how to define this.
 2              Of course, what you're defining is reasonable
 3    assurance of adequate protection, not just adequate
 4    protection.
 5              DR. KRESS:  Yes, that's true.
 6              MR. MURPHY:  And the problem is that this is a
 7    multidimensional thing.  It is not measured entirely in
 8    terms of risk.  I have my own personal example of this, and
 9    I'm not sure everybody would buy it, but I'll give you the
10    example.  Suppose we have a plant that has a core damage
11    frequency of ten to the minus six and a LERF of the ten to
12    the minus seven, and they haven't changed the plant.
13              We've reviewed the living devil out of the PRA; we
14    are sure it's right, as sure as we can be.  We believe we
15    have an extremely safe plant.
16              And they arbitrarily take the feeling that, you
17    know, we don't want resident inspectors anymore, so we want
18    to throw them off the property.
19              As soon as that happens, the Commission would lose
20    reasonable assurance of adequate protection, and under the
21    adequate protection standard, we would take -- we would be
22    able to issue an order saying you put our inspectors back on
23    the site or you shut down.
24              DR. KRESS:  Yes, but the other side of that coin
25    is, suppose they are conforming with all the regulations,
.                                                               403
 1    doing a really good job in the inspection, passing all of
 2    the oversight colors, green, but their CDF is ten to the
 3    minus -- five times ten to the minus three?
 4              That was, in our opinion, a place that ought to be
 5    incorporated into the concept of adequate protection, along
 6    with these other things, not make a quantitative CDF and
 7    LERF the measure of adequate protection, but include it.
 8              And that's the other --
 9              MR. MURPHY:  I don't disagree with that.  I think
10    eventually we will have to do some sort of quantitative
11    definition of adequate protection.  I think that if you look
12    at the -- not so much at the SRM, but at the vote sheets
13    associated with the SRM.
14              Several of the Commissioners have a similar view,
15    but they feel that we need to have more experience with what
16    we're doing today and develop standards for PRA and know
17    what we're talking about before we're ready to do that.
18              And I think when we do it, it will not be the
19    answer -- it won't be sufficient, but it will be part of the
20    definition, and I think that will come, but basically, the
21    Commission has told us we're not ready yet, and I think
22    they're right on that.
23              DR. BONACA:  But it seems as if, you know, it
24    still is expanding on the old definition that PRA will be
25    used to fill gaps in the regulation.  And that's really --
.                                                               404
 1    there is no reciprocity there.
 2              The question is, well, there isn't any PRA there
 3    that has been performed on a plant, using margins recognized
 4    by the deterministic regulation, so we don't know what the
 5    PRA for a plant that just meets the regulation would give us
 6    for core damage frequency.
 7              MR. MURPHY:  Well --
 8              DR. BONACA:  For example, if I assume the
 9    containment only capable of meeting its design pressure,
10    which is the requirement, I will get a different result in
11    some of the LERF analysis I've had.  That's just a simple
12    example.
13              MR. MURPHY:  Let me make a wild statement --
14              DR. BONACA:  With the margins that we discussed,
15    so I'm saying that still this is coherent with the old
16    approach that PRA will be used just to fill gaps in the
17    regulation.
18              MR. MURPHY:  Yes.  I'll make a controversial
19    statement about what you just said.  I've been asked the
20    question a number of times, what -- if a plant just met the
21    regulations, how safe would it be?
22              And my answer to that, basically, is that the core
23    damage frequency would be zero.
24              The reason for that, the plant would never go
25    critical.
.                                                               405
 1              DR. KRESS:  Okay.
 2              MR. MURPHY:  Because of all the things that we
 3    don't regulate that are required for a power plant to
 4    operate and so if you just -- so it's difficult to say
 5    exactly where you are.
 6              There are a few of them that were enacted on the
 7    basis of adequate protection.  I believe the last time we
 8    did one on adequate protection was the PTS rule back around
 9    '76, '77.  Most of them have been enacted -- since the
10    backfit rule was initiated, most of them have been enacted
11    under the backfit rule as cost beneficial actions.
12              Some of those perhaps could have been also
13    justified as adequate protection but the backfit test --
14    they could meet the backfit test and that was an easier
15    argument to make.
16              What it means is that the regulations, the various
17    parts take us to different levels of safety and what we are
18    saying is when you are looking for relaxations you really
19    ought to be looking for the cases where the regulations
20    drove you well below the safety goals, because now you
21    require them to be more than safe enough and this is an
22    obvious area where you have room for relaxation, so think of
23    the goal in that light as something you aim for.  You don't
24    don't have to meet --
25              DR. WALLIS:  -- like adequate protection to me,
.                                                               406
 1    but something that you back off from when you go too far.
 2              MR. MURPHY:  Well, adequate protection, as we are
 3    using the term, means that if you had, if you lose
 4    reasonable assurance of adequate protection, the minute Sam
 5    Collins reaches the conclusion that he has lost adequate
 6    assurance of adequate protection there's only one action he
 7    can take.  He shuts the plant down immediately.
 8              He cannot let it operate when he doesn't have
 9    adequate protection, when he doesn't have reasonable
10    assurance of adequate protection, but in between is the area
11    where you would like to make -- it's like of like an ALARA
12    principle.  You would like to make it safer but you can only
13    do so if it is cost beneficial, and we have the various
14    guidelines to tell us how to do that cost benefit
15    analysis --
16              DR. KRESS:  But since adequate protection is
17    multidimensional, like you said.
18              MR. MURPHY:  Yes.
19              DR. KRESS:  I am having trouble defining what the
20    words "in between" mean.  In between five dimensions -- I
21    don't know what "in between" means.
22              The safety goals is one dimension.
23              MR. MURPHY:  I can define it as long as I keep in
24    that dimension that has one -- that acts as risk, but when I
25    get into the other end then it gets harder and that's why
.                                                               407
 1    this question of how we define the three region.  It may be
 2    a three region and five or six dimensional space before we
 3    are done, and that is going to be difficult to not only draw
 4    but also to explain.
 5              DR. KRESS:  Yes, that's the trouble about it.
 6              That is why I wanted something that I could
 7    determine what in between meant.
 8              DR. WALLIS:  If you define it in terms of the
 9    actions the agency performs I can understand it -- if you
10    cross this line we do this, if you cross this line we do
11    that, if you cross this line we don't do anything.
12              I understand that.  It is when you start getting
13    into these words about defining these lines that I get very
14    confused because --
15              DR. APOSTOLAKIS:  See, in reality it is not just a
16    matter of going above or below the line.  That is the
17    problem.
18              There are at least 19 PWR units now that have a
19    CDF greater than the minus 4, right?  Documented, according
20    to their numbers, and the IPE is -- that's why I say at
21    least, because if you go back and do some more stuff maybe
22    others would, and yet they are not shut down.
23              I don't know that they are doing any studies to
24    see whether they can reduce it.  I suspect they are just
25    happily going along and operating having met the regulations
.                                                               408
 1    and the Agency is not doing --
 2              DR. SHACK:  Planning a power uprate.
 3              DR. APOSTOLAKIS:  What?
 4              DR. SHACK:  Planning a power uprate.
 5              [Laughter.]
 6              DR. APOSTOLAKIS:  Some of them are planning a
 7    planning uprate or a license extension.
 8              CHAIRMAN POWERS:  Not so much in PWR range.  Life
 9    extension, yes, but power uprate --
10              DR. APOSTOLAKIS:  So it is not a clear-cut thing. 
11    That is the whole point.  I mean pushing the Agency to
12    declare that if you are above or below you do certain things
13    is no --
14              I have some questions that perhaps were answered
15    20 years ago, but still, when you look at the first bullet,
16    and maybe we can all look at Attachment 1.  This is the
17    goals, the document itself.  It is Tab 15 of your book.
18              Attachment 1, Nuclear Regulatory Commission, 10
19    CFR, Part 50 Summary:
20              "The qualitative safety goals are as follows. 
21    Individual members of the public" -- and so on -- "that
22    individuals bear no significant additional risk to life and
23    health."
24              And then for the societal risks we refer to viable
25    competing technologies for producing electricity.  I wonder
.                                                               409
 1    why we don't do that for individual risk as we did for
 2    societal.
 3              Is that something that was resolved 20 years ago
 4    so we don't need to --
 5              MR. MURPHY:  Well, you know, these words are
 6    directly out of '86 --
 7              DR. APOSTOLAKIS:  I know, I know, but I am just
 8    curious.  Do you remember?
 9              I mean in one case we compare it with all those
10    other risks and in the second case we make it specific to
11    other means for generating electricity and I wonder what is
12    the difference between individual risk and societal risk
13    that requires that.
14              MR. MURPHY:  Well, I think -- you know, I can
15    speculate, George, and I remember those, writing them back
16    in the '86 timeframe.
17              The only good news about the modification, and
18    we'd only been on that for about three years, the first ACRS
19    document that I remember seeing on the policy statement was
20    dated 1980.  It was the David Okrent report, and so the
21    committee debated this for six years when it was coming out
22    the first time.
23              Going back on that speculation, I think the first
24    goal was to just say the impact of a nuclear plant on an
25    individual should be minimal.  It has no significant
.                                                               410
 1    addition to the risk that you have in your normal day to day
 2    life, but then they were worried about competing forms in
 3    energy.
 4              They said -- and also it should not be an addition
 5    to society compared to any other energy source.
 6              DR. APOSTOLAKIS:  Yes, I mean that is what it
 7    says.
 8              MR. MURPHY:  But I think that is the message they
 9    were trying to get across.  It's twofold.
10              If other energy sources were cheaper -- were
11    safer, appreciably safer, this might not apply.
12              DR. WALLIS:  But this doesn't make sense unless
13    you have a measure of that.
14              DR. SHACK:  An individual doesn't look at a cost
15    benefit relation, because the individual that bears the cost
16    may have no relation to the individual that gets the
17    benefit.
18              DR. KRESS:  Society --
19              DR. SHACK:  Society as a whole has a cost benefit.
20              DR. WALLIS:  But look, Joe, if you take this
21    seriously, you have to then go out and measure the risk from
22    these other ways of generating electricity and make a
23    comparison and then make a decision based on that if that is
24    your statement of principle, and I don't see the Agency
25    doing that.  It is just an empty statement unless it leads
.                                                               411
 1    to something, unless it is worked out in terms of some
 2    criteria, numbers, decisions.
 3              MR. MURPHY:  I think in terms of how the
 4    Commission went forward there were studies and they have
 5    both good and bad features -- they were done back in the
 6    '80s -- that attempted to estimate the risk from various
 7    forms of energy.
 8              They have some flaws in them but you could get a
 9    feel, and I think what the Commission was saying when it
10    issued this and now I am speculating, I am going back 14
11    years to try to figure out what they were thinking and
12    actually it's more like 20 years from the time the thing
13    started, and I think if they felt they could not say the
14    second statement that the risk to generating electricity by
15    competing technologies, that the risk was not comparable to
16    that, then you might well have had a different decision made
17    by the Commission as to how far, when and how it regulated
18    nuclear power.
19              DR. APOSTOLAKIS:  I guess if I were to rewrite
20    this, I would try to avoid reference to other --
21              DR. KRESS:  Well, they did rewrite it.  They
22    rewrote it and called it the qualitative health objectives
23    and then it disappears in that, doesn't show up.  They did
24    rewrite it.
25              DR. APOSTOLAKIS:  I think that Graham is correct. 
.                                                               412
 1    This is really an empty statement because the Agency never
 2    attempted --
 3              DR. KRESS:  Yes.  That is why it doesn't show up
 4    when they do the qualitative health objectives, so they
 5    rewrote it.
 6              DR. APOSTOLAKIS:  This is a qualitative -- a
 7    quantitative --
 8              DR. KRESS:  This is a qualitative safety goal.
 9              DR. APOSTOLAKIS:  Right.
10              DR. KRESS:  Then they convert it into qualitative
11    health objectives and it disappears.
12              DR. APOSTOLAKIS:  This is smooth.
13              DR. KRESS:  Yes.
14              DR. APOSTOLAKIS:  Now regarding the first one,
15    individual members of the public -- such that individuals
16    bear no significant additional risk to life and health.  The
17    way that this is interpreted in the quantitative goals
18    later, the one-tenth of one percent and then it becomes a
19    number, I believe we are taking the average risk from
20    accidents over the whole country, right?
21              DR. KRESS:  Yes.
22              DR. APOSTOLAKIS:  Average over time, population --
23              DR. KRESS:  Divided by the total population,
24    right.
25              DR. APOSTOLAKIS:  That is one way of interpreting
.                                                               413
 1    this.
 2              DR. KRESS:  That is the way it is interpreted.
 3              DR. APOSTOLAKIS:  How about another way that says
 4    if there is an earthquake in New York and Indian Point
 5    suffers a core meltdown that is sufficiently strong to do
 6    that, at the same time there will be no New York, so should
 7    I take now the allowed seismic risk for the plant and derive
 8    that from the fact that I have lost New York, because then I
 9    compare that there will be no significant additional risk,
10    or should I still compare it with the average over the
11    country?
12              In other words, there is a strong time element
13    here.
14              MR. MURPHY:  We pondered that question and it is
15    not that simple.
16              [Laughter.]
17              MR. MURPHY:  Nothing ever is.  The real question
18    is if you had the epicenter at Indian Point Manhattan might
19    survive.
20              DR. APOSTOLAKIS:  Yes.
21              MR. MURPHY:  If you had the epicenter at Manhattan
22    Indian Point might survive, so the question is you have got
23    to fold in a whole bunch of distributions and try to figure
24    out what it means.
25              DR. APOSTOLAKIS:  Yes, sure, and you don't do it
.                                                               414
 1    on an epicenter by epicenter basis.  You take the average
 2    horizontal ground acceleration with the plant, then you have
 3    all these maps that tell you what is the acceleration in
 4    other places.
 5              CHAIRMAN POWERS:  This has always been a
 6    puzzlement to me, and let me move away from Indian Point and
 7    Manhattan and move to Browns Ferry and Decatur.
 8              If I have an earthquake such that Browns Ferry has
 9    a meltdown, you get a pretty strong motion then that might
10    well affect Decatur, but the people that are going to be
11    affected by the core meltdown aren't there.  They are
12    someplace else, and I don't think there is an earthquake big
13    enough to affect them.
14              All their risk is going to come from the core
15    meltdown.
16              DR. KRESS:  They are going to be 50 miles away.
17              CHAIRMAN POWERS:  Well, they are going to be
18    within 10 miles --
19              DR. KRESS:  Well, if you are talking about
20    cancers, everybody will be 50 miles --
21              DR. APOSTOLAKIS:  No, we limited that now to 10.
22              CHAIRMAN POWERS:  We're within 10 miles and I mean
23    there are not huge buildings crashing down.  They may lose
24    electrical power and the outhouse may be kicked over, but --
25              DR. KRESS:  Well, that is a disaster.
.                                                               415
 1              [Laughter.]
 2              CHAIRMAN POWERS:  But I mean there's these
 3    horrible, these visions of all kinds of buildings collapsing
 4    and hospitals collapsing that you have from the Los Angeles
 5    earthquakes and things like that -- really doesn't affect
 6    them.
 7              Most of those houses survive and they survive very
 8    large earthquakes, but they are still affected by the core
 9    meltdown and so to say that, well, the core meltdown of big
10    earthquakes is a small perturbation -- not to them, it
11    isn't.
12              DR. APOSTOLAKIS:  Wait, wait, wait.  I mean
13    typically in a PRA to really have a major core meltdown you
14    have to go significantly above the SSC level --
15              CHAIRMAN POWERS:  Sure.
16              DR. APOSTOLAKIS:  -- and whether you limit the
17    consequences to 10 or 15 miles is really an artifact of the
18    regulations.  You have to look at the real risks.
19              If you are now at .45g horizontal acceleration, I
20    don't know that you will have negligible impact on the
21    nearby cities because the uniform building code does not
22    require that residences and office buildings be designed
23    against and SSC.
24              Those will be long gone before the record suffers
25    a core meltdown.
.                                                               416
 1              Now there may be an issue of who is getting what
 2    but the real issue is when we say no significant additional
 3    risk to life and health, are we going to take into account
 4    these time dependent situations and maybe relax a little bit
 5    the requirements of the nuclear plant or the one-tenth of
 6    one percent applies no matter what?  In which case we run
 7    the risk of shutting them down --
 8              MR. MURPHY:  The way that we have done it has been
 9    we applied the one-tenth of one percent.
10              DR. APOSTOLAKIS:  I'm sorry?
11              MR. MURPHY:  The way we have traditionally done it
12    is apply the one-tenth of one percent.  What we have done in
13    things like seismic events, we have not assumed a large part
14    of the population may be killed in the earthquake.
15              We have looked and said a large portion of the
16    population may not be able to evacuate and we have evaluated
17    it in those terms as we have done our PRAs.  Above some g
18    level we have assumed the bridges are down and you are not
19    going to evacuate.
20              DR. APOSTOLAKIS:  So you are always taking the
21    pessimistic view, that all these known nuclear consequences
22    can only hurt the nuclear accident, because now you cannot
23    evacuate.  You are taking a conservative approach in the
24    PRA.
25              The fact that the guys may already be dead is not
.                                                               417
 1    taken into account.
 2              MR. MURPHY:  I don't think we would ever want to
 3    take a regulatory position that says we don't have to worry
 4    about the risk from great seismic events because everybody
 5    is going to be dead.
 6              DR. APOSTOLAKIS:  No.  I understand that, but it
 7    is also a fact that a lot of these guys will be dead.
 8              MR. MURPHY:  Oh, yes.
 9              DR. APOSTOLAKIS:  Okay, so if we advertise PRA as
10    a realistic assessment, I mean in this case we are taking a
11    conservative approach, which may not be conservative enough
12    in fact, if you bring other things into account.
13              The truth of the matter is that we are not taking,
14    applying the one-tenth of one percent to these temporal
15    events that may raise the average risk significantly for
16    that part of the country.
17              DR. KRESS:  My favorite phrase is we are not
18    regulating a certain number of deaths.  We are regulating
19    risk.
20              DR. APOSTOLAKIS:  That's right, and the risk is
21    high for all the surrounding population during those times.
22              MR. MURPHY:  Well, remember, you are looking at
23    the risk, the probability and consequence.
24              The probability is exceedingly low that you are
25    going to have a great earthquake in Decatur.
.                                                               418
 1              DR. APOSTOLAKIS:  Right.
 2              MR. MURPHY:  So that the risk is small and we are
 3    still saying it may well be dominated by other causes.
 4              DR. WALLIS:  Can I ask you a simpler question?
 5              DR. APOSTOLAKIS:  Wait a minute.  Can we resolve
 6    this?
 7              DR. WALLIS:  Well, okay, go ahead.
 8              DR. APOSTOLAKIS:  What is the answer?  That we
 9    don't care?
10              DR. KRESS:  Especially if it is in New York.
11              [Laughter.]
12              DR. APOSTOLAKIS:  That these time-dependent
13    changes in the background risk are really not -- are local
14    effects.  These are really local.  I mean the average risk,
15    individual risk to life and health becomes very high during
16    those conditions, doesn't it?
17              DR. KRESS:  No.
18              DR. APOSTOLAKIS:  Does the probability of
19    occurrence of the earthquake?
20              DR. KRESS:  Yes, and that is what gets us out of
21    this conundrum is the earthquake normally doesn't add --
22    does not dominate the risk normally, so that you can have
23    this --
24              DR. APOSTOLAKIS:  It does not?
25              DR. KRESS:  Normally it doesn't, so you can have
.                                                               419
 1    this inconsistency in terms of --
 2              DR. APOSTOLAKIS:  Which risk does it not dominate? 
 3    The nuclear plant risk it dominates.  That's what the PRAs
 4    say.
 5              DR. KRESS:  Not in my part of the country.
 6              DR. APOSTOLAKIS:  Doesn't the PRA say that in most
 7    plants --
 8              MR. MURPHY:  No, I think what you are missing is
 9    that you have to look at the conditional probabilities that
10    falls to risk.
11              In some cases seismic risk is important, but when
12    I look at it and say what is the chance of my having an
13    accident and being killed?
14              Living in Maryland is probably non-earthquake --
15              DR. WALLIS:  This is my question then.  This
16    second part, the easy part, this prompt fatality risk
17    resulting from other accidents which members of the
18    population are generally exposed, do you have any handle on
19    that and is that site specific?
20              MR. MURPHY:  What we have used to come up with
21    that number is to go look at the accidental death rates in
22    the United States.
23              DR. WALLIS:  That is average for the whole
24    population.
25              MR. MURPHY:  Average for the whole population.
.                                                               420
 1              DR. WALLIS:  So I don't see that you can apply any
 2    other criterion to the other one.
 3              Otherwise, you're comparing two different things. 
 4    If you want to be site-specific, you go to look at the risk
 5    of dying from gunshot wounds in Tennessee or something, and
 6    make it very specific.  That's ridiculous.
 7              DR. APOSTOLAKIS:  Well, now the new statement does
 8    allow the use of safety goals on a plant-specific basis.
 9              MR. MURPHY:  Yes.
10              DR. APOSTOLAKIS:  But it does not change the
11    actual numbers.
12              MR. MURPHY:  That's correct.
13              DR. APOSTOLAKIS:  Which might be relevant to that. 
14    When you do it on a generic basis for the whole country,
15    it's okay to take 1/10th of one percent of the average risk.
16              DR. WALLIS:  But the average risk is quite
17    different, that's what I'm saying.
18              DR. APOSTOLAKIS:  The new thing now is that you
19    can actually use them on a plant-specific basis, yet the
20    comparison is to the average national risk.
21              MR. MURPHY:  What we are really saying is that
22    what we will operate off, basically, is subsidiary
23    objectives.
24              DR. APOSTOLAKIS:  Yes.
25              DR. KRESS:  Absolutely.
.                                                               421
 1              MR. MURPHY:  Core damage frequency and LERF. 
 2    We'll apply those to individual plants.
 3              There would be an argument if we talked that the
 4    risk was appreciably different, the risk to the population
 5    was appreciably different in one county than it is in the
 6    rest of the United States.  There could be an argument that
 7    you'd need to change those numbers to reflect that.
 8              But I don't think that exists, so you need to
 9    figure out --
10              DR. KRESS:  You control that with your siting
11    regulations to a certain extent.  And what saves you from
12    all of this is the fact that .1 percent, even if you apply
13    it on a national basis, when you smear it out to any local,
14    it's still pretty small.
15              I mean, you, of course, get variations locally,
16    but it's still going to be a small fraction, if you do it on
17    a national average; it's still going to be a small number.
18              DR. APOSTOLAKIS:  Anyway, so there is no --
19              DR. KRESS:  If you tried to do this on a local
20    basis you'd have such a mess that I don't see how you could
21    ever keep track of it.
22              DR. WALLIS:  I think you need statements like
23    this, because this is what the public might understand. 
24    What's my risk from nuclear power?  If it's less than .1
25    percent than from all other causes, that's an understandable
.                                                               422
 1    kind of statement.
 2              DR. KRESS:  I agree.
 3              DR. WALLIS:  If you say this plant has LERF of one
 4    point ten to the minus seven or something, they don't
 5    understand what on earth you're talking about.  So you've
 6    got to put in these terms at some level.
 7              DR. BONACA:  One thing that I would like to point
 8    out just for the record, because it helps maybe in thinking
 9    about this, these goals were endorsed by the Commission in
10    1986.  But they were really proposed well before, and, for
11    example, the AIF used them, and we used them for Millstone
12    3, which was a high population density site.
13              And the NRC required a Level III PRA, and the
14    statement was, you shall perform a Level III PRA; we will
15    look at the results, and determine whether or not you need
16    additional improvements.  Site-specific was a specific
17    issue.
18              The industry was very interested in the numeric,
19    quantitative determination of this because they didn't want
20    to have an open -- by which they are still coming in and
21    saying, well, based on some criteria, we are going to impose
22    this and that.
23              And at that time, all those units which were being
24    evaluated for high density population risk, endorsed the EIF
25    goals, which were exactly this, ten to the minus four, ten
.                                                               423
 1    to the minus five.
 2              And they essentially said we will modify the plant
 3    to meet these goals, if we don't meet them.
 4              I'm just mentioning this because it wasn't only
 5    that it came in as a regulatory document.  There was a real
 6    negotiation there, and it was used by the industry to
 7    essentially draw a line and say this is good enough and we
 8    will not want to do more than that.
 9              So, there is some historical perspective on where
10    this line was drawn, and I'm not sure how much the
11    Commission considered that.
12              DR. APOSTOLAKIS:  They were at this for six years,
13    so a lot of things --
14              DR. BONACA:  Of course, but what I'm trying to say
15    is, however, I just mentioned it because historically we had
16    a reasonable way that it got there, and it was almost a
17    negotiated line.
18              DR. APOSTOLAKIS:  Now, moving on, the risk to the
19    population, I mean, that's a mystery to me, how that's
20    calculated.  And one of these days, I'll find out.  I think
21    it's an individual risk, but --
22              DR. KRESS:  They're both individual.
23              DR. APOSTOLAKIS:  It's societal, to anybody else.
24              DR. KRESS:  That's just a misnomer.  They're both
25    in there, in risk.
.                                                               424
 1              DR. APOSTOLAKIS:  Okay.  Can we change the
 2    misnomer then, or is it too late?  Societal risk to
 3    everybody else means number of deaths.
 4              DR. KRESS:  It's not that, because they divide it
 5    by the population again.
 6              DR. APOSTOLAKIS:  Okay, we don't want to create a
 7    problem there.
 8              MR. MURPHY:  Well, we always change, but what I've
 9    tried to do as we went through this.
10              Going back and rereading it after about five
11    years, I found that the 1986 policy statement was actually
12    quite a good document.  And it's amazing how forward-looking
13    it was and how many of the insights from things like Reg
14    Guide 1.174 were already in there.
15              It was really a masterpiece of work, so I tried to
16    change it very little as I went through it.
17              DR. APOSTOLAKIS:  I understand.
18              MR. MURPHY:  Yes, we could get into that argument
19    of societal risk, but it's societal risk averaged over a
20    given distance.  When you do the calculation, I agree, what
21    you come up with is a --
22              DR. APOSTOLAKIS:  Let me give you a few other
23    comments because we have to get moving. Some of this
24    discussion really took place years ago.
25              I still am a little bit confused by the statement
.                                                               425
 1    on page 5, real 5, not the handwritten, on the top.
 2              The third paragraph from the bottom, the last
 3    line, the quantitative health objectives are not a
 4    substitute for existing regulations; we all see that.
 5              Then on page 9, one of the changes is stated to be
 6    that the use of safety goals to define how safe is safe
 7    enough.  And I'm a little bit confused now, and I think
 8    there is a question in here, but, Joe, maybe you can help
 9    me.
10              I mean, if the safety goals -- first of all, where
11    does in this policy statement, we define how safe is safe
12    enough?  You say this is a change.  Does it say anywhere,
13    this is safe enough?  I couldn't find it.
14              MR. MURPHY:  What that statement on page 9, if I
15    can find it --
16              DR. APOSTOLAKIS:  Yes, that's the very last page
17    where it says Summary of Changes.
18              MR. MURPHY:  Yes.
19              DR. APOSTOLAKIS:  Yes.  Anyway, that's the first
20    question.  The second question is, if, indeed, the safety
21    goals define how safe is safe enough, how can we claim that
22    these goals are not a substitute for existing regulations?
23              In other words, if I convince you that I have met
24    the goals, and the goals are safe, defined as safe enough,
25    why aren't they a substitute for existing regulations?
.                                                               426
 1              DR. WALLIS:  Because they went through the whole
 2    thing that goals and safe enough are completely different
 3    things.
 4              DR. APOSTOLAKIS:  Then maybe the statement that a
 5    change is -- the use of safety goals to define how safe is
 6    safe enough is not correct.  That's not a change.
 7              DR. KRESS:  That how safe is safe enough is an
 8    abbreviation, and it should say how safe is safe enough so
 9    that NRC no longer imposes more regulations?  That's the
10    whole statement.
11              DR. WALLIS:  How safe is safe enough to the public
12    means the minimum standard.
13              DR. KRESS:  I know, but that's what that term
14    means.
15              DR. APOSTOLAKIS:  It is intended to mean.
16              DR. WALLIS:  But it's wrong.  How safe is too
17    safe?
18              MR. MURPHY:  To respond to your question, where I
19    tried to get that in was in the second paragraph of the
20    first page.
21              DR. APOSTOLAKIS:  Of the first page.
22              MR. MURPHY:  Under summary.
23              DR. WALLIS:  To the public, adequate protection
24    and how safe is safe enough mean exactly the same thing.
25              DR. KRESS:  That's right.
.                                                               427
 1              DR. APOSTOLAKIS:  Let's let Joe finish his
 2    thought.
 3              MR. MURPHY:  Well, what we added was the safety
 4    goals are not limited goals.
 5              DR. APOSTOLAKIS:  Yes.
 6              MR. MURPHY:  The Commission believes the Staff
 7    should strive for a risk level consistent with the safety
 8    goals in developing or revising regulations.
 9              DR. APOSTOLAKIS:  And I like that.
10              MR. MURPHY:  In developing and applying such new
11    requirements to existing plants, the backfit rule should
12    apply.  So it doesn't say you have to be there; it says you
13    use the backfit rule.
14              DR. APOSTOLAKIS:  If you're above.
15              MR. MURPHY:  If you're above, yes.  You aim for
16    that level, but if you can't get there without violating the
17    backfit rule, you don't go there.
18              DR. APOSTOLAKIS:  You don't go there; that's true.
19              MR. MURPHY:  As long as you still believe you have
20    reasonable assurance of adequate protection.  The minute you
21    think you don't have reasonable assurance of adequate
22    protection, then they don't apply at all.
23              DR. BONACA:  And that's a way to measure whether
24    an outlier is worth to be dealt with.
25              DR. APOSTOLAKIS:  As long as it's below adequate
.                                                               428
 1    protection.
 2              DR. KRESS:  And the problem is that there is no
 3    objective measure of what you mean by reasonable assurance
 4    of adequate protection.
 5              DR. APOSTOLAKIS:  Yes, we all agree to that.  But
 6    there is no question about the thought on page 1.  The
 7    question is whether it is correct that the use of safety
 8    goals to define how safe is safe enough is, in fact, a
 9    change.  I don't think so.
10              MR. MURPHY:  It is a change to this document.  It
11    is not a change in philosophy.
12              DR. APOSTOLAKIS:  And it's consistent with the
13    statement that the goals are not a substitute for existing
14    regulations.
15              MR. MURPHY:  Yes, I think so.
16              DR. SHACK:  Go back to his Resident Inspector
17    example.
18              DR. KRESS:  That's exactly what that means.
19              DR. APOSTOLAKIS:  I guess the problem is the same
20    that Graham has.  You can't call it safe enough.  It's in
21    the sense that you just described, safe enough; that if
22    you're above, you go through the backfit rule to try to go
23    below.
24              DR. KRESS:  There's a lot of little baggage in
25    there.
.                                                               429
 1              DR. APOSTOLAKIS:  But it's not really safe enough.
 2              DR. KRESS:  Yes.
 3              DR. APOSTOLAKIS:  Safe enough means you don't do
 4    anything else.
 5              DR. WALLIS:  I think you really have to avoid the
 6    impression of weasel-wording.
 7              DR. APOSTOLAKIS:  I agree they should not be a
 8    substitute for regulations.  I'm just trying to make the
 9    document self-consistent.
10              DR. WALLIS:  You've got the baggage of all this
11    history of interpreting words in an unusual way; that's the
12    problem.
13              DR. SHACK:  Well, the merciful thing is that
14    nobody says safe enough in here, really.  The statements are
15    really much more carefully phrased.
16              DR. APOSTOLAKIS:  Right, and this last little
17    paragraph on page 9, is this 6, Summary of Changes, where
18    this would be part of the statement?
19              MR. MURPHY:  No.
20              DR. APOSTOLAKIS:  Okay, this is not part of the
21    statement?
22              DR. KRESS:  No, that's just for --
23              DR. APOSTOLAKIS:  What is the statement, by the
24    way?  Where does it start?  The statement ends where it says
25    for further information, contact...; is that the statement?
.                                                               430
 1              MR. MURPHY:  Let me look.  I'm in Attachment 2
 2    instead of Attachment 1.
 3              DR. APOSTOLAKIS:  Attachment 2 has the changes
 4    highlighted.  These are the changes, but it doesn't tell me
 5    where the statement ends.
 6              What is it that's going to appear in the --
 7              MR. MURPHY:  The thing that's labeled, 6, Summary
 8    of Changes Made in Revision 1, right now, I would leave that
 9    in, just to help somebody understand what the revision was,
10    so I would leave that in.
11              But it's an explanatory thing to the changes,
12    rather than part of the policy statement itself.
13              DR. APOSTOLAKIS:  So where does the policy
14    statement end?
15              MR. MURPHY:  The policy statement would end right
16    above it.
17              DR. APOSTOLAKIS:  Everything above it?
18              MR. MURPHY:  Yes.
19              DR. APOSTOLAKIS:  By the way, where does this
20    appear?  Is it part of Part 50?
21              MR. MURPHY:  It winds up in the Statements of
22    Consideration associated with Part 50.  The policy
23    statements are not regulations, but they do appear in the
24    Federal Register, and there's a -- in the looseleaf version
25    of the regulations, there is a volume of Statements of
.                                                               431
 1    Consideration, which includes the statements that go with
 2    the rules when they're being enacted so you can help
 3    interpret what they mean, and it becomes a part of that
 4    document.
 5              DR. APOSTOLAKIS:  But they probably pick up the
 6    volumes that they keep receiving one every day now that says
 7    10 CFR Part 50, that's all.
 8              MR. MURPHY:  If you get the little white book, it
 9    won't be in there.
10              DR. APOSTOLAKIS:  Okay.
11              MR. MURPHY:  But the looseleaf version that they
12    pass out that's volumes that are in --
13              DR. APOSTOLAKIS:  There is one part that they're
14    not getting.
15              DR. KRESS:  George, what they should have done is,
16    instead of calling that How Safe is Safe Enough?  They
17    should have called that Below Regulatory Concerns.
18              [Laughter.]
19              DR. APOSTOLAKIS:  To avoid --
20              DR. KRESS:  Right, that way there would be no
21    controversy at all.
22              DR. APOSTOLAKIS:  On page 7 -- and I'm going to be
23    done soon, by the way -- on page 7, useful surrogate
24    subsidiary objectives, right; are you there --
25              MR. MURPHY:  Yes.
.                                                               432
 1              DR. APOSTOLAKIS:  Right in the middle it says, in
 2    this light, the core damage frequency of less than ten to
 3    the minus four per year reactor pressure appears to be a
 4    very useful subsidiary benchmark.  Is everybody happy with
 5    words like "appears to be"?
 6              Why don't we say "is".  We all know what "is" is.
 7              [Laughter.]
 8              DR. APOSTOLAKIS:  My goodness, in the policy
 9    statement, appears to be -- you see, Joe?
10              DR. KRESS:  Because the large early release
11    frequency is a useful on.
12              DR. APOSTOLAKIS:  Yes, is a useful one.
13              DR. KRESS:  It's a little wishy-washy.
14              MR. MURPHY:  I have no objection to the change you
15    suggest.
16              DR. APOSTOLAKIS:  You mean you can still make
17    changes?
18              MR. MURPHY:  It's always possible to make changes.
19              DR. APOSTOLAKIS:  It depends on how difficult it
20    is.  I think it just struck me as I was reading it, "appears
21    to be," my goodness --
22              DR. KRESS:  We know it is.
23              DR. APOSTOLAKIS:  We know all sorts of things with
24    Regulatory Guide 1.174 and so on, and now we say it appears
25    to be?  I would propose to say "is."
.                                                               433
 1              MR. MURPHY:  I have no objection.
 2              DR. APOSTOLAKIS:  Okay.  Then, I must say that the
 3    discussion on Section 4, Treatment of Uncertainties, is a
 4    little bit aleatory.  I mean, it doesn't have the --
 5              DR. KRESS:  Oblique, you mean?
 6              DR. APOSTOLAKIS:  -- coherence that one would
 7    expect from a document of this importance.  For example, it
 8    says that the Commission has adopted -- on page 8 -- the use
 9    of mean estimates for purposes of implementing the
10    quantitative objectives, blah, blah, blah, blah, blah.
11              And then it goes on and says use of mean estimates
12    does not, however, resolve the need to quantify, to the
13    extent reasonable, and understand those important
14    uncertainties involved in the reactor accident risk
15    predictions, which now, again, makes a separation between
16    mean estimates and the distribution.
17              And the question is, can you get the mean if you
18    don't have the distribution?
19              DR. APOSTOLAKIS:  For an important document, this
20    is not a very well-thought-out phraseology; let's put it
21    that way.
22              MR. MURPHY:  What I tried to do here was blend the
23    discussion on uncertainties, which I actually thought was
24    very good, that was in the earlier document with the
25    information in 1.174.
.                                                               434
 1              So if you look at Attachment 2, you can see where
 2    the changes were made.
 3              DR. APOSTOLAKIS:  I mean --
 4              MR. MURPHY:  That's on page 8 of attachment 2.
 5              DR. APOSTOLAKIS:  This is also done in the ASME
 6    standard, and that was a comment the Committee made, how can
 7    you talk about mean values if you have not developed the
 8    distributions?
 9              It's just that you take a point estimate and you
10    declare it to be a mean value?  And to put it here in a
11    document of this significance, I --
12              MR. MURPHY:  Well, the definition of mean values
13    is from the original policy statements.
14              DR. APOSTOLAKIS:  Original what?
15              MR. MURPHY:  The definition that the safety goals
16    apply to the mean is in the original policy statement.
17              DR. APOSTOLAKIS:  That's correct; I do agree with
18    that.  But then going on and making a distinction that I can
19    do all these mean values, but let's not forget that I also
20    have to worry about the uncertainty, and then immediately
21    you say, my god, he's making a distinction.
22              MR. MURPHY:  I intended to say you're making the
23    distinction.
24              DR. APOSTOLAKIS:  I understand that.
25              MR. MURPHY:  Like here we don't believe you can
.                                                               435
 1    get a mean without a distribution.
 2              DR. APOSTOLAKIS:  Exactly.  So all I'm suggesting
 3    is to wordsmith this.
 4              DR. KRESS:  Try to sharpen up the wording.
 5              DR. APOSTOLAKIS:  Yes, to make sure that -- and
 6    you don't have to do it right now, but are you willing to do
 7    this?
 8              MR. MURPHY:  Of course.  You know, what is the
 9    suggestion?  I'll be glad to do it.
10              DR. APOSTOLAKIS:  Then it goes on to the next
11    paragraph where it talks about sensitivity studies.  I have
12    a problem with the sensitivity studies, but the results of
13    sensitivity of studies -- we've got a typo here -- should be
14    displayed, showing, for example, range variation and so on,
15    depending on the decision needs the probabilistic results
16    should also be reasonably balanced and supported through the
17    use of deterministic arguments, defense-in-depth, and so on.
18              Is this language consistent with that of
19    Regulatory Guide 1.174?
20              MR. MURPHY:  I think it is.  What I did is, I just
21    added the phraseology at the end of the sentence you're
22    referring to about defense-in-depth considerations,
23    maintenance and safety margins, and pro forma management
24    strategies.  That was added.
25              The other words are from the original policy
.                                                               436
 1    statement.
 2              DR. APOSTOLAKIS:  All right.  The last sentence is
 3    a little troublesome.  This defense-in-depth approach is
 4    expected to continue to ensure the protection of public
 5    health and safety.
 6              DR. KRESS:  Well, one of these days, we're going
 7    to have to take up this question of uncertainties and
 8    sensitivities.
 9              DR. APOSTOLAKIS:  Yes.
10              DR. KRESS:  What is always done in sensitivity
11    analysis is vary the parameters one at a time while the
12    range is about what they think is the baseline values to get
13    the importance of each one.
14              But that's supposedly done at what might be called
15    the mean.  But those numbers change all over the map, and if
16    you get off on the curve a little bit, you move up, you get
17    a different sensitivity, and it's not an appropriate way to
18    really conduct a full sensitivity analysis.
19              And one of these days, we ought to have that on
20    our agenda to discuss.
21              DR. APOSTOLAKIS:  Right, right.  Also, I don't
22    think it's really true that the Commission is making its
23    decisions using the mean estimates.  I mean, there is much
24    more at stake here.
25              DR. KRESS:  They use the point estimates coming
.                                                               437
 1    out of the PRA.
 2              DR. APOSTOLAKIS:  You start out with the means,
 3    but then you look at other things, defense-in-depth.  You
 4    look at --
 5              DR. SHACK:  You use mean estimates for comparison
 6    with these objectives.
 7              DR. APOSTOLAKIS:  Yes, but then you have shades of
 8    gray.
 9              I don't know but there are certain -- I was
10    surprised, for example, by the first paragraph under Section
11    5.  The Commission recognizes that a safety goal can provide
12    a useful tool by which the adequacy of regulations regarding
13    changes can be judged.
14              Likewise, the safety goals could be of benefit in
15    the much more difficult task of assessing whether existing
16    plants -- and so on, comply with -- conform adequately with
17    the intent of the safety goal policy.
18              DR. KRESS:  That's confusing; isn't it?
19              MR. MURPHY:  Those are words from the '86
20    statement.
21              DR. APOSTOLAKIS:  Can we either take them out or
22    --
23              MR. MURPHY:  We can take them out.  What I tried
24    to do here was not to do too much violence to the '86
25    statement, except where there was later guidance that would
.                                                               438
 1    change it.
 2              DR. APOSTOLAKIS:  But, Joe, it's not much more of
 3    a difficult task any more.  See, at that time, there was
 4    concern that, you know, maybe we can't use the new animal of
 5    PRA to assess what we've been doing with the plants, but 20
 6    years later, I think everybody is comfortable with the idea.
 7              MR. MURPHY:  I have no problem with taking that
 8    out.
 9              DR. APOSTOLAKIS:  Okay, or rephrase it at least.
10              MR. MURPHY:  I have not problem.
11              DR. APOSTOLAKIS:  Oh, I also saw for the first
12    time, on page 9, a word that I had not seen before.  In the
13    paragraph just above Section 6, second line, "nuclear power
14    plant permitees and licensees."
15              I thought they were all licensees.
16              DR. KRESS:  No, you've got permits in the NMSS
17    area and permits for construction.
18              DR. APOSTOLAKIS:  But this is for nuclear power
19    plants.
20              DR. KRESS:  Well, if you're under construction,
21    you get a permit.
22              DR. APOSTOLAKIS:  So this is a hopeful document. 
23    It hopes.  Okay, permitees then.
24              I don't know, I could argue about little details
25    in the last two Sections, 4 and 5, but I'm not sure it's
.                                                               439
 1    worth it.
 2              So, now, Joe, are you promising to go back and
 3    look at the stuff you said you have no objection to doing? 
 4    Because that will depend on whether -- actually, our
 5    decision on whether to write the letter or not.
 6              MR. MURPHY:  Depending on the constraints on the
 7    Committee, if you have time to write a letter, it would help
 8    me make sure I get the changes you want in there.
 9              But certainly, based on what I heard today, you've
10    given suggestions that I will try to incorporate into the
11    changes.
12              The paper right now is at the various Offices for
13    concurrence, and I'm sure there are going to be some changes
14    coming out as a result of that process.  I've never taken a
15    document to the lawyers without having them change
16    something, and I suspect that will be true, particularly for
17    OGC again.
18              And I'm glad to incorporate the ones that I think
19    I agreed to today.  The letter helps me make sure that we're
20    both talking on the same wavelength.  But that's your
21    choice.
22              DR. APOSTOLAKIS:  Yes.
23              MR. MURPHY:  Certainly, I would prefer a letter,
24    but I understand the constraints that the Committee operates
25    under as well.
.                                                               440
 1              DR. APOSTOLAKIS:  A letter, could it be a
 2    Larkins-gram, a little note referring to the transcript?
 3              DR. KRESS:  A letter that spells out the
 4    specifics.  That's normally documented.
 5              CHAIRMAN POWERS:  If it's going to have any
 6    technical content to it, it has to be a letter.
 7              DR. APOSTOLAKIS:  Well, first of all, are there
 8    any other comments from the members?
 9              DR. SEALE:  It can be a letter, though, and not a
10    report.
11              CHAIRMAN POWERS:  This would be a letter.
12              DR. APOSTOLAKIS:  Well, how do the members feel?
13              DR. KRESS:  I think, personally, that Joe can
14    remember most of the things you said, and take care of it
15    without a letter, but --
16              DR. APOSTOLAKIS:  It's not Joe.  I never
17    understood why this Committee on generic requirements,
18    reviews all these documents after us, and actually makes
19    changes to things we have approved.
20              I never understood that.  It happened before to
21    Regulatory Guides.
22              DR. SEALE:  It just lets you know where you sit.
23              DR. KRESS:  If you can, we'll write a letter,
24    George.
25              DR. APOSTOLAKIS:  I'm not willing to write a
.                                                               441
 1    letter.
 2              DR. SHACK:  Well, that settles it.
 3              DR. KRESS:  I agree with most of your comments.
 4              DR. APOSTOLAKIS:  Yes.  Does anyone feel we have
 5    to write a letter?  This is a very important document, by
 6    the way, that will be forgotten after it's published, but
 7    it's very important.
 8              MR. MARKLEY:  George, within the context of a
 9    Larkins-Gram, you could say something to the effect, you
10    know, that you support the change, and that the Staff agreed
11    to make certain changes.
12              DR. APOSTOLAKIS:  And then refer to the
13    transcript.  Can we do that?
14              MR. MARKLEY:  I wouldn't refer to the transcript. 
15    I'd just say that they agreed to make certain changes;
16    therefore, you have no objection or something like that. 
17    That's about the extent of it.
18              DR. APOSTOLAKIS:  Can we say that?
19              MR. MARKLEY:  Yes.
20              DR. APOSTOLAKIS:  I would much rather do that.
21              DR. SEALE:  Joe, can you defend these rash changes
22    that you've volunteered to make, without a letter?
23              MR. MURPHY:  Oh, I think so.  I'll still use the
24    Committee's name in vain.  It would be helpful, George, if
25    you could give me at least a handwritten note as to what you
.                                                               442
 1    think I agreed to, so we have some sort of --
 2              DR. APOSTOLAKIS:  There's the transcript.
 3              MR. MURPHY:  I mean, I'd have the transcript, but
 4    I'd like to start on it kind of immediately.
 5              DR. APOSTOLAKIS:  Oh.  Can I do that?
 6              DR. KRESS:  Sure.
 7              MR. MARKLEY:  That these are George's suggested
 8    changes and not the Committee's, though; that's the only
 9    distinction.
10              DR. KRESS:  We could even take a vote on whether
11    the Committee agrees to them.
12              DR. APOSTOLAKIS:  You mean I should bring them to
13    --
14              DR. KRESS:  Just give them as your comments.
15              MR. MURPHY:  I will be glad, when the transcript
16    comes out, to go over the transcript and make sure that's
17    what's in there is what I agreed to with the Committee.
18              MR. SIEBER:  The transcript is just comments, and
19    that's no different than if George just pencils them out.
20              DR. KRESS:  That's right, it's the same thing. 
21    George could include some of the ones that he didn't bring
22    up, if he wanted to.
23              DR. APOSTOLAKIS:  Well, what we can do is, I can
24    xerox the marked copy here and give it to you.
25              DR. KRESS:  Why don't you do that.
.                                                               443
 1              DR. APOSTOLAKIS:  And then you and I can -- can we
 2    do that?
 3              DR. KRESS:  Yes.
 4              DR. APOSTOLAKIS:  Okay.
 5              MR. MURPHY:  Just let me get an earlier start and
 6    get it to the Program Offices for their review earlier.
 7              DR. APOSTOLAKIS:  Are there any other comments,
 8    suggestions by the members?
 9              DR. KRESS:  Yes.  I think we lost the battle.
10              DR. APOSTOLAKIS:  Yes.
11              DR. KRESS:  And I hope we don't lose the war,
12    though.  I think we ought to stick to our guns and leave the
13    Safety Goal Policy Statement alone.  It's a sacred document;
14    it's cast in iron now.
15              And we can talk about the needs for a policy
16    statement on risk-informed regulation or something like that
17    that still incorporates these concepts we've mentioned.
18              DR. APOSTOLAKIS:  The three-Region approach?
19              DR. KRESS:  Yes.
20              DR. APOSTOLAKIS:  Well, they have not rejected the
21    concept.  They just say it's --
22              DR. KRESS:  It's just a big deal to change the
23    Safety Goal Policy Statement, and it's really not needed in
24    there.  It's not the place -- it doesn't belong in there. 
25    I've changed my mind on that.  It really belongs in a policy
.                                                               444
 1    statement on risk-informing regulations or something, which
 2    we don't really have a policy statement on, by the way.
 3              DR. APOSTOLAKIS:  We do have a de facto --
 4              DR. SEALE:  So you really changed the war, you've
 5    just defined or redefined your terms of peace.
 6              DR. APOSTOLAKIS:  I think as your CDF goes to, you
 7    know, values of two, three, ten to the minus three and
 8    above, you are now in the region of inadequate protection.
 9              DR. SEALE:  Yes.
10              DR. APOSTOLAKIS:  And I think the Staff acts that
11    way, but I guess --
12              DR. KRESS:  But it's de facto.
13              DR. APOSTOLAKIS:  It's a de facto thing.  When you
14    look at the history of the Agency, anytime we find something
15    that's above those numbers, immediately there is action. 
16    What's going on?
17              Anyway, is there anything else?
18              DR. LEITCH:  Just a small typo, page 5, line 5, in
19    the smooth version.  It says file industry, and I think it's
20    just meant to be --
21              MR. MURPHY:  Where was that?
22              DR. LEITCH:  The smooth version, page 5, line 5. 
23    I'll point it out to you.
24              MR. MURPHY:  Oh, yes, thank you.
25              DR. APOSTOLAKIS:  Okay, so this session is 21
.                                                               445
 1    minutes ahead of schedule.  Back to you, Mr. Chairman.
 2              CHAIRMAN POWERS:  Thank you very much, and we will
 3    recess until 10:15, and we can dispense with the
 4    transcription at this point.
 5              [Whereupon, at 9:40 a.m., the meeting was
 6    recessed, to reconvene in an unrecorded session.]
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

Page Last Reviewed/Updated Wednesday, February 12, 2014