ML20235R509
| ML20235R509 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 10/02/1987 |
| From: | Fierce A MASSACHUSETTS, COMMONWEALTH OF |
| To: | NRC ATOMIC SAFETY & LICENSING APPEAL PANEL (ASLAP) |
| Shared Package | |
| ML20235R461 | List: |
| References | |
| OL, NUDOCS 8710080081 | |
| Download: ML20235R509 (13) | |
Text
- _ _ _ _ _ _ - - _
t f
r i.
(
f 5C 1
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 17 DCT -5 P3 :41 BEFORE THE
~
ATOMIC SAFETY AND LICENSING APPEAL BOARDggg g.
E S CFL 7t s a m j f' BRAS.y
)
In the Matter of Docket Nos. 50-443 OL) l PUBLIC SERVICE COMPANY OF
)
50-444 OL NEW HAMPSHIRE, et al.,
)
(On-site Emergency (Seabrook Station 1, Units 1 and 2) )
Planning Issues)
)
q ATTORNEY GENERAL JAMES M. SHANNON'S MEMORANDUM IN RESPONSE TO APPEAL BOARD ORDER OF SEPTEMBER 17, 1987 Attorney General James M.
Shannon sets forth in this memorandum, in response to this Appeal Board's order of September 17, 1987, his detailed position with respect to
" matters on which agreement has not been achieved" with respect to the testing of emergency sirens this winter in New Hampshire.
1.
Scope Of The Tests.
As we understana this area of disagreement, SAPL maintains that the Appeal Boaro Order of July 30, 1987, calls for a test of all the sirens in the EPZ.
The applicants and the NRC staff interpret paragraph 2 of this Order to apply only to the East Kingston sirens.
Obviously, the Appeal Board knows best what it intenced by its Order of July 30, 1987.
But both common sense and the I
requirements of NUREG-0654 suggest strongly that a test of all j
i sirens in the EPZ should occur.
l 1
l 87100B0001 071002 PDR ADDCK 05000443 G
PDR l
i.
'i 4'
'As th'is-Board recognized in its July 30 Order at pp. 4-5, applicants have planned to take'" corrective action" for all sirens and activation antennae within the Seabrook alert and notification system.
Therefore, a-test of a couple of sirens j
in East Kingston alone hardly provides " reasonable assurance" l
that.the " corrective action" was accomplished system-wide in a 1
manner which' uniformly corrects the gross' malfunctions exhibited during the tests last winter in East Kingston.
Keeping in mind that the siren problems observed (and heard) in I
EastLKingston last winter were (a) so serious as to cause i
doubts about the ability of the sirens to notify the public at all after severe winter weather, and (b) were diagnosed as having been caused at least in part by generic features in the equipment, it makes little sense to restrict tecting this coming winter'to the sirens in East Kingston alone.
The public needs'to know, and the NRC ought to want to know as well, these things:
that the suggested equipment modifications have been attempted in sirens throughout the EPZ, that these attempted l
l modifications have been done correctly on each siren, and that the corrections " work,"
i.e.,
they permit each siren to function effectively during and after severe winter weather.
I Through an inspection and visual obs7rvation alone, the NRC staff may be able to confirm the first of these -- that the modifications to each siren in the EPZ have been attempted (the new pieces are there).
But only through an EPZ-wide test after - - - _ _ _ _ _ - _ _ _.
4 a bout"of typical New England' winter weather will the public, I
the NRC staff, and even the applicants know that the modifications were done correctly on each siren.and that these modifications " work" to solve the prior problems.
Furthermore,'the Attorney General contends that apart from this Board's July 30 Order,'the applicants are required by NUREG-0654 to conduct a " complete cycle" siren test "at least annually and as required for formal exercises."
NUREG-0654/
FEMA-REP-1, Revision 1, at p.-3-12.
This requirement is repeated in FEMA-REP-10,. " Guide f or the Evaluation of Alert and Notification Systems for Nuclear Power Plants," as follows:
" Full-scale testing should be conducted at least annually and as required.for formal exercises."
FEMA-REP-10 at E-9.
Given that the applicants must soon conduct a " full-scale" siren test anyway (not having' conducted one in the past year),
and given that a full-scale, adverse weather wintertime test (which has never occurred at Seabrook) makes such emminent sense for a siren system that is situated at this northern location in New England, the Attorney General urges the Board to seize this opportunity and direct the applicants to conduct an EPZ-wide siren system test this winter.
Finally, to the extent that the applicants and the NRC staff now seek to construe the late-filed SAPL Contention (SAPL
- 38) as one which originally challenged only the adequacy of the East Kingston sirens, we believe that is an unfairly crabbed )
i
a
- reading of that contention.
A fair-minded reading of SAPL 38 clearly reveals that SAPL was contending that-the East Kingston tests revealed generic wintertime problems to the sirens throughout'the EPZ.. Applicants. response -- to make generic corrections throughout the EPZ -- confirms this view.
As a matter of'both lega1' requirement and common sense, the tests.this winter must be system-wide.
I 2.
Timing of The Test.
The Attorney General believes that the applicants are being i
unreasonable when they claim that they cannot ensure that the test will occur during or very soon after an episode of severe winter weather.
Obviously, they would prefer to have a test under less challenging conditions.
But it la not cold weather alone that was the culprit last winter; it was driven snow, sleet,'and freezing rain.
hithout the presence of these (or at least some of these) reasonably common seacoast conditions, no none will know whether the problems exhibited last year have been corrected.
We suggest that the applicants pick a series of dates, say. ten or so (on consecutive Saturdays if they wish), and that the test be planned for the first of these dates and then be continued from date-to-date until a date is reached which is during or just after a severe winter storm.
l 3.
Tests In The Public Address Mode, i
l The test procedure which has been proposed includes tests of the sirens in each mode, including the public address mode, 1
I I i
and we assert that this is appropriate.
First, NUREG-0654, as noted above, requires a " complete cycle" test at least annually.
NUREG-0654/ FEMA-REP-1, Revision 1, at 3-12.
The l
test proposed meets this requirement because it calls for a complete cycle of tests in all of the sirens' modes.
- Second, we unaerstand that the applicants and/or governmental officials may, on an ad hoc basis during an emergency, wish to send special messages to the public using the public address mode.
Although this is not a " planned" use of the sirens (except perhaps in the beach areas), and there is no dispute that public " notification" of an emergency will occur using the
" alert" mode, using the sirens in the public adaress mode is an optional ad hoc use which has some significant public safety value.
The sirens have this " voice mode" capability, but public officials need to know whether this ad hoc capability exists during and after severe winter weather.
Appropriate wintertime tests of the public address mode will allow public officials throughout the EPZ to know whether the voice mode is worth using at all as an ad hoc option during or after severe weather.
4.
The Test Procedures Themselves.
In our letter to the Board dated September 22, 1987, we describea three (3) significant problens with the test procedures themselves:
the expectant observers problem, the observer location problem, and the voice mode intelligibility _ _ _ - - _ - - _ _ -
A problem.
These problems were' pointed-out.to us by Gregory C.
Tocci of the acoustics consulting firm Cavanaugh Tocci Associates, Inc., after he reviewed the proposed test-procedure.
Rather than repeat our description of these three problems, we have attached our letter 4 of September 22, 1987, as Attachment "A" to this memorandum.
After sending this letter to the Board we asked Cavanaugh Tocci Associates to' review it and to offer us any additional thoughts or details they had regarding the problems identified.
Gregory Tocci of Cavanaugh Tocci Associates has now written to us expressing additional details regarding these problems.
His letter'is attached hereto as Attachment "B."
With respect to the " expectant observer problem," Mr. Tocci points out that we may have been overzealous in describing the subjective assessments of expectant observers as " worthless."
Instead, he notes that "it would be more to the point to indicate that expectant observer response is an imprecise measure of unexpectant public response."
In other respects, he agrees with the statement of this problem presented in our letter of September 22, 1987.
4 With respect to the " observer location problem," Mr. Tocci supplements our letter by suggesting that the EPZ test procedure should include a map indicating observer locations relative to horn coverage patterns in order to ascertain whether or not observations were conducted at residential or occupied locations predicted to have the least audibility.
]
4 Regarding the third problem identified in our letter --
"the voice mode intelligibility problem" -- we noted that acoustics experts utilize a variety of test procedures to test
" voice intelligibility," none of which are included in the proposed test procedure, Acoustics experts assess voice intelligibility using.a measurement described as the
" articulation index."
In his attached comments, Mr. Tocci describes three methods for estimating articulation index at the observer locations selected.
Any test of the sirens in the public address mode should utilize one of these methods of measuring the articulation index.
CONCLUSION Because of the siren problems exhibited last winter in tests in East Kingston, New Hampshire, serious questions arose regarding whether all the sirens in the EPZ needed modification to ensure that they would function appropriately under severe winter weather conditions.
Indeed, the applicants indicated that system-wide modifications would be made.
The late-filed SAPL contention which addressed these concerns had substantive merit.
The objections raised were procedural only and concerned SAPL's ability to " assist in developing a sound record," the extent to which admission of the contention would broaden the issue or delay the proceeding, whether significant safety issues were being addressed, and whether a different result would be or would have been likely.
The course of events to date demonstrate that these 1 l
_.___._.____________________a
.c e
, procedural concerns have dissipated substantially and that this
' late-filed contention should now be admitted and remanded: to the Licensing Board for further proceedings.
Interveners
experts from Cavanaugh Tocci Associates stand ready to assist in developing a sound record.
The recent issuance of ALAB-875, remanding other on-site issues to the Licensing Board, means that the admission of:this contentioh will not necessarily delay the proceedings to any significant degree.
Significant safety issues are at stake here,'and the applicants' proposed unsatisfactory test procedure coupled with the NRC statr's narrow-minded-lackadaisical approach to its oversight responsibilities regarding this issue suggests that its resolution cannot be~ entrusted to the applicants and NRC staff alone.
The Appeals Board should admit this contention and remand it to the Licensing Board for appropriate proceedings.
JAMES M.-SHANNON ATTORNEY GENERAL COMMONWEALTH OF MASSACHUSETTS By:
Allan R.
Fierce Assistant Attorney General Nuclear Safety Unit Department of the Attorney General One Ashburton Place Boston, MA 02108 (617) 727-2220 DATED:
October 2, 1987 ____ __ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
b THE COMMONWEALTH OF MASSACHUSETTS U,i DEPARTMENT OF THE ATTORNEY GENERAL l
JOHN W. McCoRMACK STATE OFFICE BUILDING oNE ASHBL,RToN PLACE, BOSTON 02108-1698 JAMES M, SHANNoN ATTORNEY GENERAL September 22, 1987 Alan S. Rosenthal, Esq., Chairman Gary J.
Edles, Esq.
Atomic Safety and Licensing Atomic Safety and Licensing Appeal Board Appeal Board U.S. Nuclear Regulatory U.S. Nuclear Regulatory Commission Commission' Washington, DC 20555 Washington, DC 20555 Howard A.
dilbur Atomic Safety and Licensing Appeal Board U.S. Nuclear Regulatory Commission Nashington, DC 20555 In the Matter of PUBLIC SERVICE COMPANY OF NEW HAMPSHIRE, ET AL.
(Seabrook Station, Units 1 and 2)
Docket Nos. 50-443, 50-444 off-Site Emergency Planning
Dear Members of the Appeal Board:
In his letter to you dated September 11, 1987, which we received September 16, 1987, Edwin Reis, the NRC's Deputy Assistant General Counsel, listed certain " areas of disagreement" regarding the proposed procedure for a test of tne Seabrook emergency notification sirens that you indicated should be conducted in your order of July 30, 1987.
At the end of the letter, Mr. Reis suggests that the Appeal Board be informed within seven days if the parties views have been misenaracterized or inadequately stated.
We are writing to you now to inform you that Mr. Reis' letter did not contain any reference whatsoever to the " areas of disagreement" we have with the proposed test procedure.
Perhaps this is because at the very time --late August-- when Mr. Reis was obtaining the views of the parties on the test procedure, Don Bronstein, the attorney who was handling this matter for our office, left his job here for a position in private practice.
In any event, we do have some strong objections to the adequacy of the test procedure, and these objections are listed briefly below.
I O
//
9' Pursuant to your recent Order of September 17, 1987, in which you direct each interested party to file a memorandum, on or before October-5, 1987, detailing its positions on the matters on which agreement has not been achieved, we intend to file a memorandum which describes these objections in greater detail.
A summary of each of our objections to the test procedure
.is as follows:
A.
The Expectant Observers Problem The proposed test procedure utilized an unspecified number of " field" and " pole" observers who are to be stationed at various distances from sirens to listen carefully at exactly specified times (requiring that they have watches) for the sirens to be activated in the various modes.
These " expectant observers," because they will be listening carefully through the ever-percent background noise for a specific kind of sound at a specified time, could easily hear faint or muffled siren noises.that the unexpectant public would not notice.
The Observer Data Sheets for " pole observer" and " field observer,"
Attachments 4 and 5 to the proposed test procedure, ask l
observers to rate the noises heard simply as either " sat." or "unsat." without utilizing any objective criteria whatsoever.
Such a test says nothing about whether the sirens are capable of providing emergency notification to the unexpectant L
populace.
Because the siren system was designed to meet certain sound coverage (deciole level) criteria which enable its noise emissions to be heard above the background noise throughout the EPZ, any test of sirens must employ sound testing equipment, not expectant observers, to objectively assess whether the sirens emit sounds which meet the system's l
objective design criteria.
The subjective assessments of L
expectant observers are worthless measures of a siren system's adequacy.
B.
The Observer Location Problem The proposed test procedure calls for three (3) observers at each pole; one at the pole, one at 1000 ft.,
and one at 5,000 ft.
The procedure notes, however, that "(t]he 1000 ft.
and 5000 ft. locations are approximate and are intended as t
guidelines.
It may not be practical (or even possible) to locate field observers at these distances."
If one seeks to have a meaningful siren audibility test in wintertime conditions, when siren audibility is possibly muffled by snow l
and ice, these test procedures provide inadequate coverage of the geographic area each siren is designed to cover.
The sound contour map for the siren system, provided in Volume 1 of the New Hampshire Radiological Emergency Response Plan, indicates that because of hills and other geographic features the sound level contours around each siren are far from uniform.
In many l
cases the 70dBC coverage contour line is not a circ 1'e but is a circle with a " wedge" or a " notch" in it, typically due to the effects of a hillside.
It is in these " gap" areas where snow-muffled sirens would least likely be heard; so observers should be stationed in these " gap" areas as the top priority.
The distance an observer is from a siren in and of itself, is a poor criteria to use for determining where observers should be placed, because it is the sound coverage contour lines, not j
distance, which describe areas most likely to receive an inaudible or unintelligible signal.
C.
The Voice Mode Intelligibility Problem Since there have been serious problems created by snow and ice accumulation when sirens were tested last winter, any fair test this winter should test the " intelligibility" of the messages emitted by the sirens when they are used in the voice mode.
Acoustics experts utilize a variety of test procedures to test voice intelligibility, but no voice mode intelligibility test whatsoever are included in the proposed test procedure.
As with the " expectant observer" problem described above, the field observers who have been carefully instructed to listen for a particular message is more likely than an unexpectant resident not only to hear that message but to understand it as well, even if garbled or muffled due to snow and ice accumulation in the sirens.
Objective voice intelligibility tests should be incorporated into the test procedure to eliminate the expectant observer problem when sirens are tested in the voice mode.
Very truly yours, Allan R. Pierce Assistant Attorney General Nuclear Safety Unit Department of the Attorney General One Ashburton Place Boston, MA 02108 (617) 727-1090 ARF /ds cc:
Service List i
E'
[
CAVANAUGH TOCCI
)
ASSOCIATES, INCORPORATED 327 F BOSTON POST RO A D. SU DBURY M ASS ACH USETTS 01776 TELEPHONE i6171443 7671 PM\\CIP tlJ M \\lt4 CON 3lLTAVT A5MWi47tts CO.ysi;1nT:
l WILLIAM.i CAVAN AUGH fASA A ANTHONY HOOVER klCH ARIT G CANN PT GREGORY C 1DCCI fl. PRI:51DEYT DWID H hAYI.
TIMOrifY l FOULKE5 THOM A.s N STEIN
-I bklON G AONING atmtwr t7Dk DONN A L RArUS 1
September 25, 1987 RECEIVED Attorney Allan R. Fierce Assistant Attorney General 3
Nuclear Safety Unit
]
Department of the Attorney General 1 Ashburton Place O C T 0 1 W 7~ '
Boston, MA 02108-1698
{
v 6
Department 0120ffdf
Dear Attorney Fierce,
j i
As requested, I have reviewed your letter dated September 22, 1987 pertaining to Seabrook Off-Site Emergency Planning Test Procedures.
My comments on the three objections you have sited are as follows:
A.
The Expectant Observer Problem In the last sentence of your comments, you indicate that:
" Subjective assessments of expectant observers are worthless measures of a siren system's adequacy."
I think it would be more to the point to indicate that expectant observer response is an imprecise measure of unexpectant public response.
I J
B.
The. Observer Location Problem I
suggest that the EPZ test procedure should include a
map indicating observer locations relative to horn coverage patterns in order to ascertain whether or not observations were conducted at residential or occupied locations predicted to have the least audibility.
I believe that this would be a practical way for Seabrook to demonstrate that it is
.providing a
fair test of the siren system.
The test procedure
- should, but doesn't, require observations from locations previously determined to have questionable siren
- coverage, n
n b D MEMBER N A TIO N A L COUNCIL OF A CO U STIC A L CON 5ULTANT5 O
e Attorney Allan R.
Fierce Page 2 C.
Ihg Voice Hodg Intelligibility Problem i
There are three methods for estimating articulation index at observer locations.
The first involves estimations based on horn distance, horn directivity,. topographical shielding, and atmospheric effects.
The second method involves reading a
phonically balanced random word list over the page system and observers recording words at various locations around each siren.
(This is the most direct measure of articulation index as it was originally developed.)
The third method is to use a Bruel & Kjaer RASTI system which involves. playing an 8-second random noise signal over the page system and using a RASTI processor to measure the signal at locations around each horn.
The RASTI processor provides a direct reading of articulation index at these locations.
Although the first method might possibly be the least expensive, use of the RASTI system would be more accurate.
Use of this
- system, however, would need to be undertaken by Seabrook.
There are many technical reasons why a horn which achieves acceptable levels at observer locations under the siren mode would not achieve acceptable articulation under the page mode.
- Hence, I
believe that articulation studies are a
reasonable requirement if articulation of a message is an important feature of emergency warning.
If I can provide you with any further information, please do not hesitate contacting.me.
Sincerely yours, CAV GH TOCCI ASSOCIATES, INC.
v60)
Gregory C.
Tocci k
CCI I
I GCT/dr/7233 I
cas
\\