ML20235S370
| ML20235S370 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 03/12/1986 |
| From: | Jamison J Battelle Memorial Institute, PACIFIC NORTHWEST NATION |
| To: | Perotti D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| Shared Package | |
| ML20235S055 | List: |
| References | |
| FOIA-87-346 NUDOCS 8707210639 | |
| Download: ML20235S370 (21) | |
Text
-
- a & za OBallelle Paotic Northwest Laboratories P O fios W1 Ric nl na. N runcton L i A. 99352 Tenennone.Sev.
Teles 15 2874 March 12, 1986
~
Mr. Donald J. Perotti Emergency Preparedness Branch Division of Emergency Preparedness and Engineering Response U.S. Nuclear Regulatory Connission Washington, D.C.
20555
Dear Don:
SEABROOK EMERGENCY ACTION LEVELS In accordance with your request of February 4,1986, Gordon Bryan of Comex has reviewed the subject emergency action levels.
The results of that review are provided in the enclosures to this letter.
If you have any questions, please call me on FTS (509) 375-3782.
Best regards, 1
J D. Jamison Technical Leader Emergency Preparedness Group Health Physics Technology Section Health Physics Department JDJ:sjs Enclosure cc: DB Matthews, w/ene. M 8707210639 870717 A)
)
/
CONNOR87-346 PDR
- _- _ _j
9 SEABROOK EMERGENCY ACTION LEVELS ENCLOSURE 1.
BACKGROUND,
SUMMARY
, AND RECOMMENDATIONS BACKGROUND An overriding caveat of the WOG CSFST symptomatic process is that it is intended to apply when the reactor is shut down or has been commanded to shut down and failed to do so. To illustrate, consider the priority suberiticality(S) path. The first decision block asks the question, reactor neutron flux <5%, yes or no? If no, the S path is red, cause for declaration of Site Area Emergency (SAE). Note that only that caveat prevents declaration of SAE during normal power operations when flux is typically >95%.
1 i
The symptomatic classification process is an excellent one. However, one criticism of the process concerns classification timeliness.
If classi-l fication is made from an event-oriented EAL, it is almost intnediate.
If the same decision is made through the symptomatic process, the malfunction must proceed from cause to effect before classification is determined.
Application of the symptomatic process in lieu of a specific HUREG-0654 event-oriented EAL is a matter of judgement.
In those instances where Sea-brook relies upon the symptomatic process to accomplish one of the event.
related items from the NUREG, and the evaluator disagrees with that treat-ment, disagreement has been stated as an opinion.
1
L.
v
SUMMARY
p.
.The Seabrook plan EAL-scheme meets the intent of 10CFR50 App. E'and' NUREG-0654. However, significant weaknesses exist in the present Seabrook EALs'.
The evaluation found significant inconsistency between the plan and the procedure, particularly in the classification flow charts where 1005 correlation should be expected. There were many EAL deficiencies in both
~
the plan and the procedures, although not always the same ones. There were also some examples of correct EALs at the improper level'(e.g., Alert rather than SAE). The evaluation concluded that Seabrook was on the proper track but.had not as' yet achieved' a satisfactory program.
RECOMMENDATIONS The licensee should be required to:
1.,
Reconcile EAL differences between the plan and the procedures.
2.
Consider the criticisms of this evaluation for application to the Sea-brook EAL program. Particular attention should be paid to application l
of event versus symptomatic-oriented EALs at the GE level for those accidents which have a significant lag time between the cause and the symptomatic process decision to classify as a GE (e.g., NUREG-0654,.
App.1.GE5).
l 1
2 i -
e
r.
3.
Conduct an internal side-by-side audit of the Seabrook EAls against guidance.from Appendix 1 to NUREG-0654. Where differences are noted and retained, the. decision to deviate should be based on an engineering evaluation.and documented internally.
Upon completion of the above, the Seabrook EALs will support licensing. ~
1 6
O e
3
ENCLOSURE 2.
COMPARISON OF THE PLAN AND PROCEDURE " EMERGENCY CLASSIFICATION SYSTEM FLOW CHARTS" EPIPs are intended to implement the emergency plan. To evaluate plan implementation, a side-by-side comparison was made between the Emergency Classification Flow Chart shown in Figure 5.6 of the FSAR (part o'f the emer-gency plan) and the equivalent (unnumbered) page in the classification EPIP, ER-1.1A(Rev.3). The following deficiencies were noted:
Miscellaneous emergency condition 7A of the plan is correctly identified as a Notice of Unusual Event (NOUE).
In the EPIP there is no steam generator tube leak condition which leads to a NOVE.
(NUREG-0554, App. 1, NOUE Initiating Condition [IC] 5).
The plan lists 8a, 8b, and 8d, omitting 8c. The EPIP correctly lists Ba-8c.
Nomenclature differs between the plan and the EPIP throughout condi-tion 8 (e.g., inclusion of the words " Fuel cladding failure / melt" in item 8 of the plan; omission of the same words in item 8 of the EPIPs).
. The nomenclature and classifications differ between the plan and the.
EPIP for item 15 (e.g., prinary or secondary coolant versus emergency 1
coolant recirculation; only one declaration results from the plan, a General Emergency (GE), whereas the EPIP lists conditions which can 1
result in an Alert or a GE).
4
Plan treatment of item 16 fire, is inconsistent with the EPIP and the NUREG. Nomenclature differences between the plan and EPIP versions of 16b-16d should be resolved (e.g., contained and controlled, degraded versus loss, etc., see NUREG-0654, App.1, Alert IC' 13; SAE IC 11).
Both flow charts label EAL 17 as " control room evacuation other than fi re'. " The appropriate sections of NUREG-0654 App. I merely discuss control room evacuation, regardless of reason. I recomend that the phrase "other than fire" be deleted.
(Thatnomenclaturedoesnot appear in the' body of the EPIP, e.g.,17c).
Item 20 of the plan cannot lead to an emergency classification. There are no subordinate entries under the accident classification headings.
~
5
I 1,
ENCLOSURE 3.
SEABROOK EMERGENCY PLAN EAL EVALUATION GENERAL-FINDINGS Figure' 5.3, Critical Safety. Function (CSF) for the Heat Sink.. is.
improperly labeled. Number 4.
It should be Number H.
The X innd Y axes of the pressure temperature graph contained in Figure 5.4 are unscaled and therefore fail to provide parameter values. (NUREG-0654IID1).
OPINION: The drawing should. remain as is. Absence of the scales on this diagram is not critical since the plan is not intended to be a working document during the classification phase of an accident.
Inclusion 'of these scales would unnecessarily complicate the plan drawing. It is in the E0Ps where the scales become mandatory for operator interpretation of the' graphs..
The plan skips miscellaneous emergency condition 8c, a typo.
e Some information remains to be provided (e.g., Figure 5.3, the SG>*1).
The reviewer was unable to detemine if Seabrook complied with the following:
10CFR50 E IV B re: state and local government agreement with Seabrook EALs.
6
[
L
]
NUREG-0654 II D2, re: inclusion of all accidents postulated in l
the FSAR.
1 1
NUREG-0654 11 D3, re: state and local classification and EAL scheme are consistent with that of Seabrook Station.
HUREG-0654 II E1, re: compatible procedures for notification, consistent with classification and EAL.
1 NUREG-0654 II E3, re: content of initial messages.
NUREG-0654 App. 1, Pg. 1-8, re:
15-min updates to offsite authorities.
The words 'other than fire" should be deleted from chart EAL 17.
PLAN Eats VERSUS NUREG-0654, APP. 1, GUICANCE NOTES:
In the remainder of this enclosure paragraph numbers conform to those of Appendix 1 of NUREG-0654. Where numbers are omitted, the reviewer found the plan to confonn to the NUREG.
If an asterisk (*) is present, provided the Seabrook technical speci-fications require a shutdown to the cold condition under the condi-tions provided, that comment may be deleted.
7 i
l*
1 NOUE 1, 3c, 5, 6, 9, and 17. Not covered by the Seabrook EALs.
3b. Classified as an Alert (not NOVE) at Seabrook, but only if based upon letdown monitor readings. The case of equivalent values from an RCS cample (as distinct from the letdown monitor) is not covered.
4, 11, & 12. Not covered.
8.
Not covered.
It appears that the intent was to cover this requirement in Seabrook 20, however, no classification can result from item 20.
l 13, 14, & 15. Covered generically as Seabrook item 18a. However, the detailed definition stated in the NUREG initiating conditions is missing. Seabrook simply discusses " hazards", a term that is undefined.
OPINION: Definition was left to the EPIPs. Unacceptable because of the NUREG-0654 II D 1 requirement for the plan to specify parameters.
This evaluator would prefer to leave it as is and have detailed para--
meters appear in the working procedure, EPIP ER-1,1.
8
e ALERT lb.
Incomplete. Provided that the Seabrook item Ba value of gross activity of 600 pCi/cc at the letdown monitor is the engineering equivalent of the NUREG requirement to detect 300 pCi/cc of equivalent I-131, part of the requirement is met. However, the Seabrook EAL is solely contingent on letdown monitor and neglects the possibility of a 300 pCi/cc input from an RCS sample, as is provided for in the more general definition of the MUREG.
Ic, 4, 6, 19. Not covered.
5.
Incomplete.
Provided that the numbers work out, it is possible that a leak of 50 GPM outside containment would invoke Seabrook EAL 12b. However, the case of the 50 GPM leak inside containment is not addressed.
l 9.
Treated under Seabrook item Ba, an item which is independent of j
cause.
OPINION: Acceptable.
~
10.
Incomplete. Although the Seabrook EALs address some facets of this item (e.g., 6c,16c, etc.), it does not appear that " COMPLETE loss of any function needed for plant cold shutdown" has been covered.
Those functions must first be defined.
1 9
i
OPINION:
Probably defined in the FSAR/EOPs.
OPINION: Acceptable.
12.
Incomplete._ The Seabrook EAL (13) states " fuel accident with release of radioactivity". Consider adding the phrase "to containment or fuel handling building".
15.
Incomplete. Typically, there are many technical specification radiological limits with one set called " instantaneous". Assuming that is the case, since no particular technical specifications are referenced in Seabrook EAL 12b omission of " instantaneous" is criti-cal.
16-18.
Incomplete. Lacking a definition of the word " hazard" in Sea-brook 17b, it does not appear that the specific circumstances of the NUREG are covered by this item.
(NUREG-0654 II D 1).
See OPINION under NOUE 13-15.
i 10 f
e SAE 1.
Unsatisfactory. This initiating condition correlates to Seabrook 15. Without procedures E-1, E-2, ECA 1.1, & ECA 1.2, it is difficult to judge compliance. However, the following is clear:
There is no SAE in Seabrook 15.
The terminology " emergency coolant recirculation" is ambiguous.
OPINION: Entry into the E0Ps would lead to an inventory diagram.
With a 150 GPM LOCA, the operator would conclude off normal but not challenge or failure. Therefore, no classification would result.
If this is the case, an event-oriented EAL should be considered.
5, 13a, 13b. Not covered.
8.
Not clear. Many of the applicable situations are covered. To ensure that " Complete loss of any function..." was addressed, defini-tion of functions needed for hot shutdown is required.
- 10. Not covered. Without input ~from radiation levels, the only classification made for a fuel handling accident by the Seabrook plan is an Alert.
1 I
- 11. Unsatisfactory. The Seabrook plan EAL (16c) requires that the fire be uncontrolled, uncontained, and actually affecting safety 11 t
systems. The NUREG merely requires that the fire compromise the -
function of safety systems.
Seabrook's alert level (16b) fire is con-trolled, contained, and potentially affecting safety systems; the SAE fire is uncontrolled, uncontained, actually affecting. There is'no proper classification for a controlled, contained fire affecting safety-related equipment.
OPINION: The entire set of fire classifications in the plan should te
~
reviewed.
- 14. Not covered unless Seabrook 18e " hazards" is defined to include
" imminent loss of physical control of the plant".
Not covered unless some document incorporates the specific detail of the NUREG (e.g., flood, tornado, etc.) into "hezards" in 18c.
GENERAL EMERGENCY (GE) 1.
Incomplete. The aspects of 5 rem /hr thyroid, projected versus actual, and actual conditions are not contained in Seabrook's plan EAL 12d.
~
l Sa-Se. Inadequate.
l OPINION: The utility has relied upon the symptomatic process to encompass all of the circumstances listed in NUREG-0654 Appendix 1, 12
GE PWR Sa-Se, as it eventually would. However, pending red or orange paths and the GE classification that would result, valuable time would be lost.
For example, to establish the C orange /H red condition for GE the reactor vessel would have to steam down to an abnormal RULIS level and all steam generators would have to drop below some (to be determined) low value. As another example, the C orange Z red GE con-~
dition is not met until containment pressure exceeds 52 psig.
It does i
not make sense to wait. The licensee should consider use of event-
~
related EALs rather than relying upon the symptomatic process to led to classification.
7.
Not covered.
A e
13
ENCLOSURE 4.
EPIP ER-1.1 EAL EVALUATION GENERAL FINDINGS The effective date has been omitted in EPIP ER-1.1.
OPINION: This is probably true for the entire family of EPIPs.
Prior to licensing some NTOL plants use this technique to ensure that the
~
emergency plan is not inadvertently invoked prior to fuel load.
If so, acceptable; otherwise, it is a minor administrative goof.
Paragraph 2.1 directs the unit supervisor to wait 5 minutes for the shift superintendent (SS) before assuming SS duties. I recommend that the unit supervisor immediately assume the SS duties, pending his arrival. This would prevent loss of 5 minutes with the 15 notification clock running.
The words "other than fire" should be deleted from the chart EAL 17.
i EPIP ER-1.1 EALs VERSUS NUREG-0654, App. 1 GUIDANCE NOTE: In the remainder of this enclosure, paragraph numbers conform.
to those of Appendix 1 of NUREG-0654. Where numbers are omitted, the reviewer found the plan to conform to the NUREG.
14
If an asterisk (*) is present, provided the Seabrook technical speci-fications require a shutdown to the cold condition under 'the conditions provided, that coment may be deleted.
NOUE l
1.
Incorrect. Seabrook EAL 14b appears to key on SI temination rather than initiation.
I J
l OPINION: The answer to this criticism is in procedure ES-1.1; not J
i held. However, even if the criticism is true, it is a minor differ-ence in timing. Since the end result is identical and the operators will be busy with the plant until termination, I recommend that the EAL be accepted as written, l
2.
Unsatisfactory. The equivalent EAL is Seabrook's 12a.
It appeared to have the following errors:
Add " body" after "whole" in paragraph 3.
1 Paragraph 3, by itself, initiates NOUE if site boundary monitor-e ingindicates>0.06midhrwh'olebody. Since most portable instruments will not discriminate <0.1 mR/hr, retention of this EAL could result in spurious NOUEs.
I recommend a reconsideration of paragraph 3.
3, 4, 5, 6, & 9.
Not covered.
15
4 10.
I recommend that Seabrook 16a2 be revised to substitute " Fire is not reported extinguished 10 minutes after initial observation". As presently worded, the EAL requires a report from the scene that the fire is still burning 10 minutes after it's start to initiate the NOUE. The fire brigade is too busy fighting the fire to mark time and make meaningless reports. Let de control room log time instead.
13b, 14e, the specific cases of 15 & 17. Not covered in Seabrook Eats, ALERT Ib. Unsatisfactory.
Partially covered but as a NOUE. Seabrook NOUE EAL Ba declares this based upon letdown >600 uCf/cc gross and chemical verification. That is acceptable. However, if the intent of Ba was to cover the NUREG 300 pCi/cc criteria, the chemical verification step should detect and report equivalent I-131, not gross activity.
OPINION: It appears that Seabrook has attempted to meet the require-ment by measuring gross activity and extrapolating to equivalent iodine. Since the gross activity letdown monitor and chemical test are both faster than sampling and detemining equivalent I-131, NOUE EAL Ba should be considered as acceptable, provided that Seabrook con-siders RCS gross activity of 600 uti/cc to be an engineering equiva-lent of 300 pCi/cc as I-131. This does not alter the fact that the EAL is at the wrong level, NOUE versus Alert).
Ic, 4, 6, 17b, 18e. Not covered.
16
5.
Uncertain. This item may fall under Seabrook EAL 15a, but that must be verified from procedures E-1 or E-2.
7.
Unsatisfactory. As written, Seabrook EAL 6c would not be declared until power was restored (in <15 min), at which point.it is academic, or until the 6c case was overcome by an outage >15 min, which falls under 6e, a SAE. The NUREG version of the EAL is time independent.
s l
9.
Treated under Seabrook item 8a, an item which is independent of cause.
OPINION: Acceptable.
- 10. Incomplete. Although the Seabrook EALs address some facets of this item (e.g., 6c,16c, etc.), it does not appear that " COMPLETE loss of any function needed for plant cold shutdown" has been covered.
The functions must first be defined.
OPINION:
probably defined in the FSAR/EOPs.
OPINION: Acceptable.
- 12. Unsatisfactory.
Seabrook's EAL 13a Ib and Ic are SAE conditions in the NUREG SAE IC 10, not Alert conditions. Although it appears that 13ala and 2 meet the requirements of the NUREG at the Alert level, 17
13alb and 13alc properly belong in a new Seabrook EAL, one which t
results in declaration of an SAE.
I recommend that 13a be revised and new 13b be added with SAE initiation included.
l
=
l 14.
Incorrect. Seabrook's EAL 9c imposes a 2-hour time requirement
~
before the Alert condition is declared. The NUREG IC has no time factor, just loss of alarms.
OPINION: Some time factor should be specified. Two hours is accept-able given steady-state conditions.
- 15. Uncertain. This condition is limited to gaseous effluents by Seabrook's EAL, and to that extent it is acceptable. The NUREG condi-tion appears to apply to both gaseous and liquid effluents.
OPINION:
It may be that Seabrook considers liquid effluents incapable of impacting site boundary readings.
?
17c. Seabrook EAL 1862 requires that the tornado damage safety struc-tures, the NUREG version, is independent of damage.
OPINION: Acceptable.
17d.
Incorrect. The Seabrook EAL requires not only hurricane winds but also, on back shifts, the SS decision that recall of personnel
/
18 m.
would enhance plant safety. The recall-feature is not included in the NUREG version..
18a & 18b.
Incorrect.
Seabrook's EAL requires aircraft or missile impact on plant structures or components or that an explosion result in damage. The NUREG does not.
l OPINION: Acceptable.
- 20. Unsatisfactory. The aspect of anticipated evacuation is neglected.
l Seabrook adds a requirement to the NUREG version which requires unstable plant conditions before a control room evacuation results in an Alert.
SAE 1.
Unsatisfactory. Seabrook has no SAE EAL.
5, 15b. Not covered.
8.
Not clear. Many of the applicable situations are covered. To ensure that " Complete loss of any function..." was addressed, defini-tion of functions needed for hot shutdown is required.
- 10. Not covered. Refer to comment 12 in the Alert section of this enclosure.
l 19
'h.
b l
'15c.
Incorrect.
Seabrook EAL 18c also requires that on the back shifts the SS determine that the safety of the plant would be enhanced by. recall of personnel..This is not a requirement in the NUREG version.
E I
la.
Incomplete.
No reference is made to actual meteorological condi-tions in Seabrook 12d3.
5.
Inadequate.
1 OPINION: The utility has relied upon the symptomatic process to encompass all of the circumstances listed in NUREG-0654 Appendix 1, GE PWR Sa-e, as it eventually would. However, pending red or orange paths and the GE classification that would result, valuable time would be lost. For example, to establish the C orange /H red condition for GE, the reactor vessel would have to steam down to an abnonnal RULIS level and all steam generators would have to drop below some (to be determined)lowvalue. As another example, the C orange Z red GE con-dition is not met until containment pressure exceeds 52 psig.
It does not make sense to wait. The licensee should consider use of event-related EAls rather than relying upon the symptomatic process to lead to classification.
20 w