ML20206A205
| ML20206A205 | |
| Person / Time | |
|---|---|
| Issue date: | 10/21/1988 |
| From: | Butcher E Office of Nuclear Reactor Regulation |
| To: | Jenkins P UNITED KINGDOM |
| References | |
| NUDOCS 8811150013 | |
| Download: ML20206A205 (5) | |
Text
_- __ _ _ . - _ _
,#jb '%q'g UNITEJ 8TATES
['
s 3 NUCLEAR REGULATORY COMMIS810N wAsMmotom, o. c, aones k,,,,, 00T 2113B8 :
r Mr. Peter D. Jenkins Safety Technology Engineer Generation and Construction Division Barnett Way ,
Barw e d, uloucester GL4 7RS Cagland
Dear Mr. Jenkins:
As we agreed in our earlier correspondence. hr. Fischer of my staff eet with ,
Dr. Horne on Septerter 23, 1988 to ber,Jn the exchange of ideas between our i two organizations on risk-based Technical Specifications. This meeting was quite useful as the starting point for this exchange and I would li6e to take '
this opportunity to thank Dr. Horne for providing his time on short notice.
I subsequently talked to Dr. Horne via telephone 3 bout our proposed trip to England and would line to confirm these discussions with you. Based on the l availability of Dr. Herne and the plant staff it was agreed that we could meet on hovember 26 30. 1988.
Dr. Horne suggested that perhaps we could spend the first day at your of fices in Gloucester where we could discuss Techntal Specification programs in each-of our crganizations and learn more about the details of your ESSM compute. code.
This could be followed by spendirg some amount of time Tuesday at the plant whure we could see the actsel set up of the E55M and talk to operations i
personnel about their use of We system. We have left Wednesday morning open '
for any discussions 10f t over f rom the first two days and plan to go back to London Wednesday afternoon. I
~
l also told Dr. Horne that we would like to invite a representative of one of l the utility cospanies that operate a nuclear power plant in the U. 5. to partic-1pate in our meetings. Dr. Horne stated that he did rot think this would be a prcblem and on tiist basis ! intend to explore this possibility with a utility !
that is beginning to integrate principles similar to tnose the ESSM is based on ,
into their operations. i To f acilitate our discussions we have put together a list of Westions or l toptes that we wculd like to discuss. A :opy of this list is enclosed, f 1
I l/ i I
g 1' ^ i p,3 / t }-
r o
d' 'yl g,J < / [,A
~
Nd**5 $23 M @22 m
p,Y II
Mr. Peter D. Jenkins 2 Please let te know if these arrangerents are acceptable ta you. I look forward to meeting with you in the near future.
Sincerely.
Origial Si lned By:
Edward J. Batcher Edward J. Butcher. Chief Technical Specifications Branch U, 5. huclear Regulatory Comission
Enclosure:
As stated cc: John McLeod. Nll Cr. Brian E. Horne.CEGB Or Nigel J. Holloway, hl:
Dr. Eahm:n Ateft. Salt OlSTRIEUT :.:h:
DT5 FCoffetn. FES 0HcFherson, NRR l M Central Filer EJButcher FCR TEHurley, hPR COEA R/F JH$nie2ek. LER 0158 R/F FJHiraglia. NFR HFaulkrer. IP CERossi hER BKGrires, hPR
($hao, hFR FCorgel, hER I JRoe, hPV 5Varga hRR CCrutchfield, hER RBarrett, hER C:Old,hEA:hKR EMcWr:rre 10/aA/88 l
r
$UGGESTED TOPICS FOR D15CUS$10N l l
1, characteristics of the E55M :
- 1. Level of Oetail of the Plant Model lecorporated in ($$M. l t
. Component Level
. Supercoetonent Level }
- Independant Subtree l r
Methodology for Representation of the Plant Model, j
- 2. i
. Independen', f ault trees ut Ing support state methodology, J
. Merged system fault trees (1 !arg! fault tree for e'ach !;
accident sequence).
. One large fault tree representing plant model.
- 3. Calculation Methodology
. Regeneration 89 minimal cutsets in each calculation. ,
f
. Manipulation of a large number of sequenae cutsets generated [
with no cotrponent out of service. [
. Use of cutoff fregtency. l l 4 Hardware Requirements
. Primary Memory ,
j . Data $torage -l l
1 l S. Speed of the Program i
. Discussion of several actual cases.
. Limiting cases. t
- 6. Data Used in Qualification of Plant Model [
l
. Generic
. Plant specific [
. How cften will it he updated? l
. Model 9 sed for ur:,vailability of standby coeponents (1/2 f) [
7 Modeling of Human Errors j
. Cognitive human errors.
1
l
- Dependencies of human error on plant status, i.e.,
availability of information, level of stress, time availability for taking actions.
- 8. Given unavailability of one component, does the model know which other components or systems would also be out of service due to single components being common to more than one system.
- 9. How are support system interactions modeled and the impact of support system component maintenance outages evaluated?
- 10. Are any of ESSH capabilities specific to AGRS or is it general encugh to be applied to any other plant.
- 11. Other Technical Specificattor.s-Related issues
- 1. Discussion of the ESSM Criteria Used for Assessment of Acceptable Allowed Outage Times 1
- Completely Probabilistic (i.e., maximum core melt f requency)
' - Probabilistic and Deterministic
- Delta Core Melt Frequency'
- 2. How are the Current Surveillance Tcst Intervals Chosen?
- Deterministic
- Reliability-Ba sed
- 3. What Methods are Used for Settir.g Technical Specifications Requirements that Cannot be Handled by Plant Risk Model and ESSM?
- Deterministic
- PartDily Risk-Based
- What portion of the requirements can be handled using risk-based methods? ,
111. Operational Interface with_the_Model
- 1. Acceptance of Risk-Based Technical Specifications by the operating staff.
- 2. What is the process used for inputting chariges to plant status?
- Number of authoriztd people
- Process for assuring accuracy of the changes f r
i r
2-4
_~ , ~ - - - - , ._y_ . - , _- ._, _ _ _ _ _ _
7
- 3. Does this process help in assuring that only t5e correct components are taken out of service for test or maintenance?
- ESSM-controlled tag out
- Other control mechanisms
- 4. How Often is the Model Updated?
- Model based on average component unavailabilities.
- Frequent updating of the model based on actual unavailabilities of standby corponents.
I' 4
I l'
1 t
I 3- l
, - - _ - _ _