ML20206G988
| ML20206G988 | |
| Person / Time | |
|---|---|
| Issue date: | 11/08/1988 |
| From: | Fischer D Office of Nuclear Reactor Regulation |
| To: | Butcher E Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20206G992 | List: |
| References | |
| NUDOCS 8811230054 | |
| Download: ML20206G988 (4) | |
Text
..
[
%e,i UNITED STATES
/
NUCLEAR REGULATORY COMMISSION 3'
n IL I.j w Asms<rfON, D C. 20555
)
NOV 08lEB HEMORAN0UM FOR:
Edward J. Butcher, Chief Technical Specifications Branch FF,CM:
David C. Fischer, chief Special Projects Section Technical Specifications Branch
SUBJECT:
TRIP REPORT - UNITED KINGDOM - PREllMINARY INF0FMAT10N EXCHANGE ON THE PRITISH ESSENTIAL SYSTEM STATUS MONITOR (ESSM)
On Friday, Septenter 23, 1988, I visited Dr. Brian E. Horn of tha 0 neration Development and Construction Division, Central Electricity Oenerating Board (CEGO) in Gloucester, England. The purpose of ry visit was to obtain information en the British Essentie.1 System Status Monitor (ESSM) in advance of a more sneanirgf ul technical exchange between the U.S. and U.K. to be concucted in Neverber.
'A are interested in the ESSM because we are currently evaluating altein3tive approaches that could add greater risk perspectives to Technical Specification requirements imposed in the U.S.
More specifically, the NRC is exploring the possibility of conducting a pilot program in the U.S. that uses a real time risk-based system for setting Technical Specification requirements.
We believe that the 6ritish ESSH, currently installed at the operating Heysham B advanced gas-cooled reactor (AGCR), night provide us with some insights for our pilot program.
The British ESSM is a renu-driven computerized operator aid for determining the risk significance of reroving a system / component from service (either due to ecuipment fault or for raintenance). The system was intended to replace the B.itish equivalent of Technical Specifications (i.e., Identified Operating Instructions) which were perceived as overly restrictive. The system is designed to improve the capacity factor of AGCRs by allowing operators (r.aintenance planners) to iccrtify systems or components that can be placed back in service to reduce the risk associated with various plant configurations. The corputer n.odel is based on the plants Final Safety Analysis Report.
it cor.putes the uncontrolled relee t 97cbability based on ranually identified equipment outates input by control room operatcr(s) or maintenance personnel. Based on this input, operators are told whether:
1.
deterministic criteria require irrnediate remedial action, 2.
prcbabilistic criteria require irrodiate %eedial action.
3.
probabilistic criteria warrants plant reconfiguration within 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> (fixed),or 4
no short tern action is required.
The ESSH ar,alyses the t.f fect of system /corponent outage on the original (licensed) fault trees. The ES%M does not work by r. edifying previously calculated lists of minimal cut-sets. Rather, it recalculates the cut-sets for each plant configuratior..
ESSM fault trees rodel eleven essential and supporting Data for the frequencies of the p) systems and forty-two initiating events.
initiating events and failure data for components are contained in plant data files.
O
)
.j f '
NRRD Q
vnu
I l
4 4 direct updating of actual equipment failure rates or initiating events into the i
ESSM model.
The ESSM takes approximately three minutes to provide guidance to the operators (or maintenance planner).
i
}
The most significant distinction that can be drawn between the ESSM and the r
pilot program envisioned by the NRC is that the focus of the NRC's pilot program i
would be :4 operational safety as opposed to improved plant reliability (or capacity factor).
However, it is believed the improved capacity factor could be j
an important concomitant benefit.
[
My meeting with Dr. Horn and his subordinate Mr. Steven Harris (who is now l
1 directly responsible for the ESSM at Heysham B) was informative.
I described for him the information I had hoped to get during my brief visit (see Enclosure 1).
He highlighted the characteristics and capabilities of the ESSM by walking j
me through a paper he recently presented to the International Atomic Energy i
Agency (Enclosure 2) and by referring to the slides he used to describe the 1
ESSM to the CEGB headquarters staff in London (Enclosure 3).
Dr. Horn and his superior, Mr. Bob Freeman, seemed particularly interest in the technical i
1 information that the NRC might provide to the CEGB as part of the November i
j information exchange meeting, q
a I provided Dr. Horn
[
i ImprovementProgram.pththedetailsoftheNRC'songoingTechnicalSpecifications I also promised to share with him the results of our 1
i ongoing and future work to apply risk perspectives to the existing U.S. Technical
]
Specifications.
A report that described how the CEGB validated the ESSM output was discussed but was unavailable to me because it was confidential.
Dr. Horn i
j indicated that if we really want that report he might be able to get approval to j
release the document to the NRC.
I was able to see a demonstration (from an IBM L
PC terminal in Gloucester that was tied into a mainframe computer in London) of how the ESSM works.
L I
j I suggest that any follow-on technical exchange meeting that we have with the l
J British on the ESSM have a clearly identified and mutually agreed upon agenda.
j I told Dr. Horn that we would call him the week of September 26 to firm up a i
date for our next meeting.
He indicated that late October would be the earliest j
they could get together again and said that a demonstration of the ESSM at l
l Heysham B could be scheduled for that meeting.
C%
% ~.
I l
David C. Fischer, Chief l
J Special Projects Section
[
j Technical Specifications Branch l
j cc:
Dr. Bahnan Atefi, SAIC I! This included back issues of the Technical Specifications Improvement I
Program Highlights (NUREG/BR-0110) and the SAIC "Work Plan for Phase I l
of the Pilot Prugram for Implementation of a Real Time Risk-Based Config-i atton Control Approach to Technical Specifications", August 31, 1988 l
l (SAIC-88).
NOV 0 8 hEB DISTRIBUTION:
0TSB R/F DOEA R/F DCS Central Files.
LCShao, NRR FDCoffman, RES TEMurley, NRR FJCongel, NRR GDMcPherson, NRR JHSniezek, NRR JWRoe, NRR EJButcher, NRR FJMiraglia, NRR SAVarga, NRR MLWohl, NRR CERossi, NRR DHCrutchfield, NRR POR HJFaulkner, IP RJBarrett.NRR OTSB Members GI 7 t 0+
TSB:00EA:NRR C:TSB:00EA:NRR DCFischer:pmc EJButcher 09/ /88 09/s/88 Il[*5 4
i ENCLOSURE 1 QUESTIONS ASKED OF DR. BRIAN HORN AT A PREt.!MINARY INFORMATION i
EXCHANGE BETWEEN THE NRC AND CEGB, 23 SEPTEMBER 1988 GLOUCESTER, ENGLAND 4
1.
What are the capabilities and characteristics of the ESSM7 j
2.
Describe the PRA (Code) model used for the ESSM. What systems and events I
are ecdeled?
i 3.
How do you gain confidence that your model is realistic / accurate? (Do you periodically compare actual system performance levels with system reliability values, calculated using the plant-specific PRA?)
t D
4.
How do you deal with plant parameters that are not modeled in the plants t
risk model (e.g., shutdown margin, initial condTtTon parameters)?
1 t
1 5.
How do you propose to deal with plant-specific characteristics?
l 6.
How often are the inputs to your ESSM pre processor module (e.g., fault j
trees failure data, initiating event defn) updated?
4 I
i l
7.
I understand that the ESSM not only addresses the integrity of the essential i
systems in terms of deterministic criteria, but also in terms of probabilistic criteria, What are your probabilistic criteria? How were they derived and how are they modified? Could such criteria be used for the calculation of i
A0Ts and STIs? (e.g., AR*TSA (some criteria) or l
ACMF*T5B (some other criteria))
[
I told him we were interested in both the form and numerical value of f
1 criteria and asked him whether their criteria were based on probability; i
j
]
consequences; or both.
8.
What have been your experiences in implementing the risk-based approach to l
i Tech Specs (e.g., problems encountered, operator assessment)?
i 9.
The ESSM appears to help operators decide what piece of equipment or system f
to get back on-line after a fault (or whether or not to take a given piece of equipment out at a particular time). What do you think of using such a i
model to specify an A0T or ST!?
[
10.
How do you define the boundary between important decisions and very important l
decisions (ref. PRISM).
3
- 11. How is average risk (or core melt frequency) per year factored into the i
code?
12.
Is there any written material that I can bring with me back to the US?
code description operator manuals assessment reports (re=ults of the program so far) l i
cost of your pilot program i
case studies j
1
!