ML021630293
| ML021630293 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 05/14/2002 |
| From: | Kolaczkowski A, Whitehead D Sandia, Science Applications International Corp (SAIC) |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| Download: ML021630293 (6) | |
Text
Preliminary Results:
Beaver Valley PTS Analyses May 14, 2002 presented by Donnie Whitehead (SNL)
Alan Kolaczkowski (SAIC)
Beaver Valley PTS Analysis
- PRA model (SAPPHIRE)
- Large event trees / small fault trees
- Nearly all data is generic based on actual experience
- Human action values considered realistic but with conservative bias (basis: BV procedures, training)
- Multiple initiators (many are BV support system losses)
- Beaver Valley systems/dependencies
- Thermal hydraulic model
- RELAP of Beaver Valley
- Probabilistic Fracture Mechanics
- FAVOR (post-2/2002 version) of Beaver Valley
PRA, T/H, PFM Interaction & Integration PRA Event Sequence Analysis Sequence definitions Sequence frequencies, [fr]
Thermal Hydraulic Analysis PFM Analysis Pressure &
temperature vs. time Conditional probability of vessel failure, [CPF]
Yearly frequency of thru-wall cracking
[CPF]x
[fr]
- In model (prelim CPI/CPF x scenario frequency is potentially high)
- LOCAs: 1.5 to 22
- 2 PORVs open/both reclose
- 1 PZR SRV open/recloses
- 2 PZR SRVs stuck-open including possible reclosures
- Feed & bleed (all 3 PORVs)
- Small (at least equivalent to 3 secondary safeties) to large main steam line break (upstream of MSIVs)
- SG overfeeds (but probably will be relatively unimportant)
- Excluded from model (prelim CPI/CPF x scenario frequency is low {~E-11/yr})
- 1 or 2 PORVs stuck-open
- 1 PZR SRV stuck-open
- 3 or less stuck-open secondary valves even with contd. feed
- Combinations of secondary valve openings & PZR PORV/SRV openings/LOCAs
- SGTR events Scenario Modeling Considerations
LOCA frequencies (<E-3-E-6/yr) x CPFs (<E-2-E-4) yield < E-7/yr thru-wall failure frequencies Feed and bleed (<E-5/yr) x CPF (<2E-4) yields ~E-9/yr thru-wall failure frequency 2 stuck-open SRVs (~E-5/yr) x CPF (<2E-5) yields ~E-10/yr thru-wall failure frequency 1 (~E-3/yr) or 2 (~E-5/yr) stuck-open SRVs and reclosures are still being looked at but could be ~E-8/yr thru wall failure frequencies Large main steam line break with failure to isolate feed (~2E-5/yr) still being checked including inadvertent RCPs shutoff (~E-7/yr)
Other scenarios still being looked at but are expected to be lower contributors Bottom line: LOCAs & multiple PZR stuck-open valves (including reclosures) appear to dominate PTS risk PRELIMINARY Results
- We Are Requesting Comments
- Initiators, event tree structures & dependencies
- Data (do any look very inappropriate?)
- Human actions (do values appear reasonable?)
(Note: HPI throttling/subsequent RCS pressure control is still a key action with a screening value of 50-50; in most cases will be ~0.1 or less in final analysis)
- Other comments / observations
- Revise/complete analysis including full integration with uncertainties Next Steps (PRA model)