EQE International

1
EQE International

• The use of Fragility Analysis in Seismic Safety
  Cases for Nuclear Power Stations

2
EQE International

• LAYOUT
• BACKGROUND
• OBJECTIVES
• OUTLINE METHODOLOGY
• FRAGILITY CURVE DEVELOPMENT
• ALARP ANALYSIS
• CONCLUSIONS

3
EQE International

• BACKGROUND
• LTSR and PSR completed to deterministic
  principles
• Regulator driven requirement to confirm that the
  risk posed is ALARP
• Not a Seismic PSA

4
EQE International

• OBJECTIVES
• To confirm that the current level of
  qualification renders the risk of unacceptable
  seismically induced damage as ALARP
• To understand the relative contributions of
  different items of plant, equipment and
  structures to the risk profile
• To identify areas of plant, the modification of
  which would decrease the annual failure frequency
  to the lowest reasonably practicable level

5
EQE International

• OUTLINE METHODOLOGY
• 1 Identify claimed safety functions and
  associated equipment/structures
• 2 Identify deterministic margins, and the methods
  of assessment used. Normalise margins.
  Characterise Variance in Capacity. Calculate
  median margins
• 3 Convolute the hazard and fragility curves and
  screen out items with an annual failure frequency
  below a defined cut off

6
EQE International

• OUTLINE METHODOLOGY cont
• 4 Identify modifications to increase the margin
• 5 Associate a bounding fault categorisation to a
  loss of safety function. Convert the fault
  category to an associated environmental cost as
  defined in the Safety Review Guidebook
• 6 Calculate Cost Benefit Ratio (CBR)
• 7 Review results, and provide information to the
  Client to assist in the decision making process

7
EQE International

• FRAGILITY CURVE DEVELOPMENT
• Identify Assessment Types
• 1 Code Based eg BS 5950, 8110, 5628, 806,
  ACI 349
• 2 Other Codes eg R6, SQUG
• 3 Comparison with available experimental data
• 4 Engineering Judgement

8
EQE International

• FRAGILITY CURVE DEVELOPMENT
• Use EPRI and other research work previously
  undertaken to derive fragility curve parameters
• Development of the application of standard
  curves for other assessment types eg R6

9
EQE International
10
EQE International
DERIVATION OF MEDIAN FRAGILITY The median
fragility â is a function of the following
parameters a) the normalised assessed
margin Ma b) the Code factor Fc (default value
1.0) c) the HCLPF factor d) the combined
logarithmic standard deviation ?c e) the peak
ground acceleration for the sites 10-4
p.a. URS event, å
11
EQE International

• NORMALISATION OF MARGINS
• Aim to place all margins on a common basis
• Examine basic input for all calculations, and
  acknowledge degree of sophistication applied to
  assessments
• Identify appropriate fragility parameters
• This process is to a certain degree iterative,
  particularly for items with apparently anomalous
  margins

12
EQE International

• NORMALISATION OF MARGINS- KEY FACTS
• Margins have been derived over c 15 years using a
  variety of approaches
• The current state of the art is more advanced
  than when the LTSR work began
• For some cases, seismic margins cannot be readily
  extracted, for example elements with high working
  stresses, where only margins against total load
  have been assessed
• Margins are often calculated against eg a code
  allowable, not a true measure of the loss of
  functionality

13
EQE International
14
EQE International
15
EQE International

• ALARP ANALYSIS
• Quantifiable Cost Elements
• Basic time and materials costing
• Costs as an overrun of outage
• Costs of safety case preparation
• Non-Quantifiable Cost Elements
• eg
• Worker Dose (qualitative ALARP)

16
EQE International

• CALCULATION OF COST BENEFIT RATIO
• Annual Safety Function Failure Frequency
• Annual Equipment Failure Frequency X Probability
  of Safety Function Failure
• Risk cost (valuation)
• Accident cost x Remaining Station Life X
  Annual Safety Function Failure Frequency

Modification Cost for the Plant Item in
Question Current Risk Cost - Risk Cost Post
Modification
C.B.R.

• Rank
• Review
• Report

17
EQE International

• CLOSURE
• The approach has been very useful in acting as a
  ranking tool to better understand the key
  contributors to the risk profile
• The approach used has shown the absolute
  necessity for normalisation of margins
• The approach used allows judgment to be applied
  in a quantifiable manner. This has overcome the
  limitations inherent in alternative, purely
  mechanical evaluation methods