Throughout this report, pipe sizes are referenced by nominal pipe size (NPS), which indicates standard pipe size without an inch symbol. The smallest pipe size considered in this study is NPS3-% (Plant A. b). All references to specific material types are made according to designations by the American Society for Testing and Materials (ASTM). The term “weld failure” is used to indicate a rejectable (non-through-wall or through — wall) flaw.

During the NRC LOCA Elicitation Kick-off Meeting [D.1], a LOCA was defined as “a breach of the reactor coolant pressure boundary which results in a leak rate greater than 380 lpm (100 gpm).” Instead of using the traditional (or historical) LOCA size classes (small — medium — large) that are based on break size, this study uses LOCA sizes that are based on leak rate threshold values as indicated in Table D. 1 (adapted from [D. 1 ]) and Table D.2.

Table D.1 LOCA Size Classification Threshold Values

The estimation of weld failure rates uses Bayesian reliability analysis methodology, and involves the development of prior and posterior failure rate distribution. In this study the term ‘prior’ refers to piping reliability characteristics before the implementation of industry programs to mitigate or eliminate susceptibilities to certain degradation mechanisms. The term ‘posterior’ refers to observed or expected reliability characteristics after reliability improvement actions have been implemented.

Day 1 — Tuesday, February 4, 2003

Welcoming Remarks, Agenda Review, and General Information

Rob Tregoning of the USNRC began the meeting with a review of the agenda and general announcements. In addition, the individuals present were asked to introduce themselves with a short background of their experience related to the issue of LOCA frequency estimations.

Mike Mayfield of the USNRC welcomed the group and offered his perspective on the subject. Hossein Hamzehee of the USNRC also stressed the importance of the LOCA frequency determination for the continuing effort to explore risk-informed revision of 10 CFR 50.46, which govern ECCS requirements.

The meeting attendance list was provided to all of the meeting participants.

Presentation: Importance of LOCA Distributions to 50.46

The first presentation on the agenda was made by Alan Kuritzky of the USNRC. Alan laid out the importance of LOCA frequency estimates with respect to the 50.46 revision effort. Some of the key points from his presentation and subsequent discussion are outlined below:

• The NRC staff proposed a plan for risk-informing the technical requirements of 10 CFR Part 50 (Option 3) in SECY 99-264.

• Stakeholder input was considered in the recommendation to focus revision on the ECCS requirements.

• These requirements are covered in three regulations: 10 CFR 50.46, Appendix K to 10 CFR 50.46, and General Design Criterion (GDC) 35.

• Potential changes to 10 CFR 50.46 fall in one of the following areas

• ECCS reliability (one of the focuses of elicitation — due to the simultaneous Loss of Offsite Power [LOOP] requirement)

• ECCS acceptance criteria

• ECCS evaluation model

• ECCS LOCA size definition (another focus of the elicitation)

• The elicitation results will impact changes to the ECCS reliability areas and the ECCS LOCA size definition.

• ECCS reliability is primarily impacted because of the effort to eliminate the simultaneous LOCA — LOOP requirement. This has two pieces:

• LOCA initiation frequencies and

• Conditional probability of LOOP given a LOCA.

The focus of the current expert elicitation effort will be to obtain robust LOCA frequencies for use in the LOCA-LOOP evaluation. Interim LOCA frequencies were developed last spring as part of an internal

NRC elicitation effort and these have been used to demonstrate the technical feasibility of a change in this requirement. The results of this interim NRC effort will not be made available to this panel until after the wrap-up meeting to ensure that the panel results are independent. The objective is to have this project completed by December 2003. More detail on this topic is provided in Rob Tregoning’s subsequent presentation.

For the LOCA size definition effort, a computational code is being developed to incorporate LB LOCA contributions from pipe breaks and other component failures. The results of this panel will be used to normalize the analytical code results and also provide distributions for important input variables. The targeted completion date for this technical feasibility study is July 2004. More detail on this topic is also provided subsequently.

Rob Tregoning indicated that the LOCA frequencies would also be used within the 10 CFR 50.61 risk — informed revision effort (PTS Rule) to ensure that current calculations are acceptable.

Split distributions necessary if UB and LB are not symmetric with respect to mid value.

Used log normal distribution since results provided by participants fit log-normal distribution; also log­normal distribution easily to manipulate; also tradition is that log-normal is used in risk based approaches; bottom line is that due to variability in responses should not make that much difference as to what distribution chosen.

Analysis will yield medians of mid values and bounds as well.

This section develops a basis for calculating conditional weld failure probabilities. In the model of piping reliability, the conditional weld failure probability, pL/F, represents the likelihood of a weld flaw propagating to a significant structural failure. However, for Code Class 1 piping there is no service experience data available to support a direct estimation of pL/F. Therefore, the options for calculating the conditional failure probability when the service experience data consists of zero events include applications of (1) probabilistic fracture mechanics modeling and (2) Bayesian modeling. Since the objective of this Base Case was to directly utilize insights from service experience data reviews, the latter approach was selected. Before defining the input to the Bayesian modeling, it is useful to organize the available service experience according to piping classification and severity of observed failures. Figure D.23 shows the conditional pipe failure probability as a function of observed through-wall flow rate for reactor coolant pressure boundary (RCPB) piping (Code Class 1), Code Class 2 and 3 piping, and ASME B31.1 (non-Code) piping.

Б

го

-Q

0

1

ф ‘го о

Б

с

о

о

Figure D.23 Likelihood of Structural Failure According to Service Experience with

Light Water Reactor Piping System

The empirical data that is used to construct the chart in Figure D.23 represents approximately 7,200 recorded pipe failure events representing almost 9,000 reactor-years of commercial nuclear power plant operation. It is important to note that the chart covers a wide variety of piping systems, from Reactor Coolant Pressure Boundary (RCPB) piping, Safety Injection and Recirculation piping, and Auxiliary Cooling piping to ASME B31.1 piping (or non-Code piping). Figure D.23 represents our current state-of-knowledge with respect to the probability of pipe failure and it provides an upper bound for estimates of conditional pipe failure probability. Noteworthy is the fact that only for ASME B31.1 (non-Code) piping systems have significant structural failures been observed (that is, failures equivalent to double-ended guillotine breaks). For the other classes of piping it is necessary to do data extrapolations beyond the Category 1 through-wall flow rates. In this Base Case study, a Bayesian approach is used for this extrapolation and by acknowledging service data insights. For the Code Class 1 piping, it is also important to recognize that according to the available service experience data, through-wall flow rates greater than 38 lpm (10 gpm) have been primarily attributed to failure of small bore piping. It is unlikely that the structural integrity of the Class 1 piping is equal to or less that that of Code Class 2 and 3 piping for which Category 1 LOCA events have been observed.

The Beta Distribution has some convenient and useful properties for use in Bayes’ updating. The analysis starts by defining a prior distribution that represents the analyst’s understanding of piping performance given the presence of some sort of degraded condition. The prior distribution is defined by selecting an appropriate set of initial values for parameters A(a) and B(P), denoted as APrior and BPrior. Then, when looking at the relevant service experience data, if there are “N” structural failures of a certain magnitude and “M” successes

(or degraded conditions that were repaired before progressing to a structural failure), the Bayes’ updated, or posterior distribution is also a Beta Distribution with the following parameters:

APosterior APrior + N BPosterior BPrior M

The above explains how the Beta Distribution can be used to estimate conditional weld failure probabilities. The challenge is to justify the selected parameters when the evidence is zero structural failures. Certainly, it can be argued the ASME B31.1 service experience data represents a very conservative upper bound for the conditional weld failure probability.

Selecting a well justified set of “A” and “B” parameters is not a trivial task. One basic ground rule should be for the “weight” of the field experience data to determine the shape of the posterior Beta Distribution. However, many different parameter combinations will produce the same predicted mean value. Where very little evidence is available about the parameters, constrained non-informative priors may be selected. For such a case, one can say that the “A” parameter has to be a small number.

In this Base Case study, the prior “A” and “B” parameters are defined by first deriving a constraint for the prior mean value of the conditional failure probability and then fixing the “A” parameter at 1.0 for stress corrosion cracking and 2.0 for thermal fatigue to account for the fact that according to available service experience data, thermal fatigue cracks propagate in the through-wall direction considerably faster than flaws caused by stress corrosion cracking. The process for developing conditional failure probabilities starts by deriving a point-estimate of pLF for small-diameter piping given susceptibility to stress corrosion cracking (SCC). This point estimate is based on Jeffrey’s non-informative prior and service experience data. There have been 42 through-wall flaws and zero large leaks in small-bore BWR piping. This gives a point estimate of 1.2×10-2, which is used as a “fix point” for determining conditional weld failure probabilities for other pipe sizes. The relationship between pipe size (diameter and wall thickness) and the conditional failure probability is assumed to follow a power law of the form:

Plif = a x DNb (D.8)

Where, “DN” is the nominal pipe size in [mm]. Decreasing trends correspond to negative values of b. Parameters a and b are determined for pLF = 1.2×10-2 and DN = 25. Point estimates of pLF for other pipe sizes are derived using the power law for conditional failure probabilities and assuming that the general shape of the curve is similar to that of piping susceptible to vibratory fatigue for which pLF = 2.5/DN. Next the predicted conditional weld failure probability using the power law approach is used to determine the Beta Distribution parameter “B.” For Class 1 piping, engineering judgment, as portrayed by Figure D.23 is used to assign values to the prior Beta Distribution parameters. The proposed Beta Distribution posterior parameters for this Base Case study are summarized in Table D. 14.

Table D.14 Proposed Beta Parameters for Code Class 1 Piping

D.5.3 Conditional Failure Probability and Flow Rate

The conditional failure probabilities derived in the previous section are assumed applicable to Cat0 LOCA. It is furthermore assumed that for a significant primary piping breach to occur there has to be a through-wall flaw coinciding with a plant operational mode change or an unusual or severe loading condition such that the leakage exceeds a Cat0 LOCA. The service data collection (e. g., PIPExp) includes numerous examples where pressure pulses or spikes caused by changing flow conditions following a plant operational mode change have resulted in non-active leaks11 becoming active leaks. The physics of such transitions from non-active to active leaks are complex and location-dependent (e. g., function of flaw size and pipe stresses). Some published work exists on the correlation between crack propagation and plant transient history [D.25]. Using available empirical data, the uncertainties in such crack growth assessments are considerable, however.

In this analysis a simple parametric approach is applied to the estimation of weighted conditional failure probabilities (CL) of a pressure boundary breach that exceeds a Cat0 flow rate threshold value. This approach is described through the event tree in Figure D.24. An undetected, or detected but monitored through-wall is exposed to a pressure pulse or unusual loading condition before a decision to perform manual, controlled reactor shutdown. The pressure pulse or unusual loading condition is characterized as a subjectively defined probability distribution.

Figure D.24 Event Tree for Definition of LOCA Categories

In the cases of “moderate-to-high” to “extreme”, the term “unusual” implies a loading condition beyond that resulting from anticipated transients including manual and automatic reactor/turbine trips. The conditional probability of an unusual or severe loading condition is described by five sets of subjective 3-bin discrete and overlapping probability distributions as summarized in Table D.15. These DPDs are combined with the weld failure rate distributions and conditional weld failure probability distributions by using a Monte Carlo merge technique. [11]

Service data on water hammer events provides a justification for the chosen DPDs. From PIPExp, a point estimate for CL-WH-Cat6 is approximately 4.9E-03, which is based on two events involving severe overloading (including plastic deformation) of a pipe section in 411 recorded water hammer events. This is taken as a best estimate CL-value for calculating a Cat6 LOCA. Figure D.24 includes the rules for how the DPDs are applied to the LOCA frequency calculation. The Cat0 and Catl LOCAs include contributions from each loading condition associated with Cat2 or larger pressure boundary breach. In other words, the calculation accounts for the possibility that an ‘unusual’ loading condition may not result in a global or catastrophic pressure boundary breach. Given a through-wall flaw and severe overload, Figure D.25 shows the conditional failure probability as a function of pipe size.

Figure D.25 Conditional Probability of Weld Failure Given Through-Wall Flaw and Severe

Overloading

A discussion was held on the confidentiality of participant’s responses during the exercise. Rob Tregoning indicated that all information provided as part of this exercise will remain confidential and will not be distributed to anyone not specifically involved in the exercise. The kick-off meeting has been videotaped, but this will not be distributed outside of the group. Elicitation sessions will be taped for accuracy, but this information will also not be made public. There will be public reporting of the assumptions, methodology, elicitation results, and calculated LOCA frequencies that stem from this exercise. However, the summary reporting will only identify the names, affiliations, and possibly credentials of the expert elicitation panel and the facilitation team early in the report. No reference to individual opinions will be documented.

PORV stands for pilot operated relief valve.

Difficult to correlate stuck open valves categories with leak rate sizes/categories; size of valves will vary between plants.

Presentation #20 — Emergency and Faulted Loading Base Case Development By Gery Wilkowski

No uncertainties applied to loads in Gery’s analysis.

Base case assumes idealized TWC geometry.

Did not do any subcritical crack growth; thus did not consider residual solutions.

LBB. ENG2 is in form of closed form solutions.

For predicting large crack growth in a pipe tests it is better to use J-M than J-D.

Duane Arnold crack would have failed if subjected to Level B type loading.

Global secondary stresses act as primary stresses if crack large enough such that failure stress is below yield strength.

Mr. Deboo has 28 years experience working in the nuclear power generation field. Mr. DeBoo’s recent experience includes fatigue, crack growth and flaw stability analyses necessary to demonstrate operability for most power plant components. These evaluations would include root cause and remaining life determinations. He has extensive experience with IGSCC and other material degradation issues. He also has performed and supervised functionality and operability evaluations of systems and components to address unanticipated operating events or conditions, which do not meet inspection or test requirements.

During his 28 years in nuclear power generation, Mr. DeBoo has worked on three major nuclear projects (six units) including all design, inspection and testing phases leading to commercial operation. This experience included system and component seismic qualification, component fatigue qualification, licensing and design review for fuel load, special analytical assessments of the safety significance of installation discrepancies, and system/equipment/component functionality/operability evaluations resulting from startup and operating test programs.

Mr. DeBoo has extensive experience in the evaluation of fatigue-related problems, material degradation issues and the assessment of remaining life in vessels, piping and supports for nuclear applications. He has performed safety evaluations for unanticipated operating events, and has developed plant-unique acceptance criteria to permit continued operation. Those events include fluid transients, thermally stratified flows, and flow-induced vibration. He supervised the AMSE Class 1 piping fatigue analysis on two BWR units and two PWR units.

Mr. Deboo has a B. S. in Mechanical Engineering from Northwestern University and a M. S. in Mechanical Engineering from the University of Illinois. He is a member of the ASME. Currently he is serving on Section XI of the ASME Boiler and Pressure Vessel Code as a member of the Working Group on Pipe Flaw Evaluations and as the Secretary of the Working Group on Flaw Evaluation.

Rob stressed that prior to their individual elicitations, each panel member needs to do their homework and answer as many of the elicitation questions as possible. Panel members can change their answers at any time during this exercise, including during the actual elicitation and afterwards.

It was indicated that the term “LOCA frequencies” should really be “LOCA probabilities” in this presentation during the analysis of the conditional emergency loading. Rob agreed to make this change to the presentation.

Pete Riccardella questioned whether we should expand the conditional seismic loads to conditional emergency and faulted loads which include transients like water hammer as well as seismic. Fred Simonen thought that we needed to talk with the PRA people. Alan Kuritsky (USNRC) indicated that these other potential LOCA causing events were currently not included in the PRAs. Rob suggested tabling this discussion until later in the agenda. Bruce Bishop thought that if we continued with the traditional seismic approach then we need to consider the fact that the snubbers may not work probably. Bruce indicated that there is a significant probability that the snubbers may not work as advertised.

Rob next addressed the top down elicitation structure. Bill Galyean used a top down approach where he assigned an overall piping LOCA contribution, and then looked at the breakdown in the contribution due to piping system, geometry, load history, mitigation, materials, and degradation mechanisms. Vic Chapman questioned what the top down approach gave us (we start with the final answer that we are looking for). Lee Abramson indicated that at the end of the day we will get numbers in each of the blocks on slide 3 so that we can decompose the problem into the small pieces. The panel members can initially choose either a top down or bottoms up approach.

Bruce Bishop asked how soon the panel members are going to have the final base case results prior to the first elicitation. Rob indicated that he would be working with the base case members the week of June 9 to finalize their answers. He hoped to have the final results by the end of June. However, the base case conditions are well-known

The panel members need to supply their answers (on a pre-established form) and the facilitation team will work with the individual panel members individually. The important point of the pre-elicitation exercise is to quantify the median value results, provide some qualitative uncertainty and also rationale. During the elicitations the panel members can change answers as they interact with the facilitation team and points are clarified.

Bruce Bishop felt that the “utility safety cultural” should be “utility operations cultural” in that safety and economic drivers both feed into the operations cultural. Karen Gott indicated that the IAEA definition of safety cultural (and how we defined safety cultural at the kick off meeting) includes both safety and economic aspects. With regards to the ratios on safety cultural issues, ratios greater than 1.0 indicate that things are getting worse; less than 1.0 means things are getting better.

Slide 6 shows a flow chart in which the question is which variables are independent. Variables in this context are geometry, load history, mitigation, materials, and degradation mechanisms. For these variables, each panelist may need to estimate the future impact of that variable, e. g., what new materials, or what new degradation mechanisms, should be expected in the future. We may want to look at the past to estimate what might happen in the future (e. g., Gery Wilkowski’s plot of new failure mechanisms with time which shows a new mechanism approximately every 7 years).

There ensued a long discussion on the comparisons of reference cases (defined at the kick off meeting) with baseline cases. We defined a reference case for each piping system (e. g., diameter, material, degradation mechanism, etc) whereas we only defined a few baseline cases to which the reference bases are to be compared (i. e., anchored). For example, the hot leg (base case) may be a natural comparison for the reference case for the cold leg.

For Questions 3A.1 and 3A.2 (slide 7) and all related questions, Rob will change the “surge line” to “cold leg” example so questions don’t refer to a system that is both a base case and a reference case, realizing that we will end up asking same question for a surge line as well. For those systems which have both a base case and a reference case, these comparisons may be more natural and easier than inter-system comparisons.

It was decided to eliminate the assessment of which variables are independent or dependent. All variables will be considered to be dependent as originally defined during the kick-off meeting. Thus, the original Question 3A.2 will be eliminated. Other References to correlated or independent variables in other questions should be eliminated in the final presentation version. For original Question 3A.3, the requirement to list at least 80 percent gets the most significant contributions, but not all of them.

There was considerable discussion on what the panel members would need to provide for Question 3A.3 and 3A.4. We need to know what variables have a major impact on LOCA frequencies to quantitative that impact as best as possible. Lee Abramson tried to make the point that things would become clearer once individual panel members got into their elicitations and tried to put numbers to the answers to the questions.

A question was asked if any attempt is going to be made to look at plants and determine how many plants have a certain combination of variables (V1, V2, V3, etc.), and how many plants have another set of variables (V2, V4, V5, etc.), and how many have another set. It was noted that some variables will be important at certain plants and other variables will be important at other plants. Rob indicated that it would be nice to have such information, but it was not practical to get such information in the time frame we have. This could possibly be a follow-on effort. However, it should be stressed that the plant design information will not likely result in a significant change in the analysis. If certain designs do not contribute to the LOCA frequencies then they are not significant contributors and the panelists can focus on designs that do as long as they exist in several plants. If the population of the significant contributors is in error by 2 to 3, it will likely not matter.

The only issue to avoid during quantification is if you believe that only a few (1 — 2) plants of a certain design, operating experience, etc. significantly contribute to the generic LOCA frequencies. These plants should not be explicitly considered in these generic estimates. However, but possibly applicability of these generic results to those design conditions should be discussed during the elicitation.

Elicitation Question (EQ) 3A. 1 compares a base case to a reference case, then EQ 3 A.4 will use the impact of the important variables to compare other similar piping systems to the reference cases. Reference cases are the link back to the base case for which we will have actual LOCA frequency estimates. The panel members don’t have to do the mapping back to the base cases, the facilitation team will do that. Then the facilitation team will filter the results up (bottoms up approach) to get an overall LOCA frequency.

There was some discussion about how the facilitation team would integrate these results and how the panelists could account for these individual contributions.

Again, it was emphasized that if panel members are not comfortable in answering specific questions, then they need to say so. If the panel members need to make a crude assumption, then do so, but indicate that during the individual elicitations so the facilitation team can help estimate the level of uncertainty. Dave Harris thought that it won’t be clear to him how this all fits together until he starts the process. Then he is sure that he will have a lot of questions. Rob Tregoning and Lee Abramson told him, and the rest of the panel members, to call them for any clarification of questions. Rob also indicated that he will be contacting each panelist prior to their elicitation to discuss issues.

The flow chart on slide 9 is for a “top down” approach, much in the motif of what Bill Galyean discussed on Day 1. As part of this approach, one only needs to tie one system (of those identified as being important contributing systems to LOCA frequencies) to the base case since previously we had identified individual piping system contributions. One can make this connection through a reference case if not a base case system, or can tie directly to the base case if the system is a base case system. This approach may be more straightforward for people who need to integrate variables in mind, which might lead to more uncertainty.

Panel members can chose which approach to follow (top down or bottoms up), and they can switch back and forth depending on different systems. As part of their homework prior to their elicitation, they only need to do one approach, but the facilitation team may ask about each approach during the elicitation. There will be different tables to fill out with each approach. The bottoms up approach may be more rigorous with less subjectivity, but the top down approach may be easier to understand. One can get top down answers from the bottoms up approach, but can’t do reverse. By doing both approaches for certain systems, panelists can search for consistency in their analysis.

Rob discussed the elicitation questions related to non-piping components starting at slide 12. At the kick­off meeting in February we didn’t spend as much effort developing the base case and reference cases for non-piping components as we did for the piping systems. Thus, these gaps needed to be filled during the remainder of this meeting.

Slide 12 illustrates the flow chart for the “bottoms up” approach for the non-piping components. At the kick-off meeting, we had identified five (5) non-piping components to consider (pressurizer, valves, pumps, RPVs, and steam generators). In order to estimate the frequencies for these non-piping components, the panel members need to pick either a piping or non-piping base case for comparison. The facilitation team will integrate the results in a manner similar to the bottoms up approach for piping systems.

There was a question about the nature of the non-piping base case conditions, especially in light of the thorough discussion about the piping base cases during the previous day. We had originally planned to have precursor data for the non-piping base cases for comparison. However, we do not have data identified yet. We may have to drop the idea of using non-piping base cases and only have piping base cases to compare to. We will come back to this issue later in the afternoon.

Bruce Bishop felt that tying non-piping components back to piping base cases would be difficult. He foresaw lots of dissimilarities between non-piping and piping in failure mechanisms, etc. He suggested that we try our best to come up with some non-piping base cases. Even if we can’t come up with base cases for all 5 of the components, if we could come up with base cases for a few, that would be better than nothing.

Pete Riccardella asked if we have defined the failure mode for these non-piping components. Rob felt that we don’t have a clear definition at this time. Rob felt that we had to take more of a mechanistic viewpoint. There was also general confusion about the definition of “failure mode” for the non-piping issues. Rob indicated that this means the failure mechanism. It was agreed that “failure mechanism” is a preferable term and Rob will change the phrasing for the elicitation questions from “failure mode” to “failure mechanism” to avoid any confusion.

It was again emphasized that the panel members will not be asked to provide absolute LOCA frequencies. However, if a panel member prefers to think in terms of frequencies, they should feel free to do so. Elicitation questions will ask for relative comparisons with the base cases and other conditions. The facilitation team will then make the calculations to get the absolute LOCA frequencies. The panel members should make the best comparisons possible and are not compelled to answer questions in areas where they have no expertise.

Rob showed an example of a table that he may provide the panel members for them to fill out for EQs 3B.1 and 3B.2 for the top down approach, see Table B.2.2.

The complete set of tables to be filled out for the elicitation will be provided electronically by Rob. There will be a space for comments in each table row to initiate discussion during the elicitation process. The tables will be provided in Excel format. If a panel member wants to change the Excel spreadsheet format they should feel free to do so as long as the cell references for each answer remains unchanged. The final calculations will be done in Excel. Therefore, the elicitation results should be provided to Rob in the excel spreadsheets if at all possible. Hardcopies or MS Word versions of the tables can provide upon request.

Rob will provide the panel members with a copy of the spreadsheet that he will use to calculate LOCA frequencies sometime during the elicitation process. Rob will attempt to complete this spreadsheet to the individual elicitations so that the panel members can see how their responses reflect their calculated LOCA frequencies. However, this will be a lower priority than coordinating the information exchange among the expert panel and finishing the base case calculations.

Table B.2.2 Elicitation Questions 3B.1 & 3B.2

This report consists of eight sections and four appendices. Section D.2 is an overview of the analysis steps. Section D.3 summarizes the service experience applicable to the BWR and PWR Base Cases, respectively. Using the PIPExp database, Section D.4 includes a summary of the data interpretation and data processing steps necessary to derive piping reliability parameters that apply to the base case definitions. Section D.5 documents the results of the pipe failure rate estimation while Section D.6 is a documentation of the models used for estimating LOCA frequency, while Section D.7 is a summary of results. Section D.8 is a list of references. Note that the Base Case results used in Table E. 1 in the main body can be obtained from Tables 16, 17, and 20 in this report.

Appendix A summarizes the PIPExp database structure. Appendix B includes the Excel spreadsheets that are used as the basis for the LOCA frequency models, and Appendix C includes the Excel spreadsheets for the calculation of time-dependent LOCA frequencies. Finally, Appendix D is a summary of selected, significant Code Class 1 and 2 pipe failures in commercial nuclear power plants worldwide.

The next presentation was made by Bill Galyean of INEEL in which he reviewed Appendix J of NUREG/CR-5750. Some of the key points from his presentation and subsequent discussion include:

• There are a number of varieties of LOCA initiating events, including:

• Traditional pipe break LOCAs

• Stuck open PORVs and SRVs

• Steam generator tube ruptures

• Reactor coolant pump seal failures

• Interfacing system LOCAs (ISLOCAs) — where primary system coolant is inadvertently introduced into the secondary side piping and a secondary pipe fails creating a leak path of primary coolant outside containment

• Reactor vessel rupture

• While failure data exists for some of those categories, data for pipe break LOCAs and other similar events simply does not exist because it has never occurred.

• There is methodology for estimating the frequency of an event that has never occurred. A Bayesian update of a non-informative prior can be employed. This assumes that the mean value for the distribution is Уг of a failure over the service life. This can be result in a very conservative estimate because the assumed failure frequency in the prior is so high (pf = 0.5). If the failure rate is not constant over time, one also needs to account for time dependency and this methodology is not equipped for this.

• A primary Appendix J assumption is that you needed a leak before you can get a break. A conditional pipe break probability given a leak was based on the Beliczey-Schulz correlation.

• There was also a presentation of passive LOCA failures that can occur in non-piping systems as well as a list of possible data sources for this information.

Discussion: The elicitation panel discussed the validity of this assumption for degradation mechanisms that result in long surface flaws which are not as likely to leak prior to failure. Also, the expectation is that leaking flaws will be fixed after they are discovered during a plant walkdown or through other leak detection methods.

Discussion: Bruce Bishop indicated the need for very clear definitions of what constitutes a large, medium, small, and very small break LOCA. The concern is that the system response to a DEGB where the flow rates can reach 860,000 gpm (3,250,000 lpm) (according to Westinghouse calculations) is very different from a 5,000 gpm (19,000 lpm) leak which is also often characterized as a large break LOCA. Rob Tregoning indicated that clear definitions will be developed as part of this exercise.

Discussion: Gery Wilkowski relayed information provided by Helmut Schulz that the Beliczey and Schulz correlation of conditional probability of a rupture given a leak was developed for cyclic fatigue crack growth.

Discussion: Rob Tregoning emphasized that Bill Galyean’s presentation was provided to recap the last NRC-sponsored work in this area. This NUREG/CR-5750, Appendix J approach is not endorsed for the expert elicitation process; however, it represents one manner in which LOCA frequencies have been developed. Tregoning also emphasized that because substantial LOCAs have not occurred, past operating experience data needs to be augmented by information from other areas. If information was available simply from operating experience, there would be no need for the elicitation.

Discussion: The point was also raised that the panel needs to consider LOCA sources other than traditional pipe LOCAs.