Tables I.4 and I.5 summarize and combine the above RPV LOCA frequency results for BWRs and PWRs, respectively. For BWRs, three RPV LOCA contributors are addressed: RPV vessel beltline region, large nozzles (6 through 28 inch diameter), and small penetrations (partial penetration welded nozzles, 4 inch in diameter or less, such as CRDs). The individual LOCA probability contributions for each of these are provided in the top three sections of Table I.4, and they are summed in the bottom section of the table. These address all LOCA categories as well as the three time periods under consideration (0-25 yrs, 25-40 yrs, and 40-60 yrs). Note, in Table 4.4 in Section 4 of the main body of this report, only the results for the BWR beltline region and the large feedwater nozzles are presented. The results for the BWR CRDs and other small penetrations listed in Table I.4 of this appendix were not included in Table 4.4 since the estimates for the BWR CRDs and other penetrations are not based on analysis, but instead, were based on engineering judgment, i. e., BWR CRDs and other penetrations LOCA frequencies were simply assumed to be a factor of 10 less than the PWR CRDM LOCA frequencies, which were based on analysis.

The first contributor is the BWR shell region, the failure probabilities for which are dominated by the irradiated reactor vessel beltline region. The upper section of Table I.4 summarizes the results of this analysis from Table I.2, in terms of the probability of leaks of various sizes due to degradation or failure of the RPV beltline region. (Note that these were modified slightly relative to those in Table I.2 to eliminate the negative time factor for Category 5 and 6 LOCAs in the 25 — 40 year period.)

The second section of Table I.4 addresses large nozzle contributions to LOCA probability, which in BWRs are assumed to correspond to the 12 inch diameter Feedwater Nozzles that experienced thermal fatigue cracking in the 1980s [I.5]. LOCA probabilities due to this contributor are given for break Categories 1 through 4, taken from Table I.3 as this nozzle size could not lead to larger break sizes. For the Category 2 through 4 LOCAs for 0-25 year time frame, an assumption was made as to the size factor that each successive LOCA size greater than Category 1 was 5 times less likely to occur than the previous size LOCA. Then, for the 25-40 and 40-60 year time frames, the same time factor as determined for the Category 0 LOCAs in Table I.3 was assumed for the larger size LOCAs. Breaks of the other, larger diameter nozzles, such as recirculation outlet nozzles, are considered to be adequately encompassed by the vessel beltline case.

Finally, the third section of Table I.4 lists LOCA probabilities due to failures of CRDs and other small penetrations in the BWR vessel. These were estimated from the detailed analysis of PWR CRDM penetrations described above (Table I.1) but they assume that the BWR penetrations have about an order of magnitude lower LOCA probability than similar penetrations in a PWR. The order of magnitude reduction is deemed appropriate, because problems in small vessel penetrations in BWRs have occurred at a much lower frequency than the recent PWSCC experience in PWRs, upon which Table 1.1 is based. The problems in BWR penetrations have also been attributed to a fairly well-understood phenomenon (IGSCC) and in most cases the nozzles of concern have been mitigated by design and materials changes.

Table I.5 provides a similar summary for PWR RPVs. In this case, LOCA probabilities are reported for only two categories of LOCA contributors, the shell region (RPV beltline) and small penetrations. Again, as was the case for BWRs, Table 4.5 in Section 4 only includes the results for PWR CRDMs. It does not include the results for the PWR beltline region as reported in Table I.5 of this appendix. As was the case for BWR CRDs and other penetrations, the PWR beltline results in Table I.5 are not based on analysis. Again, they are based on engineering judgment, i. e., the PWR beltline LOCA frequencies in Table I.5 of this appendix were simply assumed to be a factor of ten greater than the BWR beltline LOCA frequencies from Table I.4. It was judged that the large nozzles in a PWR RPV do not pose a significant LOCA risk because they are not subject to significant thermal cycles such as the BWR Feedwater nozzles, and except for the safe-ends (which are covered in the piping elicitation), they have not experienced any degradation mechanisms to date. The contributions for the two PWR RPV LOCA contributors are summed in the bottom section of Table I.5.

For the PWR beltline region, results from a prior analysis of a PWR vessel using a third version of the VIPER software (VIPER-PWR) were reviewed. Based on this review, it was estimated that the PWR RPV beltline region presents about an order of magnitude increase in large rupture probability relative to that of a BWR, because PWR beltlines are more highly irradiation embrittled, and because they are potentially subject to PTS transients. Thus, the BWR RPV beltline region LOCA frequency entries in Table I.4 were multiplied by a factor of ten and entered in the upper section of Table I.5.

PWR CRDM penetrations results were entered directly from the above PFM analysis results in Table I.1.