The participants generally thought that the important degradation mechanisms for BWR piping were thermal fatigue, FAC, and IGSCC. It was argued that BWR plants are more prone to thermal fatigue problems than the PWRs because they experience a greater temperature fluctuation during the normal operating cycle. In BWRs, thermal fatigue is a concern for the feedwater lines, the main steam lines, and the RHR system. From a LOCA perspective, thermal fatigue is an important aging mechanism because it does not manifest itself as a single crack, but as a family of cracks over a wide area. As such, it can lead
to a large LOCA. Thermal fatigue cracks also tend to propagate rapidly, and since it is not material sensitive (i. e., it can attack a number of materials), it is difficult to prioritize critical areas for inspections.

Only the feedwater piping system is highly susceptible to FAC. The main steam line is the other major carbon-steel piping system which experiences constant fluid flow. However, it is not as susceptible to FAC because the erosion rates associated with two-phase flow are less severe. While FAC caused a serious accident in the secondary side piping at Surry 15 years ago, the panel members generally thought that the industry had inspection programs in place today to prevent the reoccurrence of such an event, especially for the primary side piping systems. However, a number of panel members expressed the concern that the water chemistry improvements which mitigate IGSCC could lead to unexpected FAC problems.

The panel consensus is that the susceptibility to IGSCC is greatly reduced compared to the past.

Measures such as improved HWC, weld overlay repairs, and pipe replacement with more crack resistant materials had essentially reduced the likelihood of IGSCC. However, there is still residual concern about the failure likelihood of the large recirculation piping material that has not been replaced. Furthermore, even for the pipe which has been replaced, the question was raised as to whether the new replaced pipe was immune to this type of degradation, or is the problem simply been move out into the future. The German experience with Type 347 stainless steel was raised in this regard. There was also concern expressed about the effects of increased sulfate levels in the future due to efforts focused at extending the life of some of the filters in the plants.

Another aging mechanism of concern is mechanical fatigue. This is primarily a problem in smaller diameter piping, especially those with socket welds, and is caused by an adjacent vibration source. From a LOCA perspective, it was noted that locations susceptible to mechanical fatigue damage were not always obvious. It is impossible to eliminate all plant vibrations, and furthermore, changing the configuration of the plant can result in newly susceptible areas.

As part of this elicitation exercise a total of 14 LOCA-susceptible piping systems were considered for the BWR plants. Of these, however, most of the participants focused on a few common systems as being the important LOCA contributors. Figure L.6 shows the Category 1 LOCA frequencies for each of these piping systems at 25 years of plant operation (present day). Note, the results for the HPCS and LPCS systems are combined as a single entry in Figure L.6 (HPCS/LPCS). For these smaller category LOCAs, the main concern is with the smaller diameter lines, such as the instrument and drain lines. Most of the participants believe that it is more likely to have a complete break of a smaller diameter line than a comparable size opening in a larger diameter pipe. One reason for this is that for a given crack size, the crack is a larger percentage of the pipe circumference in the smaller diameter pipes, and it was thought that a small diameter pipe was just as likely to have a crack of a certain length as a larger diameter pipe. Furthermore, smaller diameter lines are often fabricated from socket welded pipe which has a history of mechanical fatigue damage from plant vibrations. These lines may also be susceptible to external failure mechanisms arising from human error (e. g., damaging with equipment, such as fork trucks). Finally, these smaller diameter lines are often subject to fabrication flaws and they are typically more difficult to inspect, if they are inspected at all. In-service inspection is not routinely performed on these lines. Conversely, the larger diameter lines are inspected more rigorously and routinely.

Besides the instrument and drain lines, the recirculation and, to a slightly lesser extent, the CRD and RHR lines are also of concern, primarily as a result of SCC susceptibility.

For larger Category 3 LOCAs, the recirculation system was the largest contributor to the overall LOCA frequencies, see Figure L.7. (Note in this figure that the instrument and drain lines, as well as the CRD lines, are no longer shown in that these smaller diameter lines cannot support a Category 3 LOCA.) The fact that the recirculation system is the largest contributor is a slight departure from the PWR estimates where the smallest diameter piping system that can support a particular LOCA category consistently had the highest LOCA frequencies. The main concern with the recirculation system piping continues to be SCC, even when considering the effective mitigation programs in place today. Of secondary importance were the feedwater, RHR, RWCU, core spray, and SRV systems. There was wide variability expressed for the feedwater system. Several participants thought that its susceptibility was similar to that of the recirculation system while others thought that it would make an inconsequential contribution. This latter group generally thought that the mitigation programs in place for the feedwater system were overall effective. The RHR system was deemed important by some panel members due to the relatively larger number of precursor events reported and the relatively high number of welds. A number of the participants used the weld census data provided to differentiate the relative contributions between systems for those systems that have similar operating experience. The SRV lines were judged to be potentially problematic by four of the eight respondents who addressed the question of BWR piping. They pointed out that the SRV lines are subject to high dynamic loads during the relatively common SRV discharge events, however, only a short section of these lines are actually susceptible to a LOCA event. Overall, in comparing Figure L.6 with Figure L.7, one can see approximately a one order of magnitude reduction in the LOCA frequency between the Category 1 and 3 LOCAs for most of the BWR piping systems considered.

For the largest category BWR piping LOCAs (Category 5), the recirculation system remains the main contributor to the overall LOCA frequencies, see Figure L.8. The RWCU system had about the same median value, however, there was a question expressed as to whether the RWCU system could actually sustain such a high flow rate LOCA. One of the participants thought that the maximum diameter for this system was only 6-inches, not 24-inches as specified in the development of the elicitation questions. Besides the recirculation, and RWCU systems, the next two largest contributors to the BWR Category 5 LOCA frequencies were the feedwater and RHR systems. As for the Category 3 LOCAs, the RHR system was deemed important due to the large number of precursor events reported and the large number of potentially susceptible welds. Several of the participants indicated that these lines are susceptible to SCC

Figure L.9 is a plot of the cumulative BWR piping LOCA frequencies (including contributions from all of the piping systems) for Category 1 through 5 LOCAs. The BWR piping LOCA frequency decrease with LOCA size is relatively shallow, i. e., approximately Уг order of magnitude per LOCA category. The results tend to be governed by the results from the recirculation system. It was noted that for the recirculation system that the mitigation programs in place for controlling IGSCC promote a more uniform residual stress field which can in turn promote longer cracks which are more likely to cause a LOCA.

This effect will potentially offset the overall reduction in crack growth due to the mitigation program. It is also of note from Figure L.9 that the variability in the results as expressed by the interquartile range and the difference between the minimum and maximum values does not vary much with LOCA size. It is also of note that the expert ranking is relatively consistent with LOCA size, i. e., Participant C always predicted the highest LOCA frequencies and Participants E and G consistently predicted the lowest LOCA frequencies.

Figure L.10 shows the effect of operating time on the cumulative Category 1 LOCA frequencies for BWR piping systems. As can be seen in Figure L.10, there is not much of an effect of operating time on the cumulative Category 1 frequency. Similar findings were evident for the larger Category 3 and 5 LOCAs. Obviously, any unabated degradation mechanism would cause an increase in the overall LOCA frequencies. However, it was generally assumed by the panel members that any new degradation mechanism that came on the scene would be aggressively met by the industry and NRC, just like the IGSCC problem in BWRs was met in the past and the PWSCC problem in PWRs is being met today. The minimal changes in LOCA frequencies with time evident in Figure L.10 were the result of a number of compensating factors considered by the panel members. From the perspective of potential decreases in the LOCA frequencies, the recirculation lines should see a decrease in the LOCA frequencies with respect to the current-day estimates that are based on an analysis of operational experience data due to improved mitigation strategies that have been put in place. The panelists generally felt that the IGSCC issue for BWRs had been effectively mitigated for the foreseeable future. In addition, the core spray systems may see a decrease in the LOCA frequencies with time as the segments of stainless steel piping potentially susceptible to IGSCC are replaced with carbon steel piping. Finally, future inspection and mitigation programs are expected to lead to additional decreases in the predicted LOCA frequencies. In this regard, having the industry focus its inspection resources on the more important systems through risk-informed ISI should help reduce the propensity for LOCAs. Counteracting these potential decreases are potential increases due to bigger thermal fatigue and FAC concerns in the future. Concern was expressed about the high usage factors that will exist near the end-of-plant license. Also, there is the concern with new, previously unknown degradation mechanisms that may arise in the future. In this regard, the inspection methods of today may not be reliable for these new mechanisms. Furthermore, these new mechanisms may not manifest themselves in the same locations of concern today. Finally, while timely and proper maintenance programs are always beneficial, there are instances in which they may prove counterproductive. The frequent opening and closing of systems for inspections increases the likelihood

for human error such as having tools and other debris left behind or bolts not being torqued properly. Also, improper service of active components (e. g., valves) can lead to passive system failures.

Figures L.11 and L.12 show the cumulative MV estimates, along with the 5% and 95% bound values for the various participants for the Category 1 and 3 LOCAs, respectively. The uncertainty range (difference between 5% LB and 95% UB values) for the Category 3 LOCAs are comparable (or slightly greater than) for the Category 1 LOCAs. Only participants A, E, and F expressed considerably more uncertainty for the Category 3 LOCAs than they did for the Category 1 LOCAs. Similar findings were found when comparing the Category 5 results with the Category 3 results. Overall, the panelists appeared more confident about their BWR estimates than they did for the corresponding PWR estimates. They had less uncertainty about future and bigger size LOCA frequencies compared with their PWR predictions. There was also less uncertainty among the panelists about the magnitude of the dominant contributing factors.

In addition, the panel members used more consistent approaches and more consistent base case estimates for the BWR estimates than they did

Figure L.14 shows the breakdown of PWR Category 3 LOCA frequencies by piping system at 25 years of plant operations (present day). The small diameter instrument and drain lines, as well as the RH lines, do not appear on this figure in that they are of such size that they could not sustain a Category 3 LOCA. Again, as was the case for the PWR Category 1 LOCAs, the smallest diameter lines that can sustain this size (i. e., category) of LOCA are the dominant contributors. These include the CVCS, SIS-DVI, RHR, surge, and PSL. This is different than what was observed for the BWR Category 3 LOCAs where the larger recirculation system was the dominant contributor, primarily due to its susceptibility to IGSCC.

The two most listed systems as being major contributors to this category of LOCA for PWR piping were the CVCS and SIS-DVI lines. For both, the primary concern was fatigue. One participant commented that the CVCS line was one of the most fatigue sensitive locations in the entire plant. Another commented that they were concerned with environmentally-assisted fatigue for this system. With regard to the SIS-DVI (and the SIS-Accumulator lines for that matter), several participants indicated that both lines had experienced thermal fatigue cracking in the past due to cold water leaking past the check valves. Another line that a number of participants thought would be a major contributor to this category of LOCA was the RHR lines. The concern with these lines was with environmental attack due to the stagnant nature of the flow in these lines. The pressurizer spray lines were of a concern due to the chance for PWSCC at one of the bimetal welds.

n = 9

1e-13 1 e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2

LOCA Frequency (yr-1)

Figure L.14 Category 3 LOCA Frequencies for PWR Piping Systems at 25 Years of Plant

Operation

For the largest categories of PWR piping LOCAs (Categories 5 and 6), the hot leg, cold leg, surge line, and RHR lines all contribute to the overall LOCA frequencies, see Figure L.15 for the Category 5 LOCAs. Of these, the median value of the LOCA frequency for the cold leg is about a half order of magnitude less than the median values for the other three piping systems. This slight reduction is primarily due to the fact that the cold leg is less susceptible to PWSCC than either the hot leg or surge line at this time (25 years of plant operations) due to the fact that it operates at a slightly lower temperature. Somewhat surprisingly in examining Figure L.15, a number of the participants felt that the hot leg would have a greater propensity for a Category 5 LOCA than the surge line. Both lines are susceptible to PWSCC due to the presence of bimetallic welds and the high operating temperatures, but the surge line was also judged to be susceptible to thermal fatigue due to thermal stratification and thermal striping stresses. Also, the surge line is smaller diameter, which based on the thought that smaller diameter lines are more prone to LOCAs than their larger counterparts, would imply that the Category 5 LOCA frequencies for the surge line should be higher. Finally, at least one participant argued that the surge line to pressurizer bimetallic weld was one of their biggest concerns in the entire plant due to its susceptibility to PWSCC and the fact that it is a very difficult weld to inspect. Counteracting these arguments, however, is the fact raised by a number of the participants that there are more hot leg to RPV bimetal welds (2 to 4 depending on the number of loops) in a plant than there are surge line to pressurizer bimetal welds (one).

LOCA Frequency (yr-1)

Figure L.15 Category 5 LOCA Frequencies for PWR Piping Systems at 25 Years of Plant

Operation

Figure L.16 is a plot of the cumulative PWR LOCA frequencies at 25 years of plant operation. Cumulative frequencies are shown for Category 1, 3, and 6 LOCAs. Based on a review of Figure L.16 there appears to be approximately a one order of magnitude reduction in LOCA frequency between each successive LOCA category.

——— 1——- 1——- 1——- 1——- 1——- 1——- 1—— 1——— 1——- 1——- 1——- 1——-

1e-15 1e-14 1e-13 1e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2

PWR LOCA Frequencies at 25 Years (CY-1)

Figure L.16 Cumulative PWR LOCA Frequencies at 25 Years of Plant Operations

Figure L.17 shows the effect of operating time on the cumulative Category 1 LOCA frequencies for PWR piping systems. Several participants felt that the operational experience is sufficient to expect the frequencies to remain relatively constant out to 60 years of life. Degradation and aging will naturally continue to occur. However, the inspection and mitigation strategies will effectively identify and temper the frequency increases caused by this aging. Some panelists expected a short term frequency increase due to PWSCC before effective mitigation is developed. This trend is consistent with the historical response to evidence of emerging degradation by the industry. Also, at least one participant expressed a concern about the high usage factors that will exist at 60 years at many locations. All of these concerns are reflected in the results showing the effects of operating time and aging in Figures L.17 and L.18 for Category 1 and 3 LOCAs, respectively. As can be seen in Figures L.17 and L.18, there is a slight increase in the cumulative Category 1 and 3 LOCA frequencies between 25 and 40 years, but not much of an effect between 40 and 60 years. The median LOCA frequencies for the Category 1 and 3 LOCAs at 40 years are an order of magnitude higher than the median LOCA frequencies for the Category 1 and 3 LOCAs at 25 years. Similar findings were evident for the larger Category 6 LOCAs. The rationale behind this is that this size of LOCA (and associated pipe size) is most affected by aging. These pipes are not as easily inspected, or as leak sensitive, as their larger counterparts and these pipes have not experienced the infant mortality as their smaller counterparts.

1e-7 1e-6 1e-5 1 e-4 1e-3 1e-2 1 e-1 1e+0 1e+1

PWR Category 1 LOCA Frequencies(CY’1)

Figure L.17 Effect of Operating Time on the Cumulative Category 1 LOCA Frequencies for PWR

Piping Systems

1e-8 1e-7 1e-6 1e-5 1 e-4 1e-3 1e-2

PWR Category 3 LOCA Frequencies(CY-1)

Figure L.18 Effect of Operating Time on the Cumulative Category 3 LOCA Frequencies for PWR

Piping Systems

Solid diamonds are 5% Lower Bound, Mid-Value, and 95% Upper Bound Values A o—o^-o * _ * B 1 C ♦——————- ♦————- ♦ *

H I . 0 ** " J L

Figures L.19 and L.20 show the cumulative MV estimates, along with the 5% and 95% bound values for the various participants for the Category 1 and 3 LOCAs, respectively. The uncertainty range (difference between 5% LB and 95% UB values) for the Category 3 LOCAs are typically greater than for the Category 1 LOCAs for most of the participants. In a similar vein, the level of uncertainty for the Category 6 estimates were much greater than for the Category 1 or 3 estimates, see Figure L.21. All of the panelists had at least two orders of magnitude difference between the LB and UB values for their Category 6 estimates, and some of the panelists (C, E, and J) had greater than four orders of magnitude difference.

1.00E-17 1.00E-15 1.00E-13 1.00E-11 1.00E-09 1.00E-07 1.00E-05 1.00E-03 1.00E-01

PWR Category 1 LOCA Frequencies (CY-1)

Figure L.19 PWR Category 1 LOCA Frequencies Showing MVs, 5% LB, and 95% UB Values for
All Participants Who Responded to the PWR Piping Questions

♦—{——— ♦ G H ‘ * . . I J L

Solid diamonds are 5% Lower Bound, Mid-Value, and 95% Upper Bound Values A * ♦—Ц—- ♦ b C

1.00E-17 1.00E-15 1.00E-13 1.00E-11 1.00E-09 1.00E-07 1.00E-05 1.00E-03 1.00E-01 PWR Category 3 LOCA Frequencies (CY-1)

Figure L.20 PWR Category 3 LOCA Frequencies Showing MVs, 5% LB, and 95% UB Values for All Participants Who Responded to the PWR Piping Questions

Solid diamonds are 5% Lower Bound, Mid-Value, and 95% Upper Bound Values о———- о—о A * ♦—————— ♦ B * ♦——————————————- ♦ C

E

G

j

L

1.00E-17 1.00E-15 1.00E-13 1.00E-11 1.00E-09 1.00E-07 1.00E-05 1.00E-03 1.00E-01

PWR Category 6 LOCA Frequencies (CY-1)

Figure L.21 PWR Category 6 LOCA Frequencies Showing MVs, 5% LB, and 95% UB Values for All Participants Who Responded to the PWR Piping Questions

In general, the results for PWR piping appear consistent. The quantitative results and the qualitative opinions and rationale were for the most part consistent. The variability between participants stems from the different approaches used and the basis for their estimates. Several different approaches with different anchoring points were used by the different panelists. The variability between the participants seems reasonable given the frequency magnitudes being computed.