The same three major non-piping components (RPV, valves, and pumps) as considered for BWRs are considered for PWRs, plus the steam generator and pressurizer are added. One of the operational experience databases showed an order of magnitude higher incident rate for PWR non-piping than BWR non-piping. This was partially attributed to the fact that there are more PWRs than BWRs. However, this comparison is also biased by the large number of steam generator tube failures reported in the databases. Steam generator tubes are subjected to a host of degradation mechanisms: fatigue, denting, external SCC, PWSCC, and overload failures. It was almost universally accepted that SGTRs would be the dominant contributor to the PWR Category 1 non-piping LOCA frequency. In fact the PWR steam generator tube failure frequency is the dominant contributor to the overall PWR small-break LOCA (Category 1 LOCAs) frequency when considering both the piping and non-piping contributions. Even so, it is the expectation of a number of the panel members that the steam generator tube contribution to the small-break LOCA frequency will decrease with time due to steam generator tube replacement programs and improvements made to the secondary side water chemistry.

In general, many of the same degradation mechanisms that are important for PWR piping are important for the non-piping components as well. PWSCC is an important degradation mechanism for many of the smaller Alloy 600 components such as the CRDMs, heater sleeves, steam generator tubes, and other penetrations. As was the case for piping, the likelihood of multiple cracks forming, and possibly coalescing, and the relatively fast propagation rates associated with this type of cracking makes this mechanism a major concern from a LOCA perspective. Also, since this mechanism is more severe at the higher temperatures associated with PWRs, it is considered to be a bigger threat for the PWRs than the BWRs, at least in the short term. As was the case for PWR piping, thermal fatigue is also a concern for PWR non-piping components. It is especially of concern at nozzle inlets and other locations where thermal stratification may exist. Furthermore, for all the reasons highlighted for piping, thermal fatigue is a mechanism that can lead to large leaks, i. e., fast propagation rates, attacks a wide area, and difficult to prioritize inspection protocols due to the fact that it can attack a variety of materials. Mechanical fatigue is another common degradation mechanism to both PWR piping and non-piping components.

Mechanical fatigue is most important for smaller components, such as heater sleeves and small penetrations that are subjected to vibratory stresses due to equipment operation.

Another mechanism of special concern to non-piping components is common cause bolting failures. This is especially relevant to manways and bolted valves. The common cause mechanism could be improper installation or maintenance of bolts, e. g., improper torque, external corrosion of multiple bolts, or possibly steam cutting of multiple bolts.

Also, boric acid corrosion of carbon steel components such as RPV and steam generators can be aggressive under certain conditions.

Figure L.31 shows the Category 1 LOCA frequencies for the major PWR non-piping components at 25 years. As expected, the expected failure frequencies are highest for the steam generator. The higher LOCA frequency for the steam generator is driven by the SGTR data.

<V§C

SG

B

G

PRESS

C

G

Valves

C

The other major contributors to the Category 1 LOCA frequencies for PWR non-piping were the RPV and the pressurizer. The main subcomponent contributing to the RPV frequency is the CRDMs while the main subcomponent contribution to the pressurizer frequency is the heater sleeves. For Category 2 LOCAs, a single SGTR cannot sustain such a leak. Thus, for the Category 2 LOCAs, the CRDM and pressurizer heater sleeves became the main contributors.

For Category 3 and 4 LOCAs there was no consistent agreement among the panelists as to the major contributors. As one can see in Figure L.32, all five major components contribute fairly equally to the Category 4 LOCA frequencies. As such, there is tremendous variability about the frequency associated with each component. This variability was also apparent for the Category 6 LOCAs. This variability reflects the inconsistent opinions and approaches followed by the panelists, as well as the difficulty of this type of assessment. As is to be expected, there was a wide array of possible failure modes for dissimilar components to be considered, and the panelists tended to gravitate towards a few of the failures that they personally thought were most credible. Given all of this, the level of variability was thought to be reasonable given the event frequencies. This was one reason for adopting the elicitation approach in the first place. The highest LOCA frequencies were for the pressurizer nozzle. In addition, many of the panelists considered manway or shell failures important, irrespective of the component type. Thus, they anticipated similar distributions for both the steam generator and pressurizer. There were also major differences of opinion among the panelists as to the most important failure modes.

Figure L.33 shows the cumulative LOCA frequencies for the PWR non-piping components at 25 years of plant operations. The Category 1 LOCA frequencies for PWR non-piping are the highest frequencies estimated by the elicitation panel for piping or non-piping, BWR or PWR. The median frequency is almost 5×10-3. The variability among the panelists was very small. The difference between the minimum and maximum predictions was less than an order of magnitude. These high frequencies and low variability were driven by the SGTR data for which ample data exist in the operational experience databases; thus explaining both the high frequencies and excellent agreement between participants. For the Category 2 LOCAs, the agreement, at least on a minimum and maximum basis, is not nearly as good. However, the agreement on the basis of the spread in the IQR is nearly as good as it is for the Category 1 LOCAs. Again, for the Category 2 LOCAs, the major contributors are the CRDMs and the pressurizer heater sleeves. The much wider variability for the Category 3 through 6 LOCAs reflects differences in opinion among the panelists as to the important failure modes and their associated frequencies.

H G*#’H

1e-141e-13 1e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1

LOCA Frequency (CY-1)

Figure L.33 Cumulative PWR Non-Piping LOCA Frequencies at 25 Years of Plant Operations

Figure L.34 shows the effect of time on the PWR non-piping Category 1 LOCA frequencies. There is a very slight decrease in the frequency between 25 years (present day) and 40 years (end of plant license) due mostly to steam generator replacement programs and improved inspection and mitigation programs, e. g., improved eddy current inspection programs and improved secondary side water chemistry. There was also an expected decrease in the LOCA frequencies associated with CRDMs due to on-going head replacement programs and improved CRDM inspection programs that may go into effect over the next few years. However, there was some concern expressed that the maintenance and inspection programs for the larger component bodies (pressurizer, steam generator, RPV) may not be as rigorous as for the piping systems. Figure L.35 shows the effect of operating time on the PWR non-piping Category 6 LOCA frequencies. As can be seen in Figure L.35, the median values remain constant with time and the variability among the participants (at least on the basis of the IQR) also remains fairly constant. This tends to indicate that the participants did not foresee any significant aging effects to occur.

1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0

LOCA Frequency (CY-1)

Figure L.34 Effect of Operating Time on the Cumulative Category 1 LOCA Frequencies for PWR

Non-Piping Components

1e-14 1e-13 1e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5

LOCA Frequency (CY-1)

Figure L.35 Effect of Operating Time on the Cumulative Category 6 LOCA Frequencies for PWR

Non-Piping Components

Contrary to what was observed for the Category 1 and 6 LOCA frequencies, the median value of the Category 4 LOCA frequencies increases an order of magnitude between 25 and 40 years and then remains constant after that, see Figure L.36. As was the case for PWR piping, aging was thought to have the largest impact on LOCA Categories 3 and 4. It was thought by some that aging could accelerate near the end of the plant license faster than the effects of mitigation and inspection could become effective, especially if the plant operators do not see a return on their investment for such inspection and mitigation programs near the end of the plant’s license.

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

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

LOCA Frequency (CY-1)

Figure L.36 Effect of Operating Time on the Cumulative Category 4 LOCA Frequencies for PWR

Non-Piping Components

Figures L.37 and L.38 show the cumulative MV estimates, along with the 5% and 95% bound values for the various participants for the PWR Category 1 and 4 non-piping LOCA frequency estimates, respectively, at 25 years of plant operating time. Of note from these figures is higher uncertainty among almost all of the participants for the Category 4 LOCAs when compared with the Category 1 LOCAs. A number of the panelists showed 3 to 4 orders of magnitude of uncertainty for the Category 4 LOCAs while all of the panelists had less than approximately 2 orders of magnitude of uncertainty in their Category 1 results. In addition, the variability in the panelist’s MV estimates was within 1 order of magnitude for their Category 1 results while the variability among their results spread over a range of almost five orders of magnitude for their Category 4 results. The fact that the agreement among the panelists was so good for the Category 1 predictions plus the low uncertainty of their individual predictions is a reflection that there was near consensus agreement that the single overwhelming dominant contributor to this class of LOCAs was SGTRs, for which ample field experience is available in the operational experience databases.

/

/

Lh*h

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

10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100

LOCA Frequency (CY-1)

Figure L.37 PWR Non-Piping Category 1 LOCA Frequencies Showing MVs, 5% LB, and 95% UB
Values for All Participants Who Responded to the PWR Non-Piping Questions

10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100

LOCA Frequency (CY-1)

Figure L.38 PWR Non-Piping Category 4 LOCA Frequencies Showing MVs, 5% LB, and 95% UB
Values for All Participants Who Responded to the PWR Non-Piping Questions