ML20005F015
| ML20005F015 | |
| Person / Time | |
|---|---|
| Issue date: | 01/02/1990 |
| From: | Dennig R NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20005F014 | List: |
| References | |
| TASK-AE, TASK-T90-01, TASK-T90-1 AEOD-T90-01, AEOD-T90-1, NUDOCS 9001120182 | |
| Download: ML20005F015 (2) | |
Text
y
-i i
j.
PIOPrietary Information Attached e
_AE0_D TECHNICAL REVIEW REPORT UNIT:
HULTIPLE TR REPORT NO.: AEOD/T90-01 DOCKET NO:
MULITPLE DATE: January 2, 1990 LICENSEE:
MULTIPLE EVALUATORS: R. Dennig T. Wolf SUMM_AR,Y:
All of the PN0s issued in the first quarter of 1989, approximately 100, were reviewed for component level events that were potentially reportable to the Nuclear Plant Reliability System (NPRDS). The PN screening process was performed by engineers familiar with NPRDS conventions. Where corresponding failure records were not found in the NPRDS, LERs were reviewed (if one was issued) for additional detail to aid in the determination of reportability.
A total of 18 cases were identified where we would expect one or more NPRDS failure reports to be filed; in 15 of these cases matching reports were found in the system. This small sample then indicates that after about 9 months following discovery date approximately 80 percent of the failures are in the system.
This estimate is a little higher than past estimates of between 65 and 75 percent, but is consistent with these earlier results.
PNs are designed to provide "early notice of events of possible safety or public interest significance." Hence, they cover a range of topics beyond equi > ment failures, such as drug testing, management changes, plant milestones l
reacied (e.g., initial criticality), and non-reactor licensee events such as I
lost sources, and transportation incidents. Thus, the relatively low yield of potentially reportable component events, 18 of 100, is not surprising.
DISCUSSION:
All of the PN0s issued in the first ovarter of 1989, approximately 100, were reviewed for component level events that were potentially reportable to l
Extensive guidance on reportability is provided in the NPRDS Reporting Guidance Manual and the four NSS-specific volumes entitled Reportable System and Component Scope Manual.
Component level events described in PNs may not be reportable to NPRDS for the following reasons:
(1) The event may not meet.the NPRDS definition of failure, which comprises failure to perform intended function and the need for repair to reverse component condition and restore function.
Examples of non-reportable situations are inoperability declared due to errors in analysis or absence of equipment environmental qualification, and valves or switches that have been mispositioned due to human error.
(2) The degree of the loss of function may not be severe enough to NPRDS works with three degrees of failure:
i require rep (orting. total loss of function); degraded (some function still I
immediate I~einains but the component is degraded to the point that it cannot meet functional specifications); and p cipient (the component is 9001120182 900104 PDR ORG NEXD
}.
l 4l
,1
?
O 2
-i still functioning at an acceptable level but shows symptoms of i
impending loss of function). Attachment 1 provides examples to illustrate these degrees of severity.
Under NPRDS rules, irmediate and degraded f ailures must be reported, but incipient
~
failure reporting is optional.
s (3) The component involved may not be covered by the reportable scope.
For exaniple, at the present time the main turbine is not in the scope, so EHC component failures are not repcrtable (The main i
turbine is scheduled to be 6dded to the scope in 1990). As another example, valves and pipes less than 1 inch in size are not reportable even if they are in a report 6ble system.
The component level events listed in Table 1 (PROPRIETARY) were selected from the PNs for the first quarter of 1989. All of the component level events described in the PNs and judged to be reportable using the level of detail available in the PN are included; several non-reportable cases are included in the table to illustrate various reasons why events in PNs may not be found in NPRDS. Not all non-reportable cases are listed.
For example, if the PN involved a failure in the turbine EHC, it was not listed since the reviewers knew it to be out of scope. The PN screening process was performed by engineers familiar with NPRDS conventions. Where corresponding failure records were not found in the NPRDS, LERs were reviewed (if one was issued) for additional detail to aid in the determination of reportability.
PNs are designed to provide "early notice of events of possible safety or public interest significe.nce." Hence, they cover a range of topics beyond equipment failures, such as drug testing, management changes, plant milestones reached (e.g.,initialcriticality),andnon-reactorlicenseeeventssuchas lost sources, and transportation incidents. Thus, the relatively low yield of potentially reportable component events in Table 1,18 of 100, is not surprising.
RESULTS:
Table 1 lists a total of 18 cases where we would expect one or mere NPRDS failure reports to be filed; in 15 of these cases matching reports were found in the system. This small sample then indicates that after about 9 months following discovery date approximately 80 percent of the failures are in the system.
This estimate is a little higher than past estimates of between 65 and 75 percent, but is consistent with these earlier results, t