ML20137J703

SER Supporting Util Identification of Root Causes for Source Range Nuclear Instrumentation Channels Inoperability During 850609 Steam & Feedwater Line Rupture Control Sys Low Level Actuation & Corrective Actions Taken.Plant Restart Approved

ML20137J703
Person / Time
Site:	Davis Besse
Issue date:	01/08/1986
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20137J676	List: ML20137J671 ML20137J689 ML20137J703
References
NUDOCS 8601230158
Download: ML20137J703 (7)

Text

. .. . . _

.~ -.

Enclosure 2 SAFETY EVALUATION REPORT DAVIS-BESSE INCIDENT OF JUNE 9, 1985 TROUBLESHOOTING REPORTS ON SOURCE RANGE NUCLEAR INSTRUMENTS

INTRODUCTION i

i Source range nuclear instrument channel NI-1 (referred to also as Channel 2) was inoperable prior to and througgout the June 9,1985 incident, in that it read full downscale (less than 10 counts /sec.). During the event when the neutron level, as indicated by the intermediate range nuclear channels, fell to a predetermined level, the source range nuclear channels were activated. However, i- remained at less than I the redundant 10~1 counts /sec.NI-2 (referred rather to also asabout than indicating Channel 10 g) counts /sec.This loss of both nuclear channels was an unnecessary problem that to the already-busy reactor operator had to cope with. He was required to verify shutdown margin requirements, which included initiating emergency boration.

Problems with this instrumentation have been chronic, with some present since l initial construction of the plant. Because of the attention given to the June 9, 1985 incident, the licensee prepared systematic and thorough troubleshooting plans

! and has now submitted a findings report for each channel. Specifically, Findings Reports Nos. 15A and I5B (Revision 2 of each) are included in the licensee's

! Supplement No. I to the report " Davis-Besse Course of Action." Review of these reports raised a number of questions in two general categories: open questions and bases for deferral of corrective actions. These questions were discussed with the licensee on October 30, 1985 and on November 6, 1985. The licensee had

reasonable answers to the open questions and had reasonable bases for action

! deferrals; however, these had not been included in the reports. Accordingly, the 4

licensee prepared Revision 3 for each of the NI findings reports, which are the objects of this evaluation. '

j EVALUATION We prepared Tables 1 and 2 from the licensee's finding reports, to list the many

anomalies that were discovered via the systematic t*oubleshooting plans for NI-1

and NI-2, respectively. The number of anomalies for NI-1 and for NI-2 constitute prima-facie evidence of lack of proper maintenance of this nuclear safety-related equi'pment. Furthermore, many of the anomalies originated from installation errors during the original construction of the plant that went uncorrected for eight years j of plant operation.

Such a long list of anomalies is clearly sufficient to have caused the malfunctions related to the June 9,1985 incident. We note that the lists have some general characteristics:

1. A large number of the anomalies were related to the preamplifier j assembly.

J'

2. A large number of the anomalies were related to triaxial connectors, both

, bulkhead mounted and on cables.

860123015e 86010e PDR ADOCK 05000346 i S PDR i

TABLE 1 TROUBLESHOOTING FINDINGS NI-1 (CHANNEL N0. 2)

1. DETECTOR ASSEMBLY (3 Anomalies) 1.1 Improper assembly of triax Amphenol connector which interfaces the integral mineral-insulated detector cable to the triax cable to/from the preamplifier.

1.2 Detector location in thimble not a core midplane.

1.3 Masonite spacers used to block area around detector signal cable in detector thimble plug were too short (i.e., 5 inches vs. 18 inches).

2. PREAMPLIFIER ASSEMBLY (7 Anomalies) 2.1 Detector cable connector had its center pin pushed in approximately 1/4" and off center.

2.2 No grounding wire was connected to outer preamp box.

2.3 Bulkhead connectors on preamp inner and outer boxes had high-resistance connections to triax shields, due to connectors being mounted on painted surfaces.

2.4 None of the cable connectors at the preamp had 0-rings installed.

2.5 Detector cable bushing on outer box had inadequate clearance from detector cable connector, causing a j potential ground loop.

2.6 Detector and high voltage connectors appear to be

! nickel instead of silver. .

2.7 Fiber shipping washers had been left in some bulkhead

, connectors, preventing proper meshing and tightening

} of connectors.

3. PENETRATION ASSEMBLY (No Anomaliesi

4. RPS INSTRUMENT CABINET (4 Anomalies) 4.1 Loose connection to station safety ground bus.

l i

1

4.2 Output connector for high voltage power supply had a crushed 0-ring.

4.3 Fiber shipping washers had been left in some bulkhead connectors, preventing proper meshing and tightening of connectors.

4.4 " Blue Ribbon" connector on high voltage power supply was chipped and cracked.

5. GENERAL (3 Anomalies) 5.1 Operation of instrument cabinet door switches (provide annunciator indication of open door) for RPS, SFAS, and SFRCS cabinets cause high level spikes at input to count rate amplifier module.

5.2 Every Amphenol connector was tarnished; many Amphenol connectors contained metal flakes.

5.3 Operation of some SFAS-controlled motor operated valves caused some spiking observable at the input of the rate-of-change amplifier module.

{

TABLE 2 TROUBLESHOOTING FINDINGS NI-2 (CHANNEL NO. 1)

1. DETECTOR ASSEMBLY (1 Anomaly) 1.1 Leaking seal plate allowed rust to form at connection box at top of detector thimble.

2. PREAMPLIFIER ASSEMBLY (9 Anomalies) 2.1 Low voltage cable connector loose on cable.

2.2 High voltage bulkhead connector extremely loose --

0-ring not installed.

2.3 High voltage cable connector did not have 0-rina installed.

2.4 Ground wire to outer box not installed.

2.5 Connectors for high voltage and detector appear to be nickel instead of silver.

2.6 Bushings not installed where cables enter outer box.

2.7 Bulkhead connectors on outer box for detector and high voltage were loose, i .e. , mounting nuts only finger tight.

2.8 Printed circuit board was not mounted securely inside preamplifier box.

2.9 Shipping washers were left in some bulkhead connectors, which prevented adequate tightening of connectors.

3. PENETRATION ASSEMBLY (3 Anomalies) 3.1 Resistance substantially high for signal cable.

3.? Signal cable connector loose on cable.

3.3 Shipping washers were left in some bulkhead connectors, which prevented adequate tightening of connectors.

3.4 Intermittent losses of continuity for center conductor of signal cable.

l i

l

4. RPS INSTRUMENT CABINET (3 Anomalies) 4.1 Coax connector to count rate amplifier not locked.

4.2 High voltage cable connector appears to be nickel rather than silver.

i 4.3 High voltage cable connector did not have 0-ring installed.

5. GENERAL (3 Anomalies) 5.1 Cabinet door switches (for annunciator) for RPS, SFAS, and SFRCS cabinets cause high level spikes.

5.2 Every Amphenol connector tarnished; many contained metal flakes.

'3. 3 Some SFAS-operated MOVs cause spiking on input to SUR meter.

i I

I I

3. The NI-2 channel containment penetration, which is generally considered to be a relatively passive component and hence not as likely to fail as more active components, had serious anomalies. (See Item 3 of Table 21.

These characteristics suggest that components that are either cumbersome to get to or not very likely to fail cannot be neglacted. Further, the characteristics suggest a generalized problem with triaxial connectors.

(We note that the licensee is addressing the generalized connector problem with improved procedures and training).

The licensee identified the following root causes for the failures of NI-1 and NI-2, respectively:

NI-1:

• High resistance connections in the bulkhead connectors on the preamp, due to being mounted on painted surfaces (Item 2.3 of Table 1).

Improper assembly of the triax connector at the detector interface (Item 1.1 of Table 1).

NI-2:

High resistant and intermittent connections related to the containment penetration assembly (Items 3.1 and 3.4 of Table 2).

Generalized poor condition of connectors, due to improper assembly, lack of proper cleaning, and poor maintenance (Item 5.2 of Table 2).

We view these items as the licensee's determination of the most significant of the anomalies discovered. In our experience, only rarely is a single factor the cause of problems in pulse-type nuclear instrumentation. Most often many factors contribute to the problem, in varying degrees of severity.

Our review of the discovered anomalies indicates that there is reasonable assurance that most, if not all, significant contributors have been identified.

The basis for our conclusion is very pragmatic. Subsequent to the corrective actions that have already been completed, the channels have been monitored continuously for substantial periods of time, with no further instances of problems. For NI-1 this period was six weeks; for NI-2, eight weeks.

The licensee's reports state that although most corrective actions will be completed prior to plant restart, certain corrective actions are not major contributors to the problems and thus are planned to be completed after plant restart. We have reviewed the specific basis provided for these deferrals.

In view of the substantial periods of proper operation of the instrumentation channels without problems, we find the deferral of additional corrective actions to be acceptable.

e CONCLUSION The systematic and thorough troubleshooting plans for the source range nuclear instrumentation channels have revealed a substantial number of causes. The most significant causes appear to be improper installation of triaxial connectors on the preamplifier (for NI-1), intermittent high-resistance connections in a containment penetration assembly (for NI-2), and a generalized poor condition of triaxial connectors. Our review of the licensee's findings reports concludes that the significant contributors to the problems have now been identified and that the corrective actions have been effective in improving the performance of these instrumentation channels. Accordingly, as far as the sources range nuclear instrumentation is concerned, plant restart and power operation is now acceptable.

Dated: January 8,1986 Principal Contributor: J. T. Beard l

1

.