Morning Report for January 9, 2004

Headquarters Daily Report

• REGION III

REGION III

Part 21 - 2003-0029-00 - Failure of Solenoid Operated Valve to Close

Licensee/Facility: AUTOMATIC VALVE AUTOMATIC VALVE NOVI, Michigan Dockets: License No:

Notification: MR Number: H-2004-0001 Date: 12/16/2003

Source Document: Fax

Reviewer: FOSTER, JACK W

Discussion:

Part 21 - 2003-0029-00 - Failure of Solenoid Operated Valve to Close (EN 40393)

On 10/16/2003, Dresden reported that a solenoid operated valve (B7122-145, serial number 64065 shipped on 12/11/2002) failed to exhaust air with both solenoids de-energized causing the control rod drive (CRD) valve to remain in the open position. The next day, this valve was sent back to Automatic Valve (the vendor) for analysis and inspection.

The valve was first functionally tested in what the vendor considered a normal manner such that 35 psig was low pressure and 145 psig was high pressure. The valve functioned without problems at low pressure but failed to return to its normally closed position at 145 psig when both solenoids were de-energized. Further functional testing revealed that at an inlet pressure of approximately 100 psig the valve would fail to return when de-energized.

When the valve was disassembled by the vendor, the plunger in the number 2 solenoid (the left hand solenoid when facing the exhaust port) was found to be approximately .020 too long. The actual length was 1.315 inches compared to the specified length of 1.290 inches +.005/-.003. Because the plunger only has a total stroke of less than .030, the natural expansion of the seal material in the plunger (due to a combination of heat and pressure) as well as the combined out of specification length created a situation where the plunger had no room to move and thus pilot air was exhausted when the solenoid was de-energized.

The specific root cause as provided by the vendor of this failure is the out of spec plunger. Procedures require the lengths of all plungers to be inspected prior to use. In addition to the measurement variation, the primary root cause of the observed plunger dimensions is changes to the length of the fluorocarbon insert after it is pressed into the plunger body. This variation is detectable and beyond measurement uncertainty.

Dissection of returned plungers revealed abnormal compression set among all plungers which were beyond specification limits and normal compression set among plungers which were within specification limits. From chemical analysis the vendor concluded that there was a cure problem with a portion of the suspect lot. The vendor also added that post cure takes place when the solenoid is continuously energized. Based on these test results, the vendor believes that a portion of the lot was not properly post cured.

The degree of growth of the plunger insert due to compression set is observed to be variable. This depends on exact dimensions of the molded insert when installed, the ID of the plunger retaining the insert and the exact cure duration variation of the bad plunger inserts. The degree of growth also seems to be terminal. Returned plungers (field failure) and other samples from the suspect lot on hand) at Automatic Valve have not changed dimensions during the period of study. Five plungers tested at 200 degrees F for five days exhibited thermal expansion and did not detectably change size when cooled down. Ten plungers subjected to 230 degrees F for ten days exhibited thermal expansion and did not detectably change size when cooled down.

The following plants have components containing the suspected lots as follows:
Nebraska Public Power - Cooper Station
Exelon - Limerick, Peach Bottom, Quad Cities, Dresden

Information for this issue as well as other Part 21s is available at the NRC's Operating Experience web page at: http://www.nrc.gov/reading-rm/doc-collections/event-status/

Accession Numbers:

Accession No	Accession Date
ML033530127	12/16/2003

Contacts:

Name	Office Abbrev	Phone No	E-Mail
FOSTER, JACK W	NRR	(301) 415-3647	JWF@nrc.gov