ML20148H790: Difference between revisions
StriderTol (talk | contribs) (StriderTol Bot insert) |
StriderTol (talk | contribs) (StriderTol Bot change) |
||
Line 1: | Line 1: | ||
{{Adams | |||
| number = ML20148H790 | |||
| issue date = 03/02/1988 | |||
| title = Insp Repts 50-324/88-06 & 50-325/88-06 on 880125-29. Violations Noted.Major Areas Inspected:Area of Inservice Testing of Pumps & Valves | |||
| author name = Blake J, Girard E | |||
| author affiliation = NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) | |||
| addressee name = | |||
| addressee affiliation = | |||
| docket = 05000324, 05000325 | |||
| license number = | |||
| contact person = | |||
| document report number = 50-324-88-06, 50-324-88-6, 50-325-88-06, 50-325-88-6, NUDOCS 8803300083 | |||
| package number = ML20148H775 | |||
| document type = INSPECTION REPORT, NRC-GENERATED, INSPECTION REPORT, UTILITY, TEXT-INSPECTION & AUDIT & I&E CIRCULARS | |||
| page count = 15 | |||
}} | |||
See also: [[see also::IR 05000324/1988006]] | |||
=Text= | |||
{{#Wiki_filter:r 1 | |||
l | |||
' [kQ 080g UNIVED STATES l | |||
* | |||
/ 'o NUCLEAR REGULATORY COMMISSION 1 | |||
[ n REGloN 11 ) | |||
j 101 MARIETTA STREET, N.W. | |||
S | |||
* '.. 'f ATL ANTA, GEORGI A 30323 | |||
't, go | |||
..... | |||
Report Nos.: 50-325/88-06 and 50-324/88-06 | |||
Licensee: Carolina Power and Light Company | |||
P. O. Box 1551 | |||
Raleigh, NC 27602 | |||
Docket Nos.: 50-325 and 50-324 License Nos.: DPR-71 and DPR-62 | |||
Facility Name: Brunswick 1 and 2 | |||
Inspection C du d- nuary 25-29, 1988 | |||
Inspecto : - // ' ' | |||
E B | |||
j f tE rd D' ate' signed | |||
Appro ed by _ _ f . | |||
7' & | |||
J.J.faiake, Chief Date Signed | |||
Pt ials and Processes Section | |||
j iv sion of Reactor Safety | |||
SUMMARY | |||
Scope: This routine, unannounced inspection was in the area of inservice | |||
testing of pumps and valves. | |||
Results: One violation was identified involving deficiencies in valve testing | |||
procedures and failures to follow the procedures. | |||
8803300083 830310 | |||
PDR ADOCK 05000324 | |||
O DCD | |||
.. . .. ~- , , | |||
e-- | |||
- | |||
- | |||
. | |||
. | |||
. | |||
REPORT DETAILS | |||
. | |||
1. Persons Contacted | |||
Licensee Employees | |||
S. Boyse, Project Engineer, Technical Support | |||
*R. Creech, Instrumentation and Control Supervisor (Unit 2) Maintenance | |||
*J. Crider, Inservice Inspection Responsible Engineer, Technical Support | |||
C. Dietz, Plant General Marager | |||
W. Dorman, Quality Assurance Supervisor, QA/QC | |||
*K. Enzor, Director, Regulatory Compliance | |||
*L. Jones, Director, QA/QC | |||
*W. Martin, Principal Engineer, On-Site Nuclear Safety | |||
*J. O'Sullivan, Manager, Maintenance | |||
C. Patterson, Inservice Inspection Project Engineer | |||
*R. Poulk, Project Specialist - NRC, Regulatory Compliance | |||
*J. Simon, Engineer, Operations | |||
*L. Wheatley, Inservice Inspection Project Engineer, Technical Services | |||
*E. Wilson, Engineering Supervisor (Nuclear), Technical Support | |||
*A. Worth, Engineering Supervisor (Discipline), Technical Support | |||
NRC Resident Inspector | |||
*W. Ruland, Senior Resident Inspector ' | |||
* Attended exit interview | |||
2. Exit Interview | |||
The inspection scope and findings were summarized on January 29, 1986, - | |||
, | |||
with those persons indicated in Paragraph 1. The inspector described the | |||
areas inspected and discussed in detail the inspection findings. - | |||
Dissenting comments were not received from the licensee. | |||
Item Number Status Title / Reference Paragraph | |||
325, 324/88-06-01 Open Violation - Inadequate stroke | |||
time testing of air operated I | |||
valves (Paragraph 4) | |||
Proprietary information is not contained in this report. | |||
3. Licensee Action on Previous Enforcement Matters l | |||
l | |||
This subject was not addressed in the inspection. | |||
l | |||
l | |||
l | |||
I | |||
! | |||
. | |||
- | |||
. | |||
, | |||
. | |||
2 | |||
l | |||
4. Inservice Testing of Pumps and Valves (73756) - Units 1 and 2 . | |||
The NRC inspector examined selected examples of the licensee's inservice i | |||
testing of pumps and valves to verify compliance with licensee commitments ' | |||
and regulatory requirements. Based on Technical Specification (TS) | |||
4.0.5.a.2,10 CFR 50.55a(g), and on NRC approval letter dated April 8, | |||
1987, the licensee is currently required to test their pumps and valves in | |||
accordance with ASME Section.XI (80W81). | |||
The inspector directed his examination primarily to the licensee's past | |||
testing and evaluation of the test results from four Unit 2 Containment | |||
Isolation Valves (CIVs) which failed to close in a recent actuation. In | |||
addition, the inspector examined the testing of the licensee's Loop B | |||
Residual Heat Removal (RHR) pumps and of four RHR valves that are tested | |||
along with those pumps. Details of the inspector's examinations and | |||
findings are described below. | |||
a. Containment Isolation Valves Which Failed to Close in Response to | |||
Isolation Signal on January 2,1988 | |||
(1) Background Information on CIVs | |||
On January 2,1988, the Unit 2 valves required for containment | |||
isolation of flow from the drywell floor drain and equipment | |||
drain failed to close in response to an automatic .(closure) | |||
actuation signal. | |||
Subsequently, the failure event has been extensively | |||
investigated by the licensee. Prior to the inspection described | |||
herein, NRC examination of the event had included a preliminary | |||
assessment by the Resident Inspector and a special inspection of | |||
the matter by a nine person NRC Augmented Inspection Team (AIT). | |||
This inspector reviewed background information, including the | |||
status of ongoing licensee investigation, through discussions | |||
with AIT members and licensee personnel and through review | |||
of the AIT Inspection Report (Report No. 325, 324/83-03). The | |||
following background data was noted by the inspector relevant to | |||
his examination of the testing. | |||
(a) Description of Valves and Event | |||
The failed valves are identified 2-G16-F003, -F004, -F019 | |||
and -F020. Valves 2-G16-F003 and -F004 are in the drywell l | |||
floor drain system piping and valves 2-G16-F019 and -F020 ; | |||
in the drywell equipment drain system piping. They are | |||
components of the Primary Containment Isolation System i | |||
(PCIS). The valves are three inch, 150 lb. gate valves | |||
manufactured of cast carbon steel. The valves are opened | |||
by an air actuation cylinder, controlled by Automatic | |||
l | |||
1 | |||
l | |||
1 | |||
F | |||
. | |||
. | |||
, | |||
. . | |||
3 | |||
. | |||
Switch Company (ASCO) solenoid valves, and are spring | |||
assisted closed upon loss of air pressure. The PCIS is a | |||
safety-related system that is designed to prevent or liLit | |||
the release of radioactive materials that may result from | |||
postulated accidents through rapid, automatic isolation of | |||
process lines which penetrate the primary containment. | |||
F003 and F019 are powered through relay K17 and F004 and | |||
F020 through relay K18. A single failure of either circuit | |||
to open would not cause a loss of containment integrity, as | |||
one valve in each line would still close. | |||
(b) Modification to Valves | |||
The solenoid valves which control the opening and closing | |||
of the air supply to the subject Unit 2 CIVs were modified | |||
to meet the Environmental Qualification (EQ) requirements | |||
of 10 CFR 50.49 in April 1986. | |||
Virtually identical Unit i valves (1G16-F003, F004, F019 | |||
and F020) had been similarly modified in April 1985. Due | |||
to an error, the licensee installed ASCO Model 206-832 G | |||
solenoids which fail open. Af ter consultation with the | |||
vendor, they modified the valves to fail closed by replacing | |||
the valve springs with the springs from a Model 206-832, | |||
Type "F", rebuild kit. | |||
(c) Previous History of Failures and Maintenance | |||
Neither the Unit i nor Unit 2 valves experienced any | |||
closure failures like those in the January 1988 event prior | |||
to the modification referred to in (b) above. The Unit i | |||
valves continued to operate satisfactorily following the | |||
modification, while the Unit 2 valves experienced six | |||
apparent failures. The closure tests, test failures and | |||
the post failure maintenance that was performed on the i | |||
' | |||
valves following EQ modification are listed in Table 1. | |||
Stroke time data included in Table 1 is discussed in 4.f | |||
below. The licensee identified two of the failures listed , | |||
in Table 1 as being the result of not holding the actuation l | |||
switch in the closed position for a sufficient period. ; | |||
That is, they were not considered hardware failures. ! | |||
4 | |||
However, the inspector notes that no similar problems were l | |||
identified in Unit 1 valve testing. Also, Region II | |||
specialists who have reviewed the actuation circuit | |||
consider it unnecessary to hold the switch in position. | |||
(d) Post Event Licensee Actions to Actuate CIVs l | |||
About 3 minutes after a manual scram of Unit 2, the Control | |||
; | |||
Operator (CO), who had started verifying the valve | |||
: | |||
\ | |||
__ | |||
l | |||
- - | |||
g | |||
c | |||
. | |||
. | |||
4 , | |||
isolations, noted that valves 2-G16-F003, F004, F019 and | |||
F020 had not automatically closed as required. He then | |||
attempted to manually close the four valves. 2-G16-F003 , | |||
and -F004 closed immediately vhen the handswitch was placed | |||
to CLOSE position (3 minutes af ter the scram). 2-G16-F019 | |||
and -F020 did not close on n.snual switch actuation. He | |||
cycled 2-G16-F019 and -F020 switch to CLOSE three to five | |||
times each, holding for 5 to 10 seconds in CLOSE position | |||
each time. 2-G16-F019 and F020 remained OPEN. | |||
' | |||
The Unit 1 Shift Foreman (SF) then also attempted to close | |||
valves F019 and F020 by holding their switches in the CLOSE | |||
position for more than 10 seconds. 2-G16-F019 and -F020 | |||
remained OPEN. | |||
. | |||
Approximately two minutes later (eix .iinutes after the scram) | |||
the SF (Unit 1) noted that 2-G16-F020 valve had CLOSED and | |||
that 2-G16-F019 was still OPEN. The SF held the manual | |||
switch for 2-G16-F019 in CLOSE . position for an additional | |||
10 seconds. 2-G16-F019 remained OPEN. | |||
' | |||
After another two minutes the SF noticed that 2-G16-F019 had | |||
CLOSED (eight minutes after the scram). | |||
, | |||
(e) Post Event Licensee Investigations into the Cause of the | |||
CIV Failures | |||
(i) Shortly after the event, the operability of the Unit i | |||
valves was verified through manual cycling and a logic | |||
test. The valves functioned nonnally. | |||
(ii) Various logic and other tests of the Unit 2 valves | |||
were conducted and, in each instance, the circuits and | |||
valves exhibited apparently normal actuation - save | |||
for an instance where the red (OPEN) light on F003 was | |||
slow to extinguish. The red light stayed on for | |||
almost 10 seconds and was illuminated concurrently | |||
with the green (CLOSE) light. The position lights are | |||
activated by limit switches located at the beginning | |||
and end of valve stroke travel. Actuation causes the | |||
valve disk to (1) move from the initial limit switch l | |||
- causing that position light to extinguish (2) to i | |||
between the open and closed limit switches - with no | |||
position light on, and finally (3) to the other limit | |||
switch which will cause that position light to l | |||
illuminate. Based on tests of the valves observed by I | |||
the NRC inspector in his examination of this matter, | |||
the open light should extinguish about one second after | |||
valve actuation. | |||
l | |||
l | |||
l | |||
- .- . | |||
, | |||
. | |||
. , | |||
. | |||
5 | |||
(iii) The solenoids were suspected as a likely cause of the | |||
failure because of the valve closure failures | |||
experienced subsequent to the solenoid replacements in | |||
April 1985. Thus, the solenoids have been closely | |||
checked in the post event investigation: | |||
- | |||
The licensee removed the solenoid from valve | |||
2-G16-F019 and tested it with 35 psi (normal) and | |||
5 psi air pressure. The solenoid performed | |||
satisfactorily at both pressures. The solenoid | |||
was disassembled and inspected. The inspection | |||
revealed minor debri, a light oil film in the | |||
solenoid body and s darkened area on one solenoid- | |||
stem. Bottom disk travel was found +7. mils out | |||
of tolerance. None of these conditions appeared - | |||
severe enough to explain the valve failure. | |||
- | |||
The solenoids from 2-G16-F003 and F020 were | |||
removed and inspected and no problems were , | |||
' | |||
identified, though minor debri and an oil film | |||
like that in F019 were noted. The vendor's | |||
representative participated in inspection of the | |||
F020 valve solenoid. The vendor representative | |||
did not identi fy any deficiencies in the 4 | |||
solenoid. | |||
(iv) The air cleanliness, pressure regulator and in-line | |||
filter serving F003 were checked and determined | |||
acceptable. | |||
(v) 'he logic relays K17 and K18 were inspected and | |||
widence of arcing and sticking were noted on K18. | |||
This might provide some bearing on the failure of F004 | |||
and F020 (both on K18) had it not been for the fact that | |||
when F004 closed in manual actuation following the | |||
event, the F020 valve would not close when similarly | |||
actuated. | |||
(vi) The air actuator of F020 was removed and inspected. | |||
The actuator was in good condition but contained a | |||
liberal amount of grease. In valve actuation, air | |||
passes from the actuator through the solenoid. The | |||
possibility that hydrocarbons from the grease may be | |||
reaching the solenoid seats and causing sticking is | |||
being investigated. | |||
(vii) The valves are stroke timed quarterly in accordance | |||
with ASME Section XI requirements. A review of stroke | |||
time data from the F003, F004, F019, and F020 valves | |||
in Unit 2 indicated some apparently abnormal | |||
I | |||
J | |||
. . . . . | |||
<. , | |||
* | |||
. | |||
. | |||
* | |||
6 | |||
unexplained changes in stroke time in past tests. | |||
There was no clear evidence of instances where these | |||
would predict impending failures prior to their | |||
occurrence. | |||
(viii) Additional licensee investigative actions planned or ; | |||
in progress include: | |||
- | |||
Additional vendor inspection of the failed ASCO | |||
solenoids | |||
- | |||
Analysis of the oily film found in the solenoids | |||
- | |||
Removal .and inspection of Unit 1 solenoids and | |||
actuators | |||
- | |||
Evaluate training on solenoid maintenance | |||
- | |||
Evaluate differences between plant maintenance | |||
procedures for the solenoids and the vendor's | |||
recommendations | |||
(f) NRC Inspector's Examination of Inservice Testing | |||
As previously stated above, unusual changes in stroke times | |||
were apparent in the licensee's data for the failed CIV. ' | |||
Concern that the licensee had not adequately addressed | |||
these changes when they occurred was expressed in the NRC | |||
AIT Report 325, 324/88-03. In the current inspection the i | |||
NRC inspector examined the stroke time data extensively, | |||
taking into account the test procedure and valve design. | |||
The inspector also observed performance of stroke timing on l | |||
Unit 1 valves 1-G16-F003, F004, F019 and F020. , | |||
1 | |||
As an almost separate item, the inspector reviewed past | |||
licensee leak test data for these valves. | |||
The inspector's review, observations and his findings | |||
are described below: | |||
(1) Stroke Timing of Units 1 and 2 CIVs G16-F003, F004, F019 i | |||
and F020 | |||
The inspector reviewed the stroke time data recorded | |||
by the lierisee in their testing of the subject Unit 2 | |||
CIVs and also reviewed the data for the similar Unit 1 | |||
CIVs. The review was performed to assess the | |||
licensee's past testing of the valves, determine if l | |||
the test data variability noted by the NRC AIT could ! | |||
be explained, and to determine whether the past test ! | |||
data could be used in predicting the valve failures or | |||
! | |||
- . - . . . | |||
, | |||
' | |||
. | |||
. | |||
I | |||
* | |||
7 I | |||
l | |||
- | |||
. | |||
1 | |||
establishing their cause. A ch*onological tabulation | |||
of the stroke time data and maintenance actions for | |||
the Unit 2 valves was prepared by the NRC inspector | |||
and is given in Table 1. A chronological listing of | |||
the similar Unit 1 stroke time data, also prepared by i | |||
the inspector, is given in Table 2. For use in | |||
identifying abnormal stroke time values, the inspector i | |||
' | |||
calculated a mean and standard deviation for the first ) | |||
five tests of the Unit 2 and Unit I valves (a standard ! | |||
normal distribution was assumed). These means and I | |||
standard deviations are identified in the Tables. In | |||
addition, the NRC inspector identified the stroke | |||
times in the tables which differed from the means by two | |||
and by three standard deviations. If the distribution of | |||
stroke timer from the licensee's tests are normally | |||
distributed. as would be expected, approximately 95% | |||
of test times should (normally) fall within twe standard | |||
deviations of the mean and 99.7% should fall within | |||
three standard deviations. It is common statistical ! | |||
practice to consider values that differ by more than i | |||
three standard deviations from the mean to be abnormal. l | |||
To aid in understanding the test data, the inspector i | |||
reviewed the design of the valves with a maintenance i' | |||
supervisor and reviewed the testing and test data | |||
evaluation procedures - | |||
identified PT-11.3 and | |||
ENP 16.1, respectively. Additionally, the inspector | |||
witnessed stroke time testing performed on Unit 1 l | |||
valves 1-G16-F003, F004, F019 and F020. From his ! | |||
review and his observation of testing, the NRC i | |||
inspector identified the following deficiencies in the l | |||
licensee's testing of the valves: l | |||
- | |||
For valves with stroke times less than 10 | |||
seconds, ASME Section XI requires that, when the | |||
, | |||
stroke time incresses 50% or more, the test | |||
i frequency shall be increased to monthly until l | |||
' | |||
corrective action is taken. ! | |||
1 | |||
Valves 2-G16-F004, F019 and F020 experienced | |||
" | |||
stroke time increases exceeding 50% on August 7, . | |||
1986 and April 27, 1987. F020 experienced over I | |||
a 50% increase in stroke time on July 28, 1987. ! | |||
Test frequency was not increased to monthly and i | |||
no corrective action was taken, j | |||
l | |||
- | |||
The licensee's engineering procedure ENP-16.1, ! | |||
; Pump and Valve Graphing, Section 5.2.12, requires I | |||
] valves with & stroke time between zero and five | |||
:' | |||
seconds to be tested monthly until the next , | |||
regularly scheduled surveillance if the stroke i | |||
g time increases by more than two seconds. | |||
- | |||
_ | |||
- ., _ . - . | |||
__. | |||
r | |||
- | |||
. | |||
. | |||
8 | |||
Although the licensee's ISI Project Engineer | |||
stated that the licensee actually adhered to the | |||
more stringent ASME requirement described above, | |||
even the ENP 16.1 requirement was not met. Test | |||
frequency was not increased to monthly when | |||
stroke time increases exceeding two seconds were | |||
experienced for valve 2-G16-F020 on August 28, | |||
1987 and April 27, 1987 and valve 2-G16-F019 on | |||
April 27, 1987 and August 7, 1986. | |||
- | |||
ASME Section XI, requires the reporting of any | |||
abnormality or erratic action that occurs in | |||
stroke timing power operated valves. By | |||
implication, the reporting process includes ; | |||
recognition of the abnormal or erratic action and ; | |||
evaluation to assess the need for corrective | |||
action. The licensee failed to report and ? | |||
evaluate many examples of abnormal changes in | |||
stroke time. Assuming a standard normal | |||
distribution and utilizing the stroke times from ) | |||
the first five tests of the subject Unit 2 CIVs 1 | |||
after their April 1986 modification the calculated | |||
stroke time mean = 3.17 seconds with a standard ! | |||
deviation = 0.27 seconds. Statistically, stroke | |||
times differing from the mean by more than 3 | |||
standard deviations are considered abnormal. On ; | |||
this basis, abnormal stroke times were i | |||
experienced but were not recognized, reported or | |||
evaluated for the following valves and test , | |||
dates: 1 | |||
2-G16-F003 tests on October 29, 1987, | |||
June 20, 1987 and October 25', 1986 | |||
2-G16-F004 tests on January 25, 1987, | |||
July 30, 1986 and July 24, 1986 | |||
2-G16-F019 tests on January 25, 1987, 1 | |||
July 30, 1986 and July 24, 1986 | |||
2-G16-F020 tests on July 28, 1987, | |||
January 25, 1987, July 30, 198C, July 24, | |||
1986 and July 17, 1986 | |||
Evaluating Unit I stroke time data in a similar j | |||
manner, the calculated stroke time mean = 3.13 i | |||
seconds with a standard deviation = 0.23 seconds | |||
(values suprisingly close to those obtained for | |||
the Unit 2 valves). Again identifying "abnormal" | |||
i | |||
l | |||
1 | |||
e | |||
. - | |||
- | |||
9 | |||
. | |||
stroke times as those that-differ from the.mean | |||
by more than 3 standard deviations, the following | |||
tests should have been (but were not) recognized, | |||
reported or evaluated as abnormal valve action: | |||
1-G16-F004, F019 and F020 tests on November 2, | |||
1987 | |||
The licensee had no criteria for identifying, | |||
reporting or evaluating abnormal- or erratic | |||
stroke times. | |||
- | |||
A$ME Section XI requires the licensee to determine | |||
the full-stroke time for the subject Unit 2 | |||
(and similar Unit 1) valves. Full-stroke time | |||
is defined as that time interval from initiation | |||
of the actuating signal to the end of the | |||
actuating cycle. The NRC inspector determined | |||
that the licensee's procedure for testing the | |||
subject Unit 2 valves and similar Unit 1 valves | |||
specified erroneous criteria for determination | |||
of full-stroke time. The procedure, PT 11.3, | |||
Drywell Drains System Valve Operability Test, | |||
specifies that stroke time is to be measured | |||
from switch actuation until the red (OPEN) light | |||
extinguishes. For the subject air operated | |||
valves, the red light extinguishes at the first | |||
stage of valve movement as valve motion leaves | |||
the first (red light actuating) limit switch. | |||
. | |||
The time required for the value to complete its | |||
stroke by moving from the OPEN switch to trip | |||
the CLOSED (green light) switch is omitted in | |||
the licensee's procedure. At the inspector's | |||
request, the licensee stroke timed the Unit 1 | |||
valves 1-G16-F003, F004, F019 and F020 in | |||
accordance with PT 11.3 and again correctly. The | |||
testing, which was observed by the inspector, | |||
demonstrated that valve full-stroke time was | |||
over twice the time obtained by following | |||
procedure. Comparative data, obtained in the | |||
testing, is shown in Table 3. | |||
- | |||
Based on discussion with a licensee test person | |||
and on the licensee's recorded test data, it is | |||
apparent that test personnel usually did not | |||
n | |||
. | |||
. | |||
* | |||
10 | |||
. | |||
follow the erroneous procedural criteria for | |||
determination of full-stroke time, but rather | |||
used correct criteria. | |||
- | |||
The current revision of PT 11.3 (Revision 12) | |||
was approved November 2,1984 and the licensee's | |||
records indicate over 30 instances of its use | |||
since that approval. Although the Operations | |||
personnel who performed the procedure apparently | |||
understood how the stroke timing should be | |||
performed (and usually performed it correctly), | |||
the erroneous method of stroke timing specified | |||
in the procedure was not promptly identified or | |||
corrected and was occasionally used. | |||
- | |||
The licensee informed the inspector that most, | |||
if not all, of their procedures for stroke timing | |||
air operated valves contained the same error as | |||
in Procedure PT 11.3. | |||
The deficiencies described above represent a violation | |||
of TS 4.0.5 requirements to perform valve testing in | |||
accordance with ASME Section XI and of TS 6.8.1 | |||
requirements to properly establish, implement and | |||
maintain written procedures for safety related- | |||
equipment (e.g. valve) testing. This violation is | |||
identified as violation 325,'324/88-06-01, Inadequate | |||
Stroke Time Testing of Air Operated Valves. | |||
The NRC inspector was able -to at least in part, | |||
explain abnormal stroke times experienced by the | |||
licensee in testing valves 2-G16-F003, F004, F019 and | |||
F020. However, the inspector did not identify any | |||
trends or indicators in the licensee's stroke times | |||
that would aid in predicting the failures experienced | |||
or their cause. The inspector notes, however, that | |||
indicators or trends may have been present that were | |||
undetected due to the insufficient control of the | |||
testing. The inspector encouraged the licensee to | |||
make an effort to perform their stroke timing more | |||
accurately. Improvements might be readily obtained | |||
through increased motivation and better training and | |||
procedures. | |||
(ii) Leak Rate Testing of Units 1 and 2 CIVs G16-F003, F004, | |||
F019 and F020 | |||
The inspector reviewed licensee leak rate test results | |||
for the subject valves to verify that testing was | |||
performed at the required frewency. was acceptable, | |||
and whether the datt contained say evidence of the | |||
, | |||
_. . _ | |||
, | |||
.- | |||
. | |||
11 | |||
. | |||
cause s' the closure failures that have been | |||
experiencea. For Unit 1 the inspector reviewed | |||
results obtained since May 19, 1977 and for Unit 2 | |||
results obtained since January 27, 1979 were reviewed. | |||
The inspector did not identify any deficiencies in the | |||
testing from this review and there was no evidence | |||
suggesting a cause of the recent closure failures, | |||
b. Loop B RHR Pumps and Valves | |||
The inspector reviewed completed records of procedure PT 8.2.2b, | |||
LPCI/RHR System Operability Test - Loop B, to verify that testing and | |||
evaivation had been performed in accordance with procedural and ASME | |||
Section XI regt.irements. The inspector checked the records for all | |||
of Units 1 and 2 Loop B RHR pumps and for four valves, The | |||
components and the test dates of the records reviewed are as follows: | |||
Component Test Dates | |||
Pump 1B 1/22/88, 10/18/87, 8/7/87, 7/26/87, | |||
7/25/87, S/1/87, 12/20/86 ; | |||
Pump 10 Same as above j | |||
Pump 28 10/31/87, 8/8/87, 5/16/87, 2/22/87, | |||
2/20/87, 2/19/87, 12/4/86 | |||
Pump 2D Same as above | |||
Valve 1-E11 - F007B Same as Dump 1B i | |||
Valve 1-E11 F0118 Same as Pump 1B i | |||
Valve 1-E11 - F017B Same as Pump 1B | |||
Valve 1-E11 - F016B Same as Pump 1B | |||
Valve 2-E11 - F0078 Same as Pump 2 B except no test 2/20/87 | |||
1 | |||
Valve 2-E11 - F0118 Same as Pump 2B except no test 2/20/87 l | |||
l | |||
Valve 2-E11 - F0178 Same as Pump 2B except no test 2/20/87 | |||
J | |||
Valve 2-E11 - F016B Same as Pump 2B except no test 2/20/87 I | |||
l | |||
No violations or deviations were identified during the above reviews with I | |||
the exception of the violation noted in paragraph 4.a(1)(f). | |||
1 | |||
\ | |||
. | |||
TABLE 1 | |||
* | |||
CHRONOLOGICAL TABULAllON Of CLOSURE STROKE TIMING AND MAINTENANCE FOR | |||
UNIT 2 VALVES 2-G16-F003, F004, F019 and F020 * | |||
Date St roke T imes ( Seconds ) for Valves Maintenance | |||
19Q1_ F004 FQ19 f_QRQ | |||
10/29/87 5.0/ Failed 3.0 2.5* E 004 so l eno i d wa s rebu i l t a r te r ra l l u re - it then passed | |||
7/28/81 3.6 2.9 3.4 4.9/ | |||
F003 solenoid was rebuilt af ter railure - it then passed | |||
7/17/87 Failed --- --- --- | |||
6/20/87 2.31/ 2.41* 3.02 2.49' | |||
4/27/87 2.38" 2.60* 3.14 2.99 | |||
None-believed caused by not holding F003 switch closed. Subsequent | |||
4/27/87 failed --- --- --- | |||
actuation resulted in closure | |||
1/25/87 3.62 .72/ .98/ .84/ | |||
10/25/86 3.81/ 2.95 3.43 2.97 | |||
G/7/86 2.95 3.13 3.45 2.74 | |||
7/31/86 --- --- --- --- f003 solenoid rebui l t a f ter ra i lure - it then passed. | |||
7/30/86 Failed 1.27/ 1.30/ 1.12/ See 7/31/86 | |||
7/25/86 2.70 3.04 3.41 2.99 | |||
7/24/86 2.6* 2.2/ 2.1/ 2.0/ | |||
7/17/86 2.8 2.90 3.29 3.80/ | |||
7/13/86 3.06 2.87 3.41 3.07 | |||
7/12/86 None - intended solenoid replacement but had wrong replacement. Valve | |||
cycled acceptably when tested with old solenoid. Test f requency was | |||
increased ( 7/15/86 request) | |||
7/11/86 Failed _ _ _ | |||
1his was an auto signal railure - see above maintenance actions | |||
6/15/86 Approximate tnd of Refueling Outage That Sta rted 11/30/65 | |||
5/16/86 3.2 3.0 3.5 3.0 ' | |||
5/15/86 3.37 2.91 3.33 2.80 | |||
4/30/86 --- --- --- -- F004 lapped to correct excessive leakage (had railed leak test) | |||
4/26/86 ,__ ___ __._ ___ | |||
F020 rewi red to di fferent limit switch and passed test - apparently not a | |||
solenoid problem (?) | |||
4/26/86 3.13 3.22 3.5 Failed | |||
Notes: (1) for the stroke times recorded in the first five valve tests (19 values during 4/26 - 7/17/86) the mean = 3.17 and | |||
the standard deviation = 0.27 | |||
(2) * indicates stroke times which dirrer from the above mean by more than 2 standard deviations | |||
(3) / indicates stroke times which dirrer from the above mean by more than 3 standard deviations | |||
. -- _ - - - - - - _ _ _ _ _ . . _ _ _ _ . _ _ _- - _ _ __ - _ . - _ _ _ . _ _ _ - . | |||
- | |||
- . - | |||
, | |||
* | |||
TABLE 2 | |||
* | |||
CHRONOLOGICAL TABULATION Of CLOSURE STROKE TIMING FOR UNIT 1 | |||
VALVES 1-G16-F003, F004, TO19 AND F020 | |||
Date Stroke Times (Seconds) fo r Va l ves | |||
F QR}_ ffRQ9 IO19 fy?J | |||
1/2/88 2.81 3.16 3.20 2.90 | |||
11/2/87 2.70 2.25/ 2.16/ 1.88/ | |||
8/9/87 3.0 3.6* 3.3 3.1 | |||
6/20/87 End of Refueling Outage That Started 2/14/87 | |||
5/9/81 3.33 3.25 3.21 3.08 | |||
4/2G/87 3.80 3.34 3.46 3.28 | |||
2/5/87 2.79 3.07 3.13 2.99 | |||
11/6/87 2.87 2.80 3.15 2.76 | |||
8/5/86 2.96 3.09 3.34 3.20 | |||
5/6/86 2.18 3 . 0 *> 3.15 3.11 | |||
2/12/86 2.90 3.05 3.23 3.13 | |||
11/2/85 2.76 3.08 3.38 3.13 | |||
9/22/85 2.80 3.39 3.61* 3.53 | |||
Notes: (1) I')r the stroke times recorded in the first five valvo tests (20 values durin9 9/22/85 - 8/5/86) the mean = 3.13 and | |||
the standaro deviation = 0.23 | |||
(2) * indicates stroke times which dirror f rom the above mean by more than 2 standard deviations | |||
(3) / indicates stroke times which dirrer f rom the above mean by more than 3 standard deviations | |||
_ _ _ _ _ _ - - . - _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . - _ _ - _ _ _ _ _ - -_ _ ._-.________J | |||
- - - - - - - . -_----_--- _ _ - - - - | |||
. | |||
. | |||
7 | |||
b | |||
* | |||
I | |||
- | |||
15 | |||
e | |||
TABLE 3 | |||
COMPARIS0N OF STROKE TIMES OBTAINED BY FOLLOWING | |||
PROCEDURE VERSUS CORRECT METHOD FOR | |||
UNIT 1 VALVES 1-G16-F003, F004, F019 AND F020 | |||
Valve No. Stroke Time Measured (Seconds) | |||
Per Procedure - (From Correct Method - (From | |||
switch closure actuation switch closure actuation | |||
until red light off) until green light on) | |||
1G16-F003 1.38#, 0.96#, 1.16 2.76 | |||
1G16-F004 1.14 2.87 | |||
1G16-F019 1.16 3.02 | |||
1G16-F020 1.06 2.76 | |||
mean = 1.13 2.85 | |||
standard deviation = .05 .12 | |||
Note: # denotes the first two measurements at the start of testing. | |||
The operator who performed the testing felt that they were not | |||
sufficiently accurate and performed the third measurement which he | |||
considered sati sf actory. The # values were not included in the | |||
calculation of mean and standard deviation. | |||
. - | |||
}} |
Latest revision as of 09:02, 27 October 2020
ML20148H790 | |
Person / Time | |
---|---|
Site: | Brunswick ![]() |
Issue date: | 03/02/1988 |
From: | Blake J, Girard E NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
To: | |
Shared Package | |
ML20148H775 | List: |
References | |
50-324-88-06, 50-324-88-6, 50-325-88-06, 50-325-88-6, NUDOCS 8803300083 | |
Download: ML20148H790 (15) | |
See also: IR 05000324/1988006
Text
r 1
l
' [kQ 080g UNIVED STATES l
/ 'o NUCLEAR REGULATORY COMMISSION 1
[ n REGloN 11 )
j 101 MARIETTA STREET, N.W.
S
- '.. 'f ATL ANTA, GEORGI A 30323
't, go
.....
Report Nos.: 50-325/88-06 and 50-324/88-06
Licensee: Carolina Power and Light Company
P. O. Box 1551
Raleigh, NC 27602
Docket Nos.: 50-325 and 50-324 License Nos.: DPR-71 and DPR-62
Facility Name: Brunswick 1 and 2
Inspection C du d- nuary 25-29, 1988
Inspecto : - // ' '
E B
j f tE rd D' ate' signed
Appro ed by _ _ f .
7' &
J.J.faiake, Chief Date Signed
Pt ials and Processes Section
j iv sion of Reactor Safety
SUMMARY
Scope: This routine, unannounced inspection was in the area of inservice
testing of pumps and valves.
Results: One violation was identified involving deficiencies in valve testing
procedures and failures to follow the procedures.
8803300083 830310
PDR ADOCK 05000324
O DCD
.. . .. ~- , ,
e--
-
-
.
.
.
REPORT DETAILS
.
1. Persons Contacted
Licensee Employees
S. Boyse, Project Engineer, Technical Support
- R. Creech, Instrumentation and Control Supervisor (Unit 2) Maintenance
- J. Crider, Inservice Inspection Responsible Engineer, Technical Support
C. Dietz, Plant General Marager
W. Dorman, Quality Assurance Supervisor, QA/QC
- K. Enzor, Director, Regulatory Compliance
- L. Jones, Director, QA/QC
- W. Martin, Principal Engineer, On-Site Nuclear Safety
- J. O'Sullivan, Manager, Maintenance
C. Patterson, Inservice Inspection Project Engineer
- R. Poulk, Project Specialist - NRC, Regulatory Compliance
- J. Simon, Engineer, Operations
- L. Wheatley, Inservice Inspection Project Engineer, Technical Services
- E. Wilson, Engineering Supervisor (Nuclear), Technical Support
- A. Worth, Engineering Supervisor (Discipline), Technical Support
NRC Resident Inspector
- W. Ruland, Senior Resident Inspector '
- Attended exit interview
2. Exit Interview
The inspection scope and findings were summarized on January 29, 1986, -
,
with those persons indicated in Paragraph 1. The inspector described the
areas inspected and discussed in detail the inspection findings. -
Dissenting comments were not received from the licensee.
Item Number Status Title / Reference Paragraph
325, 324/88-06-01 Open Violation - Inadequate stroke
time testing of air operated I
valves (Paragraph 4)
Proprietary information is not contained in this report.
3. Licensee Action on Previous Enforcement Matters l
l
This subject was not addressed in the inspection.
l
l
l
I
!
.
-
.
,
.
2
l
4. Inservice Testing of Pumps and Valves (73756) - Units 1 and 2 .
The NRC inspector examined selected examples of the licensee's inservice i
testing of pumps and valves to verify compliance with licensee commitments '
and regulatory requirements. Based on Technical Specification (TS)
4.0.5.a.2,10 CFR 50.55a(g), and on NRC approval letter dated April 8,
1987, the licensee is currently required to test their pumps and valves in
accordance with ASME Section.XI (80W81).
The inspector directed his examination primarily to the licensee's past
testing and evaluation of the test results from four Unit 2 Containment
Isolation Valves (CIVs) which failed to close in a recent actuation. In
addition, the inspector examined the testing of the licensee's Loop B
Residual Heat Removal (RHR) pumps and of four RHR valves that are tested
along with those pumps. Details of the inspector's examinations and
findings are described below.
a. Containment Isolation Valves Which Failed to Close in Response to
Isolation Signal on January 2,1988
(1) Background Information on CIVs
On January 2,1988, the Unit 2 valves required for containment
isolation of flow from the drywell floor drain and equipment
drain failed to close in response to an automatic .(closure)
actuation signal.
Subsequently, the failure event has been extensively
investigated by the licensee. Prior to the inspection described
herein, NRC examination of the event had included a preliminary
assessment by the Resident Inspector and a special inspection of
the matter by a nine person NRC Augmented Inspection Team (AIT).
This inspector reviewed background information, including the
status of ongoing licensee investigation, through discussions
with AIT members and licensee personnel and through review
of the AIT Inspection Report (Report No. 325, 324/83-03). The
following background data was noted by the inspector relevant to
his examination of the testing.
(a) Description of Valves and Event
The failed valves are identified 2-G16-F003, -F004, -F019
and -F020. Valves 2-G16-F003 and -F004 are in the drywell l
floor drain system piping and valves 2-G16-F019 and -F020 ;
in the drywell equipment drain system piping. They are
components of the Primary Containment Isolation System i
(PCIS). The valves are three inch, 150 lb. gate valves
manufactured of cast carbon steel. The valves are opened
by an air actuation cylinder, controlled by Automatic
l
1
l
1
F
.
.
,
. .
3
.
Switch Company (ASCO) solenoid valves, and are spring
assisted closed upon loss of air pressure. The PCIS is a
safety-related system that is designed to prevent or liLit
the release of radioactive materials that may result from
postulated accidents through rapid, automatic isolation of
process lines which penetrate the primary containment.
F003 and F019 are powered through relay K17 and F004 and
F020 through relay K18. A single failure of either circuit
to open would not cause a loss of containment integrity, as
one valve in each line would still close.
(b) Modification to Valves
The solenoid valves which control the opening and closing
of the air supply to the subject Unit 2 CIVs were modified
to meet the Environmental Qualification (EQ) requirements
of 10 CFR 50.49 in April 1986.
Virtually identical Unit i valves (1G16-F003, F004, F019
and F020) had been similarly modified in April 1985. Due
to an error, the licensee installed ASCO Model 206-832 G
solenoids which fail open. Af ter consultation with the
vendor, they modified the valves to fail closed by replacing
the valve springs with the springs from a Model 206-832,
Type "F", rebuild kit.
(c) Previous History of Failures and Maintenance
Neither the Unit i nor Unit 2 valves experienced any
closure failures like those in the January 1988 event prior
to the modification referred to in (b) above. The Unit i
valves continued to operate satisfactorily following the
modification, while the Unit 2 valves experienced six
apparent failures. The closure tests, test failures and
the post failure maintenance that was performed on the i
'
valves following EQ modification are listed in Table 1.
Stroke time data included in Table 1 is discussed in 4.f
below. The licensee identified two of the failures listed ,
in Table 1 as being the result of not holding the actuation l
switch in the closed position for a sufficient period. ;
That is, they were not considered hardware failures. !
4
However, the inspector notes that no similar problems were l
identified in Unit 1 valve testing. Also, Region II
specialists who have reviewed the actuation circuit
consider it unnecessary to hold the switch in position.
(d) Post Event Licensee Actions to Actuate CIVs l
About 3 minutes after a manual scram of Unit 2, the Control
Operator (CO), who had started verifying the valve
\
__
l
- -
g
c
.
.
4 ,
isolations, noted that valves 2-G16-F003, F004, F019 and
F020 had not automatically closed as required. He then
attempted to manually close the four valves. 2-G16-F003 ,
and -F004 closed immediately vhen the handswitch was placed
to CLOSE position (3 minutes af ter the scram). 2-G16-F019
and -F020 did not close on n.snual switch actuation. He
cycled 2-G16-F019 and -F020 switch to CLOSE three to five
times each, holding for 5 to 10 seconds in CLOSE position
each time. 2-G16-F019 and F020 remained OPEN.
'
The Unit 1 Shift Foreman (SF) then also attempted to close
valves F019 and F020 by holding their switches in the CLOSE
position for more than 10 seconds. 2-G16-F019 and -F020
remained OPEN.
.
Approximately two minutes later (eix .iinutes after the scram)
the SF (Unit 1) noted that 2-G16-F020 valve had CLOSED and
that 2-G16-F019 was still OPEN. The SF held the manual
switch for 2-G16-F019 in CLOSE . position for an additional
10 seconds. 2-G16-F019 remained OPEN.
'
After another two minutes the SF noticed that 2-G16-F019 had
CLOSED (eight minutes after the scram).
,
(e) Post Event Licensee Investigations into the Cause of the
CIV Failures
(i) Shortly after the event, the operability of the Unit i
valves was verified through manual cycling and a logic
test. The valves functioned nonnally.
(ii) Various logic and other tests of the Unit 2 valves
were conducted and, in each instance, the circuits and
valves exhibited apparently normal actuation - save
for an instance where the red (OPEN) light on F003 was
slow to extinguish. The red light stayed on for
almost 10 seconds and was illuminated concurrently
with the green (CLOSE) light. The position lights are
activated by limit switches located at the beginning
and end of valve stroke travel. Actuation causes the
valve disk to (1) move from the initial limit switch l
- causing that position light to extinguish (2) to i
between the open and closed limit switches - with no
position light on, and finally (3) to the other limit
switch which will cause that position light to l
illuminate. Based on tests of the valves observed by I
the NRC inspector in his examination of this matter,
the open light should extinguish about one second after
valve actuation.
l
l
l
- .- .
,
.
. ,
.
5
(iii) The solenoids were suspected as a likely cause of the
failure because of the valve closure failures
experienced subsequent to the solenoid replacements in
April 1985. Thus, the solenoids have been closely
checked in the post event investigation:
-
The licensee removed the solenoid from valve
2-G16-F019 and tested it with 35 psi (normal) and
5 psi air pressure. The solenoid performed
satisfactorily at both pressures. The solenoid
was disassembled and inspected. The inspection
revealed minor debri, a light oil film in the
solenoid body and s darkened area on one solenoid-
stem. Bottom disk travel was found +7. mils out
of tolerance. None of these conditions appeared -
severe enough to explain the valve failure.
-
The solenoids from 2-G16-F003 and F020 were
removed and inspected and no problems were ,
'
identified, though minor debri and an oil film
like that in F019 were noted. The vendor's
representative participated in inspection of the
F020 valve solenoid. The vendor representative
did not identi fy any deficiencies in the 4
solenoid.
(iv) The air cleanliness, pressure regulator and in-line
filter serving F003 were checked and determined
acceptable.
(v) 'he logic relays K17 and K18 were inspected and
widence of arcing and sticking were noted on K18.
This might provide some bearing on the failure of F004
and F020 (both on K18) had it not been for the fact that
when F004 closed in manual actuation following the
event, the F020 valve would not close when similarly
actuated.
(vi) The air actuator of F020 was removed and inspected.
The actuator was in good condition but contained a
liberal amount of grease. In valve actuation, air
passes from the actuator through the solenoid. The
possibility that hydrocarbons from the grease may be
reaching the solenoid seats and causing sticking is
being investigated.
(vii) The valves are stroke timed quarterly in accordance
with ASME Section XI requirements. A review of stroke
time data from the F003, F004, F019, and F020 valves
in Unit 2 indicated some apparently abnormal
I
J
. . . . .
<. ,
.
.
6
unexplained changes in stroke time in past tests.
There was no clear evidence of instances where these
would predict impending failures prior to their
occurrence.
(viii) Additional licensee investigative actions planned or ;
in progress include:
-
Additional vendor inspection of the failed ASCO
solenoids
-
Analysis of the oily film found in the solenoids
-
Removal .and inspection of Unit 1 solenoids and
actuators
-
Evaluate training on solenoid maintenance
-
Evaluate differences between plant maintenance
procedures for the solenoids and the vendor's
recommendations
(f) NRC Inspector's Examination of Inservice Testing
As previously stated above, unusual changes in stroke times
were apparent in the licensee's data for the failed CIV. '
Concern that the licensee had not adequately addressed
these changes when they occurred was expressed in the NRC
AIT Report 325, 324/88-03. In the current inspection the i
NRC inspector examined the stroke time data extensively,
taking into account the test procedure and valve design.
The inspector also observed performance of stroke timing on l
Unit 1 valves 1-G16-F003, F004, F019 and F020. ,
1
As an almost separate item, the inspector reviewed past
licensee leak test data for these valves.
The inspector's review, observations and his findings
are described below:
(1) Stroke Timing of Units 1 and 2 CIVs G16-F003, F004, F019 i
and F020
The inspector reviewed the stroke time data recorded
by the lierisee in their testing of the subject Unit 2
CIVs and also reviewed the data for the similar Unit 1
CIVs. The review was performed to assess the
licensee's past testing of the valves, determine if l
the test data variability noted by the NRC AIT could !
be explained, and to determine whether the past test !
data could be used in predicting the valve failures or
!
- . - . . .
,
'
.
.
I
7 I
l
-
.
1
establishing their cause. A ch*onological tabulation
of the stroke time data and maintenance actions for
the Unit 2 valves was prepared by the NRC inspector
and is given in Table 1. A chronological listing of
the similar Unit 1 stroke time data, also prepared by i
the inspector, is given in Table 2. For use in
identifying abnormal stroke time values, the inspector i
'
calculated a mean and standard deviation for the first )
five tests of the Unit 2 and Unit I valves (a standard !
normal distribution was assumed). These means and I
standard deviations are identified in the Tables. In
addition, the NRC inspector identified the stroke
times in the tables which differed from the means by two
and by three standard deviations. If the distribution of
stroke timer from the licensee's tests are normally
distributed. as would be expected, approximately 95%
of test times should (normally) fall within twe standard
deviations of the mean and 99.7% should fall within
three standard deviations. It is common statistical !
practice to consider values that differ by more than i
three standard deviations from the mean to be abnormal. l
To aid in understanding the test data, the inspector i
reviewed the design of the valves with a maintenance i'
supervisor and reviewed the testing and test data
evaluation procedures -
identified PT-11.3 and
ENP 16.1, respectively. Additionally, the inspector
witnessed stroke time testing performed on Unit 1 l
valves 1-G16-F003, F004, F019 and F020. From his !
review and his observation of testing, the NRC i
inspector identified the following deficiencies in the l
licensee's testing of the valves: l
-
For valves with stroke times less than 10
seconds, ASME Section XI requires that, when the
,
stroke time incresses 50% or more, the test
i frequency shall be increased to monthly until l
'
corrective action is taken. !
1
Valves 2-G16-F004, F019 and F020 experienced
"
stroke time increases exceeding 50% on August 7, .
1986 and April 27, 1987. F020 experienced over I
a 50% increase in stroke time on July 28, 1987. !
Test frequency was not increased to monthly and i
no corrective action was taken, j
l
-
The licensee's engineering procedure ENP-16.1, !
- Pump and Valve Graphing, Section 5.2.12, requires I
] valves with & stroke time between zero and five
- '
seconds to be tested monthly until the next ,
regularly scheduled surveillance if the stroke i
g time increases by more than two seconds.
-
_
- ., _ . - .
__.
r
-
.
.
8
Although the licensee's ISI Project Engineer
stated that the licensee actually adhered to the
more stringent ASME requirement described above,
even the ENP 16.1 requirement was not met. Test
frequency was not increased to monthly when
stroke time increases exceeding two seconds were
experienced for valve 2-G16-F020 on August 28,
1987 and April 27, 1987 and valve 2-G16-F019 on
April 27, 1987 and August 7, 1986.
-
ASME Section XI, requires the reporting of any
abnormality or erratic action that occurs in
stroke timing power operated valves. By
implication, the reporting process includes ;
recognition of the abnormal or erratic action and ;
evaluation to assess the need for corrective
action. The licensee failed to report and ?
evaluate many examples of abnormal changes in
stroke time. Assuming a standard normal
distribution and utilizing the stroke times from )
the first five tests of the subject Unit 2 CIVs 1
after their April 1986 modification the calculated
stroke time mean = 3.17 seconds with a standard !
deviation = 0.27 seconds. Statistically, stroke
times differing from the mean by more than 3
standard deviations are considered abnormal. On ;
this basis, abnormal stroke times were i
experienced but were not recognized, reported or
evaluated for the following valves and test ,
dates: 1
2-G16-F003 tests on October 29, 1987,
June 20, 1987 and October 25', 1986
2-G16-F004 tests on January 25, 1987,
July 30, 1986 and July 24, 1986
2-G16-F019 tests on January 25, 1987, 1
July 30, 1986 and July 24, 1986
2-G16-F020 tests on July 28, 1987,
January 25, 1987, July 30, 198C, July 24,
1986 and July 17, 1986
Evaluating Unit I stroke time data in a similar j
manner, the calculated stroke time mean = 3.13 i
seconds with a standard deviation = 0.23 seconds
(values suprisingly close to those obtained for
the Unit 2 valves). Again identifying "abnormal"
i
l
1
e
. -
-
9
.
stroke times as those that-differ from the.mean
by more than 3 standard deviations, the following
tests should have been (but were not) recognized,
reported or evaluated as abnormal valve action:
1-G16-F004, F019 and F020 tests on November 2,
1987
The licensee had no criteria for identifying,
reporting or evaluating abnormal- or erratic
-
A$ME Section XI requires the licensee to determine
the full-stroke time for the subject Unit 2
(and similar Unit 1) valves. Full-stroke time
is defined as that time interval from initiation
of the actuating signal to the end of the
actuating cycle. The NRC inspector determined
that the licensee's procedure for testing the
subject Unit 2 valves and similar Unit 1 valves
specified erroneous criteria for determination
of full-stroke time. The procedure, PT 11.3,
Drywell Drains System Valve Operability Test,
specifies that stroke time is to be measured
from switch actuation until the red (OPEN) light
extinguishes. For the subject air operated
valves, the red light extinguishes at the first
stage of valve movement as valve motion leaves
the first (red light actuating) limit switch.
.
The time required for the value to complete its
stroke by moving from the OPEN switch to trip
the CLOSED (green light) switch is omitted in
the licensee's procedure. At the inspector's
request, the licensee stroke timed the Unit 1
valves 1-G16-F003, F004, F019 and F020 in
accordance with PT 11.3 and again correctly. The
testing, which was observed by the inspector,
demonstrated that valve full-stroke time was
over twice the time obtained by following
procedure. Comparative data, obtained in the
testing, is shown in Table 3.
-
Based on discussion with a licensee test person
and on the licensee's recorded test data, it is
apparent that test personnel usually did not
n
.
.
10
.
follow the erroneous procedural criteria for
determination of full-stroke time, but rather
used correct criteria.
-
The current revision of PT 11.3 (Revision 12)
was approved November 2,1984 and the licensee's
records indicate over 30 instances of its use
since that approval. Although the Operations
personnel who performed the procedure apparently
understood how the stroke timing should be
performed (and usually performed it correctly),
the erroneous method of stroke timing specified
in the procedure was not promptly identified or
corrected and was occasionally used.
-
The licensee informed the inspector that most,
if not all, of their procedures for stroke timing
air operated valves contained the same error as
in Procedure PT 11.3.
The deficiencies described above represent a violation
of TS 4.0.5 requirements to perform valve testing in
accordance with ASME Section XI and of TS 6.8.1
requirements to properly establish, implement and
maintain written procedures for safety related-
equipment (e.g. valve) testing. This violation is
identified as violation 325,'324/88-06-01, Inadequate
Stroke Time Testing of Air Operated Valves.
The NRC inspector was able -to at least in part,
explain abnormal stroke times experienced by the
licensee in testing valves 2-G16-F003, F004, F019 and
F020. However, the inspector did not identify any
trends or indicators in the licensee's stroke times
that would aid in predicting the failures experienced
or their cause. The inspector notes, however, that
indicators or trends may have been present that were
undetected due to the insufficient control of the
testing. The inspector encouraged the licensee to
make an effort to perform their stroke timing more
accurately. Improvements might be readily obtained
through increased motivation and better training and
procedures.
(ii) Leak Rate Testing of Units 1 and 2 CIVs G16-F003, F004,
F019 and F020
The inspector reviewed licensee leak rate test results
for the subject valves to verify that testing was
performed at the required frewency. was acceptable,
and whether the datt contained say evidence of the
,
_. . _
,
.-
.
11
.
cause s' the closure failures that have been
experiencea. For Unit 1 the inspector reviewed
results obtained since May 19, 1977 and for Unit 2
results obtained since January 27, 1979 were reviewed.
The inspector did not identify any deficiencies in the
testing from this review and there was no evidence
suggesting a cause of the recent closure failures,
b. Loop B RHR Pumps and Valves
The inspector reviewed completed records of procedure PT 8.2.2b,
LPCI/RHR System Operability Test - Loop B, to verify that testing and
evaivation had been performed in accordance with procedural and ASME
Section XI regt.irements. The inspector checked the records for all
of Units 1 and 2 Loop B RHR pumps and for four valves, The
components and the test dates of the records reviewed are as follows:
Component Test Dates
Pump 1B 1/22/88, 10/18/87, 8/7/87, 7/26/87,
7/25/87, S/1/87, 12/20/86 ;
Pump 10 Same as above j
Pump 28 10/31/87, 8/8/87, 5/16/87, 2/22/87,
2/20/87, 2/19/87, 12/4/86
Pump 2D Same as above
Valve 1-E11 - F007B Same as Dump 1B i
Valve 1-E11 F0118 Same as Pump 1B i
Valve 1-E11 - F017B Same as Pump 1B
Valve 1-E11 - F016B Same as Pump 1B
Valve 2-E11 - F0078 Same as Pump 2 B except no test 2/20/87
1
Valve 2-E11 - F0118 Same as Pump 2B except no test 2/20/87 l
l
Valve 2-E11 - F0178 Same as Pump 2B except no test 2/20/87
J
Valve 2-E11 - F016B Same as Pump 2B except no test 2/20/87 I
l
No violations or deviations were identified during the above reviews with I
the exception of the violation noted in paragraph 4.a(1)(f).
1
\
.
TABLE 1
CHRONOLOGICAL TABULAllON Of CLOSURE STROKE TIMING AND MAINTENANCE FOR
UNIT 2 VALVES 2-G16-F003, F004, F019 and F020 *
Date St roke T imes ( Seconds ) for Valves Maintenance
19Q1_ F004 FQ19 f_QRQ
10/29/87 5.0/ Failed 3.0 2.5* E 004 so l eno i d wa s rebu i l t a r te r ra l l u re - it then passed
7/28/81 3.6 2.9 3.4 4.9/
F003 solenoid was rebuilt af ter railure - it then passed
7/17/87 Failed --- --- ---
6/20/87 2.31/ 2.41* 3.02 2.49'
4/27/87 2.38" 2.60* 3.14 2.99
None-believed caused by not holding F003 switch closed. Subsequent
4/27/87 failed --- --- ---
actuation resulted in closure
1/25/87 3.62 .72/ .98/ .84/
10/25/86 3.81/ 2.95 3.43 2.97
G/7/86 2.95 3.13 3.45 2.74
7/31/86 --- --- --- --- f003 solenoid rebui l t a f ter ra i lure - it then passed.
7/30/86 Failed 1.27/ 1.30/ 1.12/ See 7/31/86
7/25/86 2.70 3.04 3.41 2.99
7/24/86 2.6* 2.2/ 2.1/ 2.0/
7/17/86 2.8 2.90 3.29 3.80/
7/13/86 3.06 2.87 3.41 3.07
7/12/86 None - intended solenoid replacement but had wrong replacement. Valve
cycled acceptably when tested with old solenoid. Test f requency was
increased ( 7/15/86 request)
7/11/86 Failed _ _ _
1his was an auto signal railure - see above maintenance actions
6/15/86 Approximate tnd of Refueling Outage That Sta rted 11/30/65
5/16/86 3.2 3.0 3.5 3.0 '
5/15/86 3.37 2.91 3.33 2.80
4/30/86 --- --- --- -- F004 lapped to correct excessive leakage (had railed leak test)
4/26/86 ,__ ___ __._ ___
F020 rewi red to di fferent limit switch and passed test - apparently not a
solenoid problem (?)
4/26/86 3.13 3.22 3.5 Failed
Notes: (1) for the stroke times recorded in the first five valve tests (19 values during 4/26 - 7/17/86) the mean = 3.17 and
the standard deviation = 0.27
(2) * indicates stroke times which dirrer from the above mean by more than 2 standard deviations
(3) / indicates stroke times which dirrer from the above mean by more than 3 standard deviations
. -- _ - - - - - - _ _ _ _ _ . . _ _ _ _ . _ _ _- - _ _ __ - _ . - _ _ _ . _ _ _ - .
-
- . -
,
TABLE 2
CHRONOLOGICAL TABULATION Of CLOSURE STROKE TIMING FOR UNIT 1
VALVES 1-G16-F003, F004, TO19 AND F020
Date Stroke Times (Seconds) fo r Va l ves
F QR}_ ffRQ9 IO19 fy?J
1/2/88 2.81 3.16 3.20 2.90
11/2/87 2.70 2.25/ 2.16/ 1.88/
8/9/87 3.0 3.6* 3.3 3.1
6/20/87 End of Refueling Outage That Started 2/14/87
5/9/81 3.33 3.25 3.21 3.08
4/2G/87 3.80 3.34 3.46 3.28
2/5/87 2.79 3.07 3.13 2.99
11/6/87 2.87 2.80 3.15 2.76
8/5/86 2.96 3.09 3.34 3.20
5/6/86 2.18 3 . 0 *> 3.15 3.11
2/12/86 2.90 3.05 3.23 3.13
11/2/85 2.76 3.08 3.38 3.13
9/22/85 2.80 3.39 3.61* 3.53
Notes: (1) I')r the stroke times recorded in the first five valvo tests (20 values durin9 9/22/85 - 8/5/86) the mean = 3.13 and
the standaro deviation = 0.23
(2) * indicates stroke times which dirror f rom the above mean by more than 2 standard deviations
(3) / indicates stroke times which dirrer f rom the above mean by more than 3 standard deviations
_ _ _ _ _ _ - - . - _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . - _ _ - _ _ _ _ _ - -_ _ ._-.________J
- - - - - - - . -_----_--- _ _ - - - -
.
.
7
b
I
-
15
e
TABLE 3
COMPARIS0N OF STROKE TIMES OBTAINED BY FOLLOWING
PROCEDURE VERSUS CORRECT METHOD FOR
UNIT 1 VALVES 1-G16-F003, F004, F019 AND F020
Valve No. Stroke Time Measured (Seconds)
Per Procedure - (From Correct Method - (From
switch closure actuation switch closure actuation
until red light off) until green light on)
1G16-F003 1.38#, 0.96#, 1.16 2.76
1G16-F004 1.14 2.87
1G16-F019 1.16 3.02
1G16-F020 1.06 2.76
mean = 1.13 2.85
standard deviation = .05 .12
Note: # denotes the first two measurements at the start of testing.
The operator who performed the testing felt that they were not
sufficiently accurate and performed the third measurement which he
considered sati sf actory. The # values were not included in the
calculation of mean and standard deviation.
. -