ML20211A292
| ML20211A292 | |
| Person / Time | |
|---|---|
| Site: | Hatch  |
| Issue date: | 02/02/1987 |
| From: | Jape F, Witener H NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20211A236 | List: |
| References | |
| 50-321-86-39, 50-366-86-39, NUDOCS 8702190139 | |
| Download: ML20211A292 (8) | |
See also: IR 05000321/1986029
Text
.
p Kf20
UNITED STATES
g
9'o
NUCLEAR REGULATORY COMMISSION
y"
REGloN 11
3
g
j
101 MARIETTA STREET, N.W.
- '-
'*
ATL ANTA, GEORGI A 30323
\\...../
Report Nos.:
50-321/86-39 and 50-366/86-39
Licensee: Georgia Power Company
P. O. Box 4545
Atlanta, GA 30302
Docket Nos.:
50-321 and 50-366
License Nos.: DPR-57 and NPF-5
Facility Name: Hatch 1 and 2
Inspection Conducted: November 15-21, 1986
Inspector: N.[ hb
/ - dB - 8 7
H.~ L. Whitener
Date Signed
Approved by:
A
%
Ab/ h
'
"
'
Date Signed
F.- Jape, Section ~ Chief .
Engineering Branch-
-
" Division of Reactor Safety'~
SUMMARY
.
p,
'
~
Scope: -This routine, announced inspection was in the area of containment leak
"
rate testing.
Results: . One violation was identified - The overall integrated leakage rate for
the "as is", (also called "as found") containment condition could not be deter-
mined due .to ,the failure to measure <the change in leakage rate which resulted
from ' repairs and adjustments made to' the containment boundary prior to the
integrated (Type A) leak rate test'.
%2]%3kb
p
G
'
I
.
REPORT DETAILS
1.
Persons Contacted
Licensee Employees
- D. S. Read, Plant Support Manager
- L. Sumner, Manager of Operations
T. Powers, Manager of Engineering
- A. Fraser, Acting QA Site Manager
- R. D. Baker, Nuclear Licensing Manager, Corporate
G. Goode, Superintendent of Plant Engineering and Services
- A. Huber, Senior Plant Engineer
.1. C. McWhorter, Engineer, Southern Company
Other licensee employees contacted included leak rate test personnel.
Other Organizations
Bechtel - Leak Rate Consultants
R. Blum
M. Burgess
H. Hill
B. Patel
A. Salley
Volumetrics, Inc.
l
D. Peyvan, Project Engineer
NRC Resident Inspector
- G.
M. Nejfelt
- Attended exit interview
2.
Exit Interview
The inspection scope and findings were summarized on November 21, 1986,
with those persons indicated in paragraph 1 above. The inspector described
the areas inspected and discussed in detail the inspection findings.
No
dissenting comments were received from the licensee.
The following new
items were identified during this inspection.
Violation (50-366/86-39-01):
Failure to measure the change in leakage
rate due to repairs or adjustments to the containment boundary prior to
the Type A test makes the calculation of the "as found" containment
,
l
integrated leak rate indeterminable (paragraph 5.e).
l
,
<
.
2
Inspector Followup Item (50-366/86-39-02): Review the licensee's action to
identify and test the pressure restraining seals in the containment boundary
,
(paragraph 5.d).
The licensee did not identify as proprietary any of the materials provided
to or reviewed by the inspector during this inspection.
3.
Licensee Action on Previous Enforcement Matters
This subject was not addressed in the inspection.
4.
Unresolved Items
Unresolved items were not identified during this inspection.
5.
Containment Integrated Leak Rate Test - Unit 2 (70313, 70307, 70323)
The inspector reviewed and witnessed test activities to determine that the
primary containment integrated leak rate test (CILRT) was performed in
accordance with the requirements of Appendix J to 10 CFR 50, ANSI-N45.4,
test procedure 42SP-TET-003-2S " Prima ry Reactor Containment Integrated
Leakage Rate Test" and the criteria of BN-TOP-1, Revision 1 for a short
duration test.
Selected sampling of the licensee's activities which were inspected included:
(1) review of the test procedures to verify that the procedures were properly
approved and conformed with the regulatory requirements; (2) observation of
test performance to determine that test prerequisites were completed, special
equipment was installed, instrumentation was calibrated and appropriate data
were recorded; and (3) preliminary evaluation of leakage rate test results
to verify that leak rate limits were met.
Pertinent aspects are discussed in the following paragraphs.
!
a.
General Observations
The inspector witnessed and reviewed portions of the test preparation,
temperature stabilization and data processing during the period of
November 15-21, 1986.
(1) The test was conducted in accordance with an approved procedure.
Procedure changes and test discrepancies were properly documented
in the procedure.
(2) Test prerequisites selected for review were found to be completed.
(3) Selected plant systems required to maintain test control were
found to be operational.
(4) Special test instrumentation was reviewed and found to be installed
and calibrated.
,
_
.__
_ _ _ .
_ _ _ .
._
-
.
. _ _
- . . -
.
3
(5) Controls for preventing pressurized air sources inside containment
or externally pressurized penetrations were established in the
test procedure.
(6)
Instructions and documentation for venting, draining and isolation
of systems were established in the test procedure.
(7) Problems encountered during the test were described in the test
~
event log.
(8) A containment temperature survey was performed for the no forced
flow test conditions.
(9) Temperature, pressure, humidity, and flow data were recorded at
15-minute intervals.
Data were assembled and retained for final
evaluation and analysis by the licensee. A final leak rate test
report will be submitted to the Office of Nuclear Reactor Regula-
tion pursuant to Paragraph V of Appendix J to 10 CFR 50.
b.
Test Performance - Unit 2 (70313)
(1) Method
The Bechtel computer program used for data analysis has the
capability for Total Time analysis in accordance with the
requirements of BN-TOP-1, Revision 1, for a short duration test
and Mass Point-Linear Regression analysis in accordance with the
recommendations of ANSI /ANS 56.8-1981 for a 24-hour test.
The
Appendix J acceptance limit of 0.75 La was met for the 95% Upper
Confidence Limit (UCL) for both the Total Time and Mass Point
analysis techniques.
A supplemental test was performed in
accordance with the specifications of Appendix C to ANSI N45.4 -
1972Property "ANSI code" (as page type) with input value "ANSI N45.4 -</br></br>1972" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process. and BN-TOP-1, Revision 1 - 1972.
(2) Test Description
The absolute method as defined in ANSI N45.4-1972 was used in
determining the containment leakage rate.
Values bounding the
test conditions were as follows:
Containment Volume
257566 cubic feet
Accident Pressure (Pa)
57.5 Psig
Maximum Allowable Leakage (La)
1.2 wt.% per Day
System conditions for performance of the integrated leak rate test
were as follows:
(a) Reactor Vessel
Water level was between 53 and 60 inches,
and vented to the containment
'ww
-
P'
a
e
- - ' ' ' '
+ ' - -
t
'
.
4
(b) RHR System
One loop was operating in shutdown
cooling mode
(c) Containment Fans
Secured
Containment pressurization was initiated at 9:10 p.m.,
on
November 18, 1986. At 3:00 a.m.,
on November 19; pressurization
was secured; test pressure of Pa was verified; and, the stabiliza-
tion period was started. Temperature stabilization criteria were
met at 7:00 a.m., but the start of the Type A leak rate measurement
was deferred due to excessive containment leakage. The leakage
path was identified as a packing leak on valve No. 2T49-F001A and
was corrected to an acceptable Type A test limit. The Type A leak
rate measurement start time was established at 11:30 p.m.,
November 19, 1986. A twelve hour short duration test in accordance
with the criteria of BN-TOP-1, Revision 1 and Appendix J to 10 CFR 50 was performed followed by a six hour supplemental (verification)
test which was successfully completed at 6:45 p.m., on November 20,
1986.
c.
Test Results - Unit 2
(1) Type A Test
The Technical Specification allowable leakage (La) for Hatch 2 is
1.2 wt.% of the containment mass at accident pressure (Pa) per
day. The acceptable test leak rate limit of 0.75 La is therefore
0.9 wt.% per day. The calculated leakage rate and 95% UCL for the
Total Time and Mass Point Analyses are shown below for three
analyses performed on test data taken at 15 minute intervals
in the period 11:30 p.m. , November 19 to 11:30 a.m. , November 20,
1986. These are:
All data included in the analysis (49 data
Case I
-
sets analyzed in period 11:30 p.m., on 11/19
to 11:30 a.m., on 11/20).
Case II
-
Three data sets (Nos.
3, 6 and 7) deleted
frein the analysis (46 data sets analyzed in
period 11:30 p.m. on 11/19 to 11:30 a.m. , on
11/20).
Case III -
Test start time advanced to 1:15 a.m.,
on
11/20 which deletes the first seven data sets
'
(42 data sets analyzed in period 1:15 a.m. ,
to 11:30 a.m., on 11/20).
l
I
.
5
Mass-Point
Total Time
(wt. %/ day)
(wt.%/ day)
LSQF
UCL
LSQF
UCL
Case I
0.498
0.527
0.644
1.64
Case II
0.46
0.475
0.481
0.589
Case III
0.437
0.453
0.419
0.614
"LSQF" is the least square fit to the data and "UCL" is the
calculated 95% upper confidence limit for the respective
analysis method and case.
Review of test conditions indicated that Case II was the most
representative leak rate.
Case I contained three data sets
(Nos. 3, 6, and 7) which appeared to be affected by external noise
rather than actual changes in the containment atmospheric conditions.
Case III represents deletion of four data sets which appear to be
valid in addition to data sets 3, 6, and 7 due to advancing the
test start time from 11:30 p.m. on November 19 to the first data
set after the noise affected data which was 1:15 a.m.,
on
November 20.
Case III is included here principally to show
the consistency of the leak rate trend with respect to time.
Exclusion of data sets 3, 6, and 7 from the leak rate analysis was
based on the following information:
(a) The character of the perturbations, affecting principally the
RTD measurements and to a lesser extent the dew cells for
data points 3, 6 and 7, gave the appearance of external noise
,
in that there was a sudden departure from and return to a
-
.
stabl a trend; the deflections occurred at the same time and
in the same direction on all RTDs; and, the deflections were
I
about five times larger than the typical random noise.
$
(b) No c,orresponding change in total containment pressure
occurred.
Pressure instrumentation has no RTDs associated
'
with the circuitry. The containment pressure showed a stable
l
decreasing trend of 0.002 psi per data set.
An actual
.
'
I
decrease in temperature of 0.3 F would have resulted in a
I
pressure decrease of 0.04 psi; an order of magnitude larger
f
,
t
than that observed.
'
,
(c) The probable noise source was traced to the use of radio
communicators near the containment and instrument room.
A
'
g
review of the security log showed that radios had been used
I
in the vicinity of the instrumentation at the time the
perturbations of the RTD signals occurred.
Subsequent to
7
the test, the licensee verified that keying a radio in the
r
-
t
vicinity of the RTD instrumentation does induce noise in the
[
data similar to that observed early in the test.
l
l
. _-
...
-
_
-
.
6
The inspector concluded that the licensee had a reasonable basis
for . deletion of data points 3, 6, and 7 from the total time
analysis. The resulting total time 95% UCL in accordance with
BN-TOP-1, Revision 1, of 0.59 wt.%/ day is well below the acceptable
limit of 0.9 wt.%/ day.
(2) Supplemental Test
Appendix J requires that a supplemental test be performed to
verify the accuracy of the Type A test and the ability of the
CILRT instrumentation to measure a change in leak rate.
An
acceptable supplemental test method is described in Appendix C
of ANSI-N45.4 - 1972, as follows:
A known leak rate (Lo) is imposed on the containment and the
measured composite leak rate (Lc) must equal, within 10.25 La,
the sum of the measured Type A leak rate (Lam) plus the known
leak rate (Lo).
The acceptance criteria is expressed as:
Lo + Lam - 0.25 La < Lc < Lo + Lam + 0.25 La
A six hour supplemental test was performed by the imposed leak
rate method described in Appendix C to ANSI-N45.4-1972 and in
accordance with the requirements of BN-TOP-1, Rev. 1.
The
following values in units of wt. %/ day were obtained using both
mass point and total time analyses.
Mass Point
Total Time
Lam
0.479
0.460
Lc
1.623
1.643
Lo
1.153
1.153
0.25 La
0.3
0.3
Using the above values, Lc must fall within the limits in wt.%/ day
as follows:
,
1
Mass Point
Total Time
!
Upper Limit
1.913
1.932
Lc
1.623
1.643
Lower Limit
1.313
1.332
i
As indicated, Lc is within the specified limits for both analysis
techniques.
!
l
l
,
l
.
7
-d.
Containment Integrated Leak Rate Test - Status
At 7:00 a.m., on November 19, 1986, the four hour temperature stabili-
zation criteria were met. At this time the containment was leaking in
excess of the allowable 0.75 La (0.9 wt.%/ day) limit. The start of the
Type A test was delayed and a search for the leakage path was initiated.
In the period 7:00 a.m.,
to 2:00 p.m., on November 19, the containment
leak rate tended to stabilize at 1.1 wt.%/ day with a UCL of 1.7 wt.%/ day
by total time calculations. The major source of leakage was identified
as a packing leak on valve No. 2T49-F001A. This valve is outside of
the containment and inboard from the isolation valve of the hydrogen
recombiner system.
The measured UCL of 1.7 wt.%/ day and leak rate of 1.1 wt.%/ day both
exceed the Appendix J acceptance limit of 0.9 wt.%/ day. Consequently,
the CILRT was classified as a failed Type A test in accordance with
Paragraph III.A.1.(a) of Appendix J to 10 CFR 50. This is considered
the first failed Type A test on Unit 2.
Pursuant to the requirements
of Paragraph III. A.6.
of Appendix J to 10 CFR 50, the licensee will
submit the test schedule applicable to subsequent Type A tests for NRC
review and approval. Based on an integrated leak rate established over
a period of about seven hours before repair of valve 2T49-F001A packing
leakage and the integrated leak rate over a period of about 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />
after the repair, the corrected leakage was 0.62 wt.%/ day from the
least square fit data and 1.1 wt.%/ day from the UCL data using the
total time analysis method.
Excessive leakage through the containment boundary raises a question
regarding the adequacy of the local leak rate testing program.
For
the hydrogen recombiner system, pressure for the Type C test is applied
between 2T49-F001A and the isolation valve in order to test the isola-
tion valve in the correct direction.
In this test configuration the
valve packing on 2T49-F001A is not exposed to local test pressure. The
rroblem was identified and corrected as a result of the Type A test;
therefore, the hydrogen recombiner system leakage is considered accept-
able for the current operating cycle for Unit 2.
The licensee will look at this problem in an ongoing leak rate test
evaluation program. This matter is identified for followup inspection
as follows:
IFI (50-366/86-39-02):
Review the licensee's action to identify
and test all pressure restraining seals in the containment pressure
boundary,
e.
Review Of the "As Found" Containment Leak Rate
The inspector reviewed the controls established by the licensee to
ensure that where repairs or adjustments were made to the containment
boundary prior to the Type A test, the change in leak rate due to these
repairs or adjustments was quantified in order to determine the "as
found" containment integrated leak rate.
__
_.
.
8
The inspection findings in this area are discussed below:
Unit 2
From discussions with management and personnel involved with the local
leak rate test program, the inspector determined that the licensee had
implemented controls to obtain local leakage rate measurements before
any repairs or adjustments to the containment boundary.
The outage
Work Planning Group coordinated the control through the maintenance
work order (MWO) system.
For those MW0s which affected containment
leakage barriers an instruction sheet was inserted into the work
package which required verification that a local leak rate was
performed prior to making repairs or adjustments. There were three
instances where a breakdown in implementation of the controls occurred.
As a result, repairs or adjustments were made to the containment
boundary prior to the Type A test without determining the change in
leak rate due to these repairs or adjustments. The conditions were as
follows:
(1) Two equipment hatches, penetrations IA and 18, were removed
without first obtaining the as found local leak rate for the
seals. On reinstallation of these hatches to perform the Type A
test, new seals were installed. Consequently, the change in leak
rate due to this repair could not be determined.
(2) A limitorque valve actuator on valve 2P42-F052 was reworked and
the seating torque values reset without first obtaining the local
leak rate.
Consequently, the change in leak rate due to this
repair and adjustment could not be determined.
The regulations regarding determination of the "as is" (also called "as
found") containment integrated leak rate are:
Paragraph III. A.1(a) of Appendix J to 10 CFR 50 and Paragraph 4.2 of
ANSI-N45.4 require, in part, that no repairs or adjustments shall be
made to the containment boundary prior to the Type A test in order to
determine the "as found" leakage condition of the containment.
Paragraph III.A.1(a)
in conjunction with Paragraph III.a.1(b) of
Appendix J to 10 CFR 50, permit repairs and adjustments to the
containment boundary prior to the Type A test, as necessary, provided:
(1) the change in leak rate resulting from any repairs or adjustments
is determined using local leakage testing methods; and (2) the
corrective action taken, the change in leak rate determined by local
tests, and the "as found" containment integrated leak rate determined
from the local tests and Type A test results are obtained and are
included in the report to the Commission.
Failure to determine the change in leak rate resulting from repairs or
adjustments to the above containment boundary components was identified
as a violation as follows:
m
._
~
,...
9
Vio1'ation (50-366/86-39-01): Contrary to the above' regulations, during
the 1986 Unit 2 refueling outage, repairs or adjustments were made to-
equipment hatches IA and 18 and to valve 2P42-F052' prior to the Type A
te st .' The- change in leakage rate as a result of .these repairs or
adjustments was not determined using local leakage test methods.
Failure to determine the change in leakage rate resulting from these
repairs or adjustments to the containment boundary renders 'the cal-
culation of the "as found" containment integrated-leak rate indeter-
minable.
While failure to determine the "as found" containment leak rate with
the plant in a shutdown condition is not an immediate safety problem,
the "as found" leak rate information is necessary to permit.the licen-
see to periodically assess the ability of the containment isolation
system to perform its safety function over a surveillance cycle.
Unit 1
The inspector reviewed the final calculation of the "as found" contain-
ment integrated leak rate for the Unit 1 Type A test performed
April 18-19, 1986.
The licensee's analysis shows . an "as found"
containment integrated leak rate of 1.538 wt.%/ day. This value exceeds
the . Appendix J to 10 CFR 50 limit of 0.75 La which is 0.9 wt.%/ day.
The inspector concluded that the April 1986 Unit 1 "as found" contain-
ment integrated leak rate test is a " failed" test. This is the first
failed test on Unit 1.
Pursuant to Paragraph III.A.6(a) of Appendix J
to 10 CFR 50, the licensee's proposed Type A test schedule and correc-
tive actions will be reviewed during review of the . leak rate test
report in the Region Office.
!
L
!