IR 05000275/1992017
| ML20059C858 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 08/24/1993 |
| From: | Johnson P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| References | |
| OLA-2-I-MFP-102, NUDOCS 9401060147 | |
| Download: ML20059C858 (16) | |
Text
T - a/ r p # " ~ '
f4FfOHI6IT10).-
dD ~
M
4 c.,39.
/o 2_
g,4g q.,,
m pggg 7/z4/93
'
~
(2eph.
.,
U.S. NUCLEAR REGULATORY COMMISSION
REGION V
'93 OC' M P 6 :21 Report Nos:
50-275/92-17 and 50-323/92-17
'
Docket Nos:
50-275 and 50-323 License Nos:
Pacific Gas and Electric Ccmoany 77 Beale Street, Room 1451 San Francisco, California 94106 Facility Name:
Diablo Canyon Units 1 and 2 Inspected at:
Diablo Canyon Site, San Luis Obispo County, California Inspection Conducted:
March 17 through April 27, 1992 Inspeci.as:
H. J. Wong, Senior " sident Inspector P. J. Morrill, Senior Resident Inspector Approved by:
k 72-Reacter@hnson, Chief P. H. /'
Date Signed Projects Section 1 Inspection Summary:
Inspection on March 17 throuah April 27. 1992 (Recort Numbers 50-275/92-17 and 50-323/92-17)
Areas inspected:
Special inspection of the circumstances surrounding deri-cien~ es discovered related to the Unit 1 and Unit 2 containment fan cooler units (CFCUs) and the associated backdraft dampers.
These deficiencies included CFCU counter-rotation and backdraft dampers stuck in the open position.
Inspection Procedures 37051, 50073, 61726, 62703 and 92701 were used as guidance during this inspection.
Safety Issues Management System (SIMS) Items:
None Results:
General Conclusions and Specific Findinos:
The history of problems associated with the containment fan cooler units (CFCUs) indicates tnat design and/or maintenance problems have existed for a long time.
The NRC inspectors have observed that several
'
conditions have allowed the current problems to evolve to the three inoperable CFCU dampers found in February 1992:
-
9401060147 930824 ADOCK0500g5 PDR C
-
_
-
-
_
-
.
--
.
--
- - _ -
D-2-
'
,
.
First, despite numerous problems over the last several years, the
quality organization did not pursue the containment ventilation system, or the CFCUs in particular, as an area requiring further attention.
Even after major deficiencies were found in Unit 1 (three CFCU dampers inoperable in February 1992), a comprehensive assessment of Unit 2 dampers was not initiated by QA or QC.
The quality organization appeared to be following up on problems, rather than identifying them.
Secondly, the maintenance organization attempted to resolve the
problem of reverse rotation and broken bolts without the involvement of corporate or system engineering personnel.
System engineers
appeared to be responsible for system design and current status.
Between the maintenance and engineering organizations, however, there was an absence of a broad perspective of the system's performance and the root causes of problems.
PG&E management did not appear to have defined clear performance expectations for system engineers, maintenance engineers, corporate engineering, and mainte-nance personnel.
Thirdly, internal communications and attention to detail appeared to
be a significant problem for Diablo Canyon.
The extra washers, along with counterwights which were installed too loose or too tight, indicated that maintenance personnel did not follow the design drawings in accomplishing work.
The existence of tight counterweights in Unit 2 was identified in January 1992, but was not communicated to PG&E management until April, after reverse rotation of CFCU 2-2 prompted the licensee to conduct additional inspections.
.
In summary, the specific problems appear to have been technically resolved.
However, PG&E management should address the potential for
,
similar organizational and personal performance problems which could result in future deficiencies.
Sianificant Safety Matters:
!
The inoperability of three Unit 1 CFCus, identified by the licensee in February 1992, was due to work done during the Unit 1 outage which ended i
in March 1991.
Consequently, these devices may have been inoperable
during ten months of Modes 1, 2, and 3 operation.
This condition appeared to violate the Technical Specifications and did not satisfy the plant's design basis, although it appears that operator action would have precluded overpressurizing the containment.
Summary of Violations:
Three apparent violations were identified:
(1) operation of Unit 1 in
,
'
Modes 1, 2, and 3 with three inoperable CFCUs between March 27, 1991 and February 22, 1992, contrary to Technical Specification (TS) 3.6.2.3.a; i
(2) failure to take appropriate corrective actions after observing i
'
reverse rotation of Unit 1 CFCOs on March 25, 1991; and (3) failure to i
follow approved procedures (and Work order instructions) while inspecting I
Unit 2 CFCU dampers in February 1992.
.
.
.
.
-
_ J
-
w
--w---+
"' ~
O Y
w f
f
[
l i
.
b
g b
$ d '/ S C Lf_ c,rewyan n 3 Ff'-/0 2 Dce et t:
s
. fa;t;&w FLfcydC QQ__-
< - -
,~ u.
'
__ _ _ _ ___renmov'
- rr
.cg f
-
. _
_
'
',,
~._,4_.
__ _ __,, k ( E SI5 3_ _
_
N-
~-
DM'a :nbin $$Ytj-... _ _...
E415
- jnD$-93
_
es __
_
,.
,
_
Peptttt !
lQ t'
l
,
.
.
. - -.. _ -
_. -. -.
._
-- -..- - --
_
- _
_
.
.
-
.
DETAILS
,
1.
Persons contacted Pacific Gas and Electric Comoany
- G. M. Rueger, Senior Vice President and General Manager, Nuclear Power Generation Business Unit
-
J. D. Townsend, Vice President and Plant Manager, Diablo Canyon Operations
- W. H. Fujimoto, Vice President, Nuclear Technical Services
- D. B. Miklush, Manager, Operations Services
- M. J. Angus, Manager, Technical Services
- B. W. Giffin, Manager, Maintenance Services
- W. D. Barkhuff, Quality Control Director R. P. Powers, Mechanical Maintenance Director
,
- D. A. Taggart, Quality Per?crmance and Assessment Director
- T. L. Grebel, Regulatory Compliance Supervisor H. J. Phillips, Electrical Maintenance Director
- R. C. Anderson, MLiager, Nuclear Engineering and Construction Services
- M. R. Tresler, Project Engineer, Nuclear Engineering and Construction Services J. A. Shoulders, Onsite Project Engineering Group Manager
,
S. R. Fridley, Operations Director
- J. J. Griffin, Senior Engineer, Regulatory Compliance
- J. B. Hoch, Manager, Nuclear Safety and Regulatory Affairs
- T. A. Moulia, Assistant to Vice President, Diablo Canyon Operations
- C. A. Dougherty, Quality Assurance Senior Supervisor
,
- J. E. Tompkins, Nuclear Safety and Regulatory Affairs Director Nuclear Reaulatory Commission. Reaion V
- S. A. Richards, Chief, Reactor Projects Branch
- Denotes those attending the exit interview on March 19, 1992.
Contain-ment fan cooler units (CFCUs) were also discussed during management meetings conducted on April 2 and 21, 1992 The inspectors interviewed several other licensee employees including
shift supervisors, shift foremen, reactor and auxiliary operators, maintenance personnel, plant technicians and engineers, and quality i
assurance personnel.
2.
Containment Fan Cooler Unit (CFCU) Damners Not Fully Closina (92701)
a.
Backaround Inspection Report Nos. 50-275 and 50-323/92-01 described deficien-cies related to the backdraft dampers for CFCUs in Units I and 2.
.
Those deficiencies included the loosening of counterweights in Unit 2 and the misassembly of some damper closure mechanisms in Unit 1.
The deficiencies were first identified by the licensee on January 22, 1992. However, an NRC review of licensee recc Js
--,
-,
.,
-
r-
--, -
-.,
-., - -
.
-2-indicated that problems with the CFCU backdraft dampers existed at least as early as 1986.
In addition, at the end of the Unit I refueling outage (March 1991), approximately 20 broken bolts were found in the dampers, but no significant follow-up was performed.
A summary of significant historical events regarding these dampers is presented below.
b.
CFCU Backdraft Damoer History 2/2/86 CFCU 2-2 was observed to rotate backwards when other fans were running (AR A0016334); corrective actions were to install new damper operating parts and retest the CFCV.
1/18/89 Reverse rotation of CFCUs when fans were not operating was found to have been observed during previous outages in Units 1 and 2; dampers were not closing. Design changes for the linkage mechanism were recommended to resolve the problem. (AR A0137870)
8/20/9u The Plant Staff Review Committee approved implementation of a design change to add " helper springs" to assist the closure of Unit 2 CFCU backdraft dampers by Design Change Package (DCP) H-44664 (approval for Unit 1 dampers by DCP H-43664 was given on 10/20/90).
2/91 - 3/91 Maintenance was performed on all five Unit 1 CFCU back-draft dampers to inspect and replace closure mechanism parts; no Quality Control (QC) hc'd points were specified.
During these maintenance activities, damper linkage bars were installed incorrectly on CFCUs 1-1 and 1-5.
3/23/91 Broken bolts, which attach damper blades to their shafts, were found on erCU backdraft damper
'-2.
Corrective actions were to replace bol+s and inspect other dampers in each unit at the next refueling outage.
3/25/91 During an inspection prior to establishing containment integrity in Unit 1, the licensee identified that CFCU 1-5 was rotating backwards; the licensee's evaluation was that "... this is not an out of normal condition for the fan cooler.
Based on this evaluation, no further action is required and the CFCU is operable." This evaluation was performed by maintenance engineering and mechanical and electrical maintenance personnel without involvement by Nuclear Engineering and Construction
.
Services (NECS) or the On-site Plant Engineering Group (OPEG).
1/22/92 During a routine Unit 2 containment entry, licensee
personnel identified that backdraft damper counter-
'
weights had fallen off dampers associated with CFCUs 2-1
.
.
. _
_.-
_
. _ _ _ __
.
.-
- _ _ _ _
.
.
-3-c and 2-5.
Other counterweights were observed to be too loose.
Engineering (NECS/0 PEG) performed an evaluation of the missing counterweights.
The evaluation assumed that the time delays in sequencing of the CFCUs would be sufficient to allow any CFCU rotating in the reverse-direction "... to coast down to zero speed to prevent potential damage to the CFCU during start up on low speed."
1/23/92 -
Additional NECS review identified that the previous conclusion was in error in that a CFCU may not coast
.
down to zero speed if it is running in reverse at high
,
speed.
.
Licensee personnel performed walkdowns of Unit 1 CFCU dampers and determined that counterweights were i
attached, but that some were installed too tightly.
'
1/24/32 NECS evaltation concluded that operability of CFCU backdraft dampers and CFCUs were maintained based on the observation of full closure of CFCU damper 2-1 during testing (the one with the most missing counterweights).
This conclusion relied on the appearance of proper functioning of the damper.
It was not based on a visual observation of damper operation. Mechanical maintenance
.
!
personnel identified tight counterweights on Unit 2 CFCUs (work order C0095999).
This information was not communicated to maintenance engineers or to engineering
personnel.
1/29/92 Observation of Unit 2 CFCU damper counterweights'identi-fied two additional sets of counterweights about to ccme off CFCU 2-1.
These additional counterweights were removed from damper 2-1.
i
!
1/31/92 Inspection of CFCU 2-1 documented that the fan was rotating in the reverse direction at approximately 30 rpm.
The NECS technical evaluation concluded that reverse rotation of up to 120 rpm would be acceptable.
,
2/7/92 NECS's evaluation of too-tight counterweights on Unit 1
!
CFCU dampers concluded that the CFCUs were not precluded from performing their safety related functions.
This i
was based on assuming that failure of the damper to
close was not credible due to counterweight installation
'
with the dampers in the closed position and the absence of identified reports of reverse rotation of non-
._. _
_
-.
.....
.
_. -.
-.
=..
-
-
-
.
..
.-
.
t
-4-
'
operating fans.
This conclusion was reached without confirming by visual inspection whether reverse rotation was actually occurring with other CFCUs.
Verification by inspection was planned.
2/12/92 Licensee personnel discussed the status of the CFCU problems with NRC Region V personnel.
Licensee managers.
stated that the problems with Unit I and 2 were resolvod and that the dampers were operable.
Region V personnel questioned whether anyone had visually examined Unit 1 CFCOs to determine if reverse rotation was occurring.
z/13/92 The licensee identified that reverse rotation of Unit 1
[
,
'
CFCUs was occurring in CFCUs 1-1, 1-2, and 1-5 (up to 143 rpm).
The subsequent NECS evaluation of reverse rotation concluded that up to 180 rpm would not prevent a CFCU fan from starting.
The licensee characterized the dam-pers in Unit I as " leaky", while in fact these dampers were later found to be stuck in the full open position.
l Part of the bases Nt;S utilized relied on a loss of coolant accident (LOCA) pre.;sure-time curve to conclude that the peak LOCA pressure would occur while a CFCU fan would still be coasting down from normal operation.
This LOCA pressure curve appeared to be a conservative approximation (with regard to ability to restart CFCUs).
2/19/92 The licensee identified that Unit 1 CFCU dampers 1-1, 1-2, and 1-5 were in a stuck open condition (damper linkage bars were found to have been incorintly
,
installed during previous assembly of dampers 1-1 and 1-5).
Damper 1-1 had to be struck with a mallet to
"
'
close the damper (AR A0258712). A one-hour, non-
'
emergency report was made to the NRC based on three
'
<
CFCUs being inoperable.
The reported stated that a design basis accident (LOCA with loss of a vital bus)
.
would place the plant in an unanalyzed condition.
A l
minimum of two CFCUs and one train of the containment spray system are required to meet design requirements.
,
I 2/20/92 The licensee prepared plans to confirm the operability
of CFCU 1-1 and to repair CFCUs 1-2 and 1-5 dampers.
The NRC inspector observed that the damper for CFCU 1-1 opened on starting of the fan, but had to be struck with
!
a hammer to close.
The licensee concluded that the damper could be considered operable when left in the closed position with the associated CFCU not running
(since the damper could complete its safety function
following a LOCA).
'
2/21/92 Scaffolding was erected for CFCU 1-2 repair work.
J
~. - - - -, - _ -.
.
--.-
-
. -
.. -.
-
.,
.. -
-
,
--.
,
,
. -.
. - - -
._.
-
-
- -
.
.-
i
- 5 --
y 2/22/92 Repairs were made to the external components of CFCU 1-2 damper.
CFCU 1-2 was restored to operable condition after testing.
2/23/92 Repairs were made to the external components of CFCU damper 1-5.
CFCU 1-5 was restored to operable condition after testing.
2/26/92 Repairs were made to the external components of CFCU damper 1-1.
CFCU 1-1 was restored to operable condition
,
after testing.
3/7-8/92 Following a Unit I trip on March 6, internal and external inspections were completed for all Unit 1 CFCUs and bolting repairs were made as necessary.
Post-
'
modification tests were conducted on each CFCU to assure there was no reverse rotation of the fan when it was shut down.
Westinghouse confirmed with the licensee that 150 RPM reverse rotation would not mechanically or
,
-lectrically damage the CFCU on a start.
3/24-25/92 After a Unit 2 shutdown on March 23 (due to a turbine stop valve problem) CFCU backdraft missing weights were replaced on damper 2-1 (Work Order 098184).
The damper was repaired and inspected by QC.
4/2/92 During an NRC management meeting, the licensee discussed
!
the CFCU problems and actions taken to date. The licensee stated that all five Unit 1 CFCU backdraft dampers and damper 2-1 of Unit 2 had been overhauled and inspected externally by QC.
All Unit 2 d mpers had been externally inspected and found to be assembled in accordance with design.
4/8/92 Monthly surveillance inspection M-51A, required as a compensatory measure by the licensee's Operability Evaluation (OE) 92-01, was completed by the licensee to identify any Unit 1 CFCUs rotating backwards or dampers which did not shut. No reverse rotation or open dampers were observed in Unit 1.
4/15/92 Monthly surveillance inspection M-51A of Unit 2 CFCUs and dampers was completed by the licensee.
An NRC inspector accompanied the licensee personnel conducting these inspections.
CFCU 2-2 was observed to be rotating backwards at 132 RPM.
Based on visual inspections (using binoculars and flashlights where necessary), all dampers appeared to close when the associated CFCU was shut down.
4/16/92 The licensee completed an inspection of damper 2-2 to determine why the CFCU had rotated in reverse.
No cause of counter-rotation was identified; however, the licen-see did find that most counterweights were installed too
_
_._
_
_
..
._
-- - __.
_
-
-
_~
-
_-
.
-.-
.-. -
-
.-
t
.
-6-
.
tightly in the counterweight arm slots.
This condition was contrary to the approved design and was not consistent with previous licensee inspection results.
4/17-20/92 The licensee conducted additional inspections of back-draft dampers 2-2, 2-3, 2-4, and 2-5.
The weights on damper 2-4 were also found to be too tight and washers which were not in accordance with the approved design were found to be installed in dampers 2-2, 2-3, 2-4, and 2-5.
These findings demonstrated that previous inspec-tions of the Unit 2 backdraft dampers, completed by licensee personnel, had not been thorough and that the associated documentation was incorrect.
.
4/20/92 The licensee met with the NRC in Region V to discuss Quality Assurance issues associated with the CFCUs and to review the improper actions taken by their personnel
'
regarding the inspections of Unit 2 backdraft dampers.
This meeting will be documented in NRC Inspection Report No. 50-275 and 50-323/92-13.
1/21-24/92 The licensee erected scaffolding in contcinment and overhauled o'ackdraft dampers 2-2 and 2-4.
On April 22,
wi.ile working on backdraft damper 2-2, the licensee found binding of one blade due to rust and wear.
Licensee personnel theorized that this binding, combined with the installation of washers, was likely to have caused the damper to stick partially open and thereby have caused the reverse rotation observed on April 15.
This sticking would not have been easily detected by the M-51A surveillance, since the damper had to be observed from a distance of about thirty feet through binoculars.
4/24-25/92 An NRC inspector examined backdraft damper 2-4 to verify installation consiste-t with the design drawings (American Warming and Ventilating Dra.iings SHW-0-9098, 9099, and 81010-001-000; PG&E documents 663079, Sheet 37, Revision 6; Sheet 38, Revision 3; and Sheet 75,
.
Revision 2; and Design Change Package DCP H-44664, Revision 0).
He observed that with the damper closed, half of the counter weights were about 1/16 of an inch above the bumper blocks.
The required clearance (if any) was not stated in the design drawings.
,
The inspector subsequently observed the licensee's performance of a functional test of this damper.
This test measured the times required for the dampers to open and close, and visually checked damper operation when the associated fan was started and stopped.
Backdraft damper 2-2 was observed to open and close smoothly over a period of several seconds.
The inspector noted that even though the counterweights did not contact the
bumper blocks, the damper did not appear to be over-stressed by normal operation.
He inquired regarding the I
-, -. - -,
,., -
.
.
-.
._. -
- - - -. - -.
_ -.- -.
_=
.
.
.
.
-7-s licensee's evaluation of the condition.
NECS concluded that the observed gaps were small enough and the components flexible enough, such that the counterweights would be supported by the bumper during a LOCA and that the inertia of the weights would be absorbed by the blocks.
4/25-27/92 The licensee completed rework and testing of the Unit 2 backdraft dampers.
Licensee personnel prepared plans to completely rework the backdraft dampers in both units during their next refueling outages (IR5 and 2R5).
c.
Ir,soection Findinos (1)
Plant Ooeration Outside the Desian Bases In January and February 1992, the licensee identified that Unit 1 CFCU backdraft damper opera:ing mechanisms had been assembled incorrectly during the Unit I refueling outage which occurred in February and March 1991. The work conducted on the CFCU backdraft dampers during this refueling outage included disassembly, cleaning, inspection, and reassembly of the d.. per operatino mechanisms of all five backdraft dampers (Work Orders R063770, R0064110, 20064103, R0063296, and R0063768).
The reassembly deficiencies discovered in Unit I during January and February 1992 include those listed below and were also described in NCR DCO-92-MM-N007:
Counterweights were installed without sufficient clearance
for free movement in a slotted hole in the lever arm (all Unit 1 CFCUs).
,
Damper linkage bars were assembled in a reversed
configuration (CFCU 1-5).
The linkas he was placed on the wrong side of the
linkage arm, causing binding of the damper (CFCUs 1-1 and 1-5).
Washers were used although none were called for in design
drawings (CFCU 1-2).
An NECS engi: eer gave improper directions during the refueling outage tc install the counterweights tightly (AR A0256869),
although design drawings called for a loose connection to allow movement of the weights in the lever arm slot.
This direction was given apparently to prevent loosening during operation of the CFCUs.
Although the work orders associated with the CFCU backdraft dampers specified that post-maintenance testing was to be performed on the dampers, these tests failed to identify the assembly problems with the dampers.
The specified post-maintenance tests appeared appropriate and included a back-
._ -_
_ - _ _ _.
.-
.
..
.--
_
t-8-
-
leakage check of the dampers during operation of the redundant fan, plus verification that the shutdown fan was not " free-wheeling." The steps for post-maintenance testing had been initialed; however, it appears that instructions given to the contractor personnel performing the post maintenance testing were inadequate to assure that the testing was performed properly.
Although there was no QC involvement in the backdraft damper work during the refueling outage, this was in accordance with the established QC inspection program.
Some of the criteria used to determine which activities would receive QC inspection included the complexity of the equipment, the need for special controls, the degree to which functional compliance can be demonstrated by independent verification or test, and the quality history of the equipment.
QC had determined during review of the backdraft damper work orders that no QC inspec-tion was necessary. The licensee considered this assessment to be appropriate based on the history of the dampers and work activities associated with the dampers at the time of the refueling outage.
The Unit 1 backdraft dampers en CFCUs 1-1,1-2 and 1-5 were found to be stuck in the open condition on February 19, 1992.
The last time that work had been performed on these dampers was during the Unit I refueling outage which ended on March 27,
.
1991 when Unit 1 entered Mode 4.
It appears that the work
'
performed during that refueling outage was deficient (e.g.,
extra washers installed, damper linkage bars incorrectly
!
installed) and caused the dampers to stick open and therefore be inoperable after the end of the outage.
.
FSAR Section 6.2.2.2.1.2.7 states that the backdraft dampers
"close on any reverse flow to prevent a pressure surge from damaging the fan motor assembly." A backdraft damper stuck in the open position could cause the CFCU to be inoperable, in that in the event of a LOCA, a pressure wave traveling through the ducting would not be stopped by the stuck open damper and would cause excessive forces on the fan motor trying to start after initiation on a safety injection signal.
The pressure wave would place a force on the fan in a reverse direction and on a fan start signal, the fan could potentially trip on overload due to the excessive forces placed on the fan motor.
Technical Specifications Limiting Condition of Operation (TS LCO) 3.6.2.3 states that in Modes 1, 2, 3, and 4 the contain-ment cooling system shall be operable with either:
(a) three groups of electrically independent CFCUs with a minimum of one unit in each group, or (b) at least two groups of electrically independent CFCUs with a minimum of two units in each group.
With CFCUs 1-1,1-2, and 1-5 inoperable due to the associated backdraft dampers being stuck open, the operable CFCUs consti-tuted only two groups of electrically independent CFCUs with one unit in each group.
TS LCO Action Statement 3.6.2.3.a l
. __
-
.
..
-
..-
-9-
.y states that with one group of the required CFCUs inoperable and both containment spray systems operable, restore the inoperable group of CFCUs to operable status within 7 days, or be in hot standby within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in cold shutdown within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />.
The licensee continued Unit 1 plant operations in Modes 1, 2, 3, and 4 from March 27, 1991 until February 22, 1992 with three inoperable CFCUs, in apparent violation of TS LC0 Action Statement 3.6.2.3.a.
(Apparent Violation 275/92-17-01).
The containment design pressure ir 47 psig and is based on the operation of one train of the containment spray system and two CFCUs.
Under the design basis accident conditions of a LOCA, assuming a single active failure (loss of a vital electrical bus), and with three CFCUs inoperable due to stuck open back-draft dampers, the plant could have been faced with only one train of the containment spray system and one CFCU.
To es.luate the effect of this condition on the plant's design basis, the licensee performed analyses of containment pressure transients for the combination of a LOCA, a failure of one elec.rical bus, and three increrable CFCils.
Three enveloping cases were considered:
(1) one CFCU and one Containment Spray (CS) loop operating during injection, followed by just the CFCU during recirculation; (2) one CFCU and one CS loop operating during injection, with a second CFCU started after 42 minutes, and no CS during recirculation; and (3) one CFCU and one CS loop operating during injection, with a second CFCU started after 55 minutes, and with no CS during recirculation.
The first case represents no operator action to restore a CFCU, while the other two cases show the effect of operator action to restore one CFCU.
These analyses were reportedly based on the current licensing basis decay heat and accident models.
Reportedly, newer decay heat curves or are.ident models would have given more favorable results.
In the first case, with no operator action, peak containment pressure would reach 63.1 psig.
The containment design pressure of 47 psig would be exceeded in about 60 minutes.
In the second case the operator was assumed to restore a second CFCU in 42 minutes.
By this analysis, containment pressure did not rise above design pressure.
In the third case the operator was assumed to restore a second CFCU in 55 minutes.
By this analysis, a containment nressure of 48.8 psig, slightly above design pressure, was reached 72 minutes after the LOCA.
Operator actions to restore a CFCU are specified in Emergency Procedures EP E-0 and EP FR-Z.l.
Under the conditions listed above, a " severe challenge (magenta path)" would be diagnosed with containment pressure between 22 and 47 psig.
This condi-
-
.
-
-
- - - -
..
--..- -
- - -.
-. - -. -.
.
f l
- 10 -
-
,
tion would cause the operators to check all other critical safety functions and then initiate functional restoration of
,
containment. Operator actions are directed to EP FR-Z.1, which
'
specifies that the CFCUs be verified to be running in low speed.
If this is not obtained, the operators are directed to
" manually start CFCUs in slow speed." While not specific as to
<
the actions necessary to accomplish this, the overload trip of a CFCU is indicated on the main control board and operators are knowledgeable of how to reset a tripped breaker.
It appears reasonable to expect that operators would be able to reset one of the CFCU breakers within 42 minutes.
CFCU breakers would be cap' ble of being reset within minutes after the breaker a
tripped.
(2)
Failure to Take Adeouate Corrective Action When reverse rotation of Unit 1 CFCU 1-5 was identified on
'
March 25, 1991, at end of the refueling outage (AR A0224682),
an incorrect evaluation concluded that the issue was not of i
This evaluation was stated in the action request (AR)
concern.
to have been done by maintenance engineering after consultation with electrical and mechanical maintenance personnel.
The
'
basis of the cc.nclusion was not stated, 1or was th're any formalization of the evaluation.
It did not appear that design engineering (NECS/0 PEG) personnel participated in the
'
evaluation.
The licensee's failure to take appropriate corrective action P
when Unit 1 CFCU was found to be rotating in the reverse direc-tion is an apparent violation of 10 CFR Part 50, Appendix B, Criterion XVI, which specifies that measures be established to assure that conditions adverse to quality are corrected.
The reverse rotation of the fan was indicative of a condition adverse to quality.
(Appar-t Violation 275/92-li-02)
QC is required by procedure NPAP C-12, Revision 20, Identifi-
!
cation and Resolution of Problems and Nonconformances, to review quality related ARs to determine whether a quality problem exists, and if so, to initiate a quality evaluation (QE) to document the root cause and corrective actions.
In
,
'
!
this case, based on discussions with the QC Manager and QC personnel, it appears that the QC reviewer accepted the evaluation and concluded as the evaluators did that no quality
problem existed. However, the AR did not provide a basis for
the conclusion, as directed by NPAP C-12.
It appears that the QC reviewer did not aggressively challenge the evaluation conclusions or require a more detailed explanation of the bases
of the evaluation.
'
(3)
Broad and Thorouch Reviews Were not Conducted Even After i
,
Problems Were Well Known
)
'
'
i
!
After identification of the loose counterweights in Unit 2 on January 22, 1992, it appeared that resolution efforts were J
$
_
__
-
.
._
_
_ _ _ _.
.
_-
_..
_
_.
_..
_ _
_
_
_
i
,
- 11 -
u
.
focused on the evaluation and resolution of individual problems.
A broad review of potential problems with the backdraft dampers was not completed until the end of April 1992.
The licensee required nearly a month, from January 23, 1992 (when the Unit I counterweights were found to be installed too tight) until February 19, 1992, to determine that the dampers for thrse CFCUs were stuck in the open position.
Analyses were iocused on the potential effects of counter-weights being too tight and not on whether the entire damper functioned properly.
It appears that miscomunications occurred regarding whether all or only some of the dampers in Unit I had been observed to be working properly. No outside organization was critically reviewing the conclusions or bases for the conclusions regard-ing operability of the CFCUs.
This could have assisted ir, identifying the miscommunications or in assuring that assump-tions were promptly confirmed.
When broken bolts were identified on March 23, 1991 in CFCU 1-2, an engineering evaluation was not done to determine the root cause of the bolt failures.
This indicated a need for a more inquisitive engineering approach to problem resolution.
Even after the Unit I backdraft dampers were found to have major problems, the discovery of previously undetected problems in the Unit 2 dampers on April 16 clearly showed that licensee personnel did not maintain an objective and inquisitive attitude. At least five documented inspections (Work Order C0096321, Numbers 1 through 4 dated February 5, 12, 20 and 27, 1992, respectively; and "As-Built Configuration of the Unit 2 CFCU Backdraft Dampers," dated March 3, 1992) documented that the Unit 2 backdraft dampers were properly installed and met the design drawings.
After the Apr.. :? discovery that the counterweights in damper 2-2 were too tight, the licensee found that damper 2-4 weights were also too tight and (contrary to the design drawings) that extra washers had been installed on dampers 2-2, 2-3, 2-4, and 2-5.
It was subsequently found that the tight counterweights found in Unit 2 on April 16 had previously been identified by mechanical maintenance personnel on January 24 (Work Order C0095999), but this was not communicated to the maintenance engineers doing the Unit 2 damper inspections. The Work Order documenting this finding had reportedly been left open (and the information not communicated) to allow completion of other work under the same Work Order number.
Reportedly, QC had not inspected dampers 2-2, 2-3, 2-4, and 2-5, since no work had
'
been done on them.
Technical Specification 6.8.1 requires the licensee to estab-lish and implement procedures recomended by Regulatory Guide 1.33, Revision 2.
Licensee Administrative Procedure C-40S3, i
,_
.. _, _
_
. _..
_
-
s.
,
- 12 -
j l
'
Use of PIMS Work Order Module, Revision 16, Paragraph 4.8.1, requires that work be performed in accordance with instructions provided in the work order.
Contrary to the above, on February 5, 12, 20, and 27, 1992, the licensee maintenance engineer (s)
failed to check the freedom of movement of the counterweights on CFCU backdraft dampers 2-2 and 2-4 as prescribed in Work Order C0096321.
(Apparent Violation 323/92-17-03)
License maintenance engineers reportedly focused on the known problems (loose counterweights) in Unit 2 and did not actually do some of the checks required in the work instructions.
Even after these issues were resolved, the NRC inspector identified a condition (some counter weights not resting on the bumper blocks) which should have been observed and resolved by licensee personnel.
The lack of broad focus and inability to identify the CFCU problems should be a significant concern to licensee management.
It also appears that the failure to promptly close Work Order C0095999 kept information from reaching the personnel who needed it.
This may indicate a weakness in the timely closure of Work Packages.
The inspector expressed concern that PG&E management expectations are either not clearly understood or not being implemented.
(4)
Lack..of Consistent Work Crews and Supervision It appears that part of the reason why the Unit 1 CFCU back-draft dampers were not installed properly was that this work was not given much supervisory attention and was considered as work to be done when personnel were available. Related maintenance procedures and work instructions also provided insufficient direction to maintenance and inspection personnel.
This resulted in numerous people and crews doing tne work, apparently without sufficient coordination.
This work was considered to be relatively simple and there was therefore minimal supervisory and no QC overview.
The inspector expressed concern that other work activities may sim * rly not have been given much supervisory attention, and therefore may not have been performed adequately.
(5)
Ventilation System Review A number of problems with the containment ventilation system have been experienced at Diablo Canyon.
These problems include significant duct cracking, dirt buildup in CFCU cooling coils, bolted connections loosening or breaking off, holes elongating, and counterweights falling off.
These problems appear to be indicative of longstanding problems with the system.
It is noted that the licensee plans to perform significant ducting work during the next Unit I refueling outage. An opportunity is apparent for the licensee to perform an overall review of the containment ventilation system to determine the root causes of the various problems and resolve them completel..
.
l
- 13 -
.
d.
Summary The licensee has initiated a task group to review the above issues related to the CFCUs. The task group is expected to provide a root cause assessment and recommend corrective actions to prevent recurrence.
The results of the task group will be reviewed during a future inspection.
3.
Exit Meetina An exit meeting was conducted on March 19, 1992 with the licensee repre-sentatives identified in Paragraph 1.
During this discussion, the inspectors summarized the scope and findings of the inspection as described in Paragraph 2 of this report.
CFCU issues were also discussed during management meetings conducted on April 2 and 21, 1992.
The licensee did not identify as proprietary any of the information reviewed by or discussed with the inspectors during the inspection.
,
>