IR 05000344/1991007
| ML20024H114 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 04/29/1991 |
| From: | Morrill P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML20024H111 | List: |
| References | |
| 50-344-91-07, 50-344-91-7, NUDOCS 9105210452 | |
| Download: ML20024H114 (27) | |
Text
.
...
-
_ ----
.. -.. - - -.
-. - _
-
~
.
<
.
.
,
e
.
V. S. NUCLEAR REGULATORY COMMISSION
REGION V
Report No.
50-344/91-07 Docket No.
50-344 License No.
NPF-1-Licensee:
Portland General-Electric Company 121 S.W. Salmon Street Portland, OR 97204 Facility Name: Troja Inspection at: Rainier, Oregon Inspection conducted:
February 10 - March 31, 1991 Inspectors:
R. C. Barr Senior F,esident-Inspector J. F.- Melfi-Resident Inspector Approved By:
$dW[
vec.
4 29-91 P. J. Morrill, Chief Date Signed Reactor Projects Section 1 Summary:
Inspection on February 10 --March 31, 1991 (Report 50-344/91-07)-
Areas Inspected:
Routine inspection of operational safety verification, maintenance, surveillance, event follow-up, system engineering, and'open item follow-up.
Inspection procedures 30703, 40500, 61726, 62700, 62703, 71707, 71710, 90712, 92700, 92701,-92702,_and 93702 were used as guidance during the conduct of the inspection.
Safety Issues Management System (SIMS) Items TMI action plan-item III.D.3.4.3.
(See paragraph 10)
- Results General Conclusions on Strengths and Weaknesses:
Strengths During_this inspection, the-1991 Refueling Outage Schedule appeared well thought out,-appropriately detailed, and effective in integrating sch_eduled maintenance activities-with plant conditions. -Additionally, the licensee
-
9105210452 910429 PDR ADOCK 05000344 O
- -,_.---. - -.,,. - _
. -
~
.
_g
..
'
.
y
.
,
,
.
appeared to have taken special steps to minimize facility operation in mid-loop-(Paragraph 13).
Weaknesses During this inspection period, the following weaknesses were identified and brought to management's attention:
o Weak Maintenance Work Practices:
Paragraph 4 describes multiple instances of weak work practices by both maintenance craftsmen and maintenance supervisors..A maintenance craftsman was observed performing work without reading or referring to work instructions.
Craft
supervisors assigned to direct maintenance activities were observed not
'
.i to be attentive to their responsibilities.
o Weak Refueling Outage and Forced Outage Planning:
Paragraph 4 describes-multiple instances of forced outage planning that resulted in poor quality work instructions and extended delays.
Paragraph 13 describes weak work planning with respect to preparation for the 1991 Refueling Outage, o
Weak Management-Involvement--in Facility Activity:
Paragraph 6 describes multiple instances where, had facility management been more involved with activities and their potential impact on undesired facility operations, plant events may have been avoided.
Paragraph 4 and 13 indicate a weakness-in. management's-involvement with outage planning.
o'
Weak Craftsman Training Program:
Paragraph 6 identified two areas where improved tra1r.ing for the craftsmen could have improved the response to equipment fai3ures.
Significant Safety Matters Four violations were identified.
Paragraph 4 described a violation associated.
with failure to promptly document. deficiencies.--Paragraph 6 described 1 violations associated with failure-to comply with procedures and failure to provide adequate instructions and drawings.
Paragraph 8 described a procedural-violation associated with instrument calibration.
Open Items Summary
--In addition to the four violations, five LERs (Paragraph 12) were closed,
- three unresolved items were closed (Paragraph 7), and one enforcement item was closed (Paragraph.9).
. -. -..
.
..,
.-
-
- - -
.
-
-..
- _. - - -.. -.
.
-
_
_ _ _ _ _ _ _ _ _ _ _
.
.
-
.
l DETAILS 1.
Persons Contacted a.
Portland General Electric
- J. E. Cross, Vice President, Nuclear
- W. R. Robinson, Plant General Manager
- G. D. Hicks, General Manager, Plant Support C. K. Seaman, General Manager, Nuclear Quality Assurance
- T. D. Walt, General Manager, Technical Functions C. P. Yundt, General Manager, Trojan Excellence A. R. Ankrum, Manager, Nuclear Security M. W. Hoffman, Manager, Nuclear Safety ar.d Regulation M. B. Lackey, Manager, Planning and Control J. W. Lentsch, Manager, Personnel Protection W. O. Nicholson, Manager, Operations
- W. F. Peabody, Manager, Nuclear Plant Engineering
- S. M. Quennoz, Manager, Technical Services R. L. Russell, Outage Manager M. J. Singh, Manager, Plant Modifications
- J. F. Whelan, Manager, Maintenance
- S. A. Bauer, Branch Manager, Nuclear Regulation G. P. Enterline, Branch Manager, Operations J.
Mody, Branch Manager, Plant Systems Engineering D. L. Nordstrom, Branch Manager, Quality Operations J. D. Reid, Branch Manager, Quality Support Services
,
- A. D. Rice, Branch Manager, Chemistry G. L. Rich, Branch Manager, Radiation Protection
- J. J. Taylor, Branch Manager, PM/EA J. A. Benjamin, Supervisor, Quality Audits
- L. W. Erickson, Supervisor, TN0B
- W. J. Williams, Compliance Engineer b.
Oregon Department of Energy A. Bless, Resident Engineer
,
The inspectors also interviewed and talked with other licensee employees
$
during the course of the inspection. These included shif t supervisors,
~
reactor.and auxiliary operators, maintenance personnel, plant technicians and engineers, and quality assurance personnel.
- Denotes those attending the exit interview.
2.
Plant Status At the beginning of the inspection period the facility was in Mode 1 at 100% power.
At 10:20 am on February 12. 1991, as the result of an electrician incorrectly installing an electrical jumper, the reactor was automatically shutdown (paragraph 6).
At 6:28 am on February 15, 1991, after repairing containment isolation valve CV-1452 (paragraph 4), the reactor was restarted. While at approximately 45% power at 12:39 pm, 12:52 pm, and 2:22 pm, the facility experienced three unexplained load
.
--
-
_. _
__ _ _ _ _ _ _, _ _ _ _ _ _ _ _ _ _ _ _ _, _ _ _ _ _ _ _ _ _
__
__
_
_
__
__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - __-___-___ _ _ _ _ _ _
.
~
.
.
n
.
,
.
oscillations of between 440 MW to 50 Md.
At 10:00 pm, plant operators O
recommenced power ascension after evaluating the performance of the main turbine control system (EHC).
Full power was achieved at 4:35 am,-
February 17, 1991.
At 5:50 pm on March 3,1991, plant operators began a reactor shutdown, as required by Technical Specification (TS) 3.8.2.1, due to the failure of the Y-28 inverter.- At 3:10 pm on March 5, 1991, the facility reached Mode 5, cold shutdevn.
On March 25, 1991, while in Mode 5 and filling the-reactor coolaat sy vem, plant operators exceedea the technical specification cooldewn limit for the pressurizer (Paragraph 6).
On March 26, 1991, with the reactor coolant system solid, the over pressure protection system tctuated when reactor operators started a actor coolant pump (Paragraph 6; On March 27, the 1991 Refueling outage began. At the conclusion of the inspection period the facility was in Mode 5 with operators conductmg reactor refueling prerequisites.
3.
Operational Safety Verification (71707)
,
During this inspection period, the inspectors observed and examined activities to verify the operational safety of the licensee's facility.
The observations and examinations of those activities were conducted on a daily, weekly or biweekly basis.
Daily the inspectors observed control room activities to verify the licensee's adherence to-limiting conditions for operation as prescribed in-the facility Technical Specifications.
Logs, instrumentation,
. recorder traces, and other operational records were examined to obtain information on plant conditions, trends, and compliance with regulations.
On occasions when a shift turnover was in progress, the turnover of information on plant status was observed to determine that-pertinent information was relayed to the oncoming shift personnel.
Each week the inspectors toured the accessible areas of the faciitty to observe the following items:
(a)' General plant and equipment conditions.
(b)- Maintenance requests and repairs..
(c) Fire hazards and fire fighting equipment.
(d) Ignition sources and flammable material control.
(e) Conduct of activities in accordance with the licensee's administrative controls and approved procedures.
(f) Interiors of electrical and control panels.
(g) -Implementation of the licensee's physical security plan.
(h) Radiation protection controls.
(i) Plant housekeeping and cleanliness.
-(j) Radioactive-waste sys tems.
(k) Proper storage of compressed gNs bottles.
Weekly, the inspectors examined the licensee's equipment clearance control with respect to removal of equipment from service to determine that the' licensee complied with technical specification limiting
conditions for operation.
Active clearances were spot-checked to ensure
!
'
that their issuance was consistent with plant status and maintenance evolutions. - Logs of jtnipers, bypasses, caution and test tags were examined by the inspectors.
.
-.---ve
-
r*-t--r
'--m
'-
ee
-,m
"*v e-
+<'-em"vv-m
--
-w-*r-"
- *-
--*=-=-e
= - - - - - - - - - + --
-
" - - - - - - - - - - - - - - - - - - - - - -
-
-
-.
-.
-
_
_ - _ _ _ _ _ _ _
-
,
.
'-
.
.
Each week the inspectors conversed with operators in the control room, and with other plant personnel.
The discussions centered on pertinent topics relating to general plant conoitions, procedures, security, training and other topics related to in progress work activities.
The inspectors examined the licensee's Corrective Action Program (CAP) to confirm that deficiencies were identified and tracked by the system.
Identified nonconformances were being tracked and followed to the completion of corrective action.
Routine inspections of the licensee's physical security program were performed in the areas of access control, organization and staffing, and detection and assessment systems.
The inspectors obser"ed the access control measures used at the entrance to the protected brea, verified the integrity of portions of the protected area barrier and vital area barriers, and observed in several instances the implementation of compensatory measures upon breach of vital area barriers.
Portions of the isolation zone were verified to be free of obstructions.
Functioning of central and secondary alarm stations (including the use of CCTV monitors) was observed.
On a sampling basis, the inspectors verified that the required minimum number of armed guards and individuals authorized to direct security activities were on site.
The inspectors conducted routine inspections of selected activities of the licensee's radiological protection program.
A sampling of radiation work permits (RWP) was reviewed for completeness and adequacy of information.
During the course of inspection activities and periodic tours of plant areas, the inspectors verified proper use of personnel monitoring equipment, observed individuals leaving the radiation controlled area and signing out on appropriate RWP's, and observed the posting of radiation areas and contaminated areas.
Posted radiation levels at locations within the fuel and auxiliary buildings were verified using both NRC and licensee portable survey meters.
The involvement of health physics supervisors and engineers and their awareness of significant plsnt activities was assessed through conversations and review of RWP sign-in records.
The inspectors verified the operability of selected engineered safety features.
This was done by direct visual verification of the correct position of valves, availability of power, cooling water supply, system integrity and general condition of equipment, as applicable.
No violations or deviations were identified.
4.
Maintenance (62703)
Auxiliary Feedwater Steam Supply Valve (CV-1G2)
The Auxiliary Feedwater (AFW) Pump steam driven turbine is supplied with steam from each of the four steam generators via normally closed supply valves CV-1451 through 1454.
These valves also serve as containment isolation valves.
The valves are identical three inch diameter, air operated, wedged disk gate valves.
Technical Specifications (TS)
applicable to these valves are 3.7.1.2, AFW System; and 3.6.3.1,
._
[
-
,
.
.
.
.
,.
Containment Isolation Valves.
The inspector observed maintenance on CV-1452 from February 12-15, and on March 5, 1991, and reviewed the maintenance history associated with this valve.
Background-Nonconformance Report (NCR)89-572 documented traceability concerns over
'
the body-to-bonnet seal ring installed in CV-1452.
On January 3, 1990, to close out NCR 89-572, the licensee generated maintenance request (MR)
90-3654 to replace the seal ring.
MR 90-3654 was planned to work during the 1990 Refueling Outage; however, due to inadequate planning and work load considerations, the work was deferred.
Also during the 1990 Refueling Outage, the air actuator fr CV-1452 failed Periodic
,
Engineering Test (PET) 16-3, " Slow Loss Oi 5 trument Air," which tests CV-1452's air actuation system's integrity.
Per MR 90-6833, the actuator was repaired and satistactorily retested.
On November 29, 1990, during the quarterly performance of Periodt Operating Test (POT) 5-1, " Auxiliary Feedwater System Pump and Valve In-Service Test," CV-1452 failed its stroke time test.
Crattsmen conducting the test concluded that the wedge disc may Dave beoh sticking to the seat.
The craftsmen generated MR 90-12729 to investigate the valve's apparent wedge disc-to-seat sticking.
On December 12, 1990, after lubricating the valve steam and subsequent cyclings of the v61ve, MR-12729 was completed and CV-1452 passed timing requirements ' even though the wedged disc still appeared to be sticking slightly to the seat.
Corrective Action Request (CAR) C90-1076 was written to doctnent CV-1452's repeated failures to meet its required stroke time.
Additionally, because CV-1452 was a valve covered by the Technical Specifications, repair of the valve per MR 90-3654 was added to the next forced outage work list.
On February 12, 1991, the facility entered a forced outage.
MR 90-3654 was worked in order to comply with the requirements of TS 3.0.4.
At the February 13, 1991 plant status meeting, personnel from the Planning and Scheduling Department noted that CAR C90-1076 corrective actions also needed to be completed.
MR 91-1121 was written to address these
,
corrective actions.
Obse vation The inspector obsseved maintenance activities on CV-1452, which included MR 91-1121 (repack valve and perform corrective actions per CAR 90-1076),
MR 90-3654 (valve repair and seal ring replacement), and MR 91-1123 (check blowby-observed on solenoid valve during MR 91-1121).
On February 13, 1991, the craf tsmen started working on CV-1452 per MR 91-1121.
While cycling the valve, CV-1452 first stuck, and then popped open.
On the second attempt at cycling the valve, the air operator continuously vented resulting in the valve's failure to operate.
The air system was isolated and MR 91-1123 was written to test the air solenoid for leakage.
j During the actuator's removal, the valve's stem and disc unexpectedly l
!
_,
, _ --
- -
- - _ _ _ _ _ _ _ _ _ _ _ _
.
j
.
,
.
,
opened.
Because the c?earance boundaries had been inadequately researched (MS-143 that had previously been noted in NCR 87-31/ as naving seat leakage was used as a boundary isolation valve), work on the valve had to be dise:ntinued until toe reactor coolant system (RCS) was cooled down to 455 degrees F. and clearan:e boundaries were reestablished.
The inspector requested to be informed when work on the valve was resumed; however, due te miscommunication between maintenance supervisors, he was not.
CV-1452's actuator was removed.
With the actuator removed from the valve, the craftsmbn evaluated the air leakage from the air cylinder per MR 90-3654.
The craf tsman noted minor air leakage around the main piston seal ring; however, because he believed the leakage was trinor the leakage was not quantified.
These seals serve as the air isolation boundary between the high and low pressure sides of the velve.
Leakage by the cup seals could impact the ability to operate the vulve.
The leakage across the cup seals was not quantified.
Investigation for leakage on the solenoid identified no leakage.
Other valve maintenance included disassembly, and seat and disk inspection to ensure proper seat to disc contact.
The valve stem was replaced.
The valve disk and seat were polished, and the valve reassembled.
During the reassembly, the inspector noted that two seal rings of different sizes were issued with the same storage code numbe".
The craf tsman initially installed one of the seal rings.
Because the seal ring appeared not to seat properly, the MR work instructions were revised to increase the bonnet assembly bolt torque value.
After completely reassembling the valve, the clearances between the valve body and bonnet were still excessive.
As a result, the valve was again disassembled.
The craftsman then attempted to install the old seal ring with a setting material (copaltite).
The craftsman then observed that the seal rings (old versus new) were different sizes.
At that time, the craftsman recognized that the incorrect seal ring had been installed, and the other seal ring was the correct size.
The Paintenance Branch Manager directed the Preventive Maintenance Manager to document, via a CAR, that two seal rings of different size had been supplied using the same parts storage code number.
A week later when the inspector had identified that a CAR had not been written, the Preventive Maintenance Branch Man 1ger told the inspector that he did not realize that he was supposed to write the CAR.
As a result of the inspector's questioning, during the NRC weekly exit of February 22, 1991, the licensee initiated a CAR.
1his is an apparent violation of Nuclear Division Procedure (NDP) 600-0, " Corrective Action Prograin," that states nonconformances are to be initiated within one business day (50-344/91-07-01).
After replacing the seal ring, the valve was reassembled by a different craftsman.
The reassembly was observed by a Quality Inspections (Ql)
inspector and two maintenance supervisors.
During the reassembly, the inspector noted that the craftsman failed to read or follow the MR instructions for torquing the valve body-to-bonnet bcits.
The inspector observed that the craftsman had the torque wrench set to 16 ft-lbs vice the 8 f t-lbs required by the work instructions and that the craf tsman was
---._----.s-,_..___,___.m
_ _,,
,
'
l
'
-
.
,
starting to torque the bolts in a sequence different than that specified by the procedure.
The NRC resident inspector, who had a copy of the work instructions, then requested the craf tsman to reverify the torcuing requirements.
The valve body-to-bonnet bolts were then torquec properly.
Because the supervisors observing the maintenance had the work package, the craftsman continued to reassemble the valve without referring to the work instructions.
H4 inspector noted that the supervisors were not effectively monitoring or directing the maintenance as it progressed.
CV-1452's actuator was then reconnected to the valve.
During the reconnection, the craftsman replaced the rod bushing 0-rings.
These 0-rings are part of the air cylinder pressure boundary.
After the valve and actuator were reassembled, the air lines and the rod bushing assembly were checked for leakage.
While checking for leakage, tht! craftsman found and removed a solenoid vent tube that potentially restricted the porting of air.
Any restriction of the ported air was contrary to vendur installation requirenents.
Licensee research identified that the tubing had been installed per a maintenance request and without proper reviews.
Plant operators conducted post maintenance testing (PMT) on the valve per POT 5-1.
The test required measuring the the time to open CV-1451 through CV-1454 using accumulator supplied air, then isolating the accumulator and closing the valves using instrument air.
However, the test was performed using the accumulator to cycle the valves.
In deviating from the procedure, CV-1453 failed the test.
CV-1452 cycled satisfactorily.
The licensee then reperformed POT 5-1 using only the instrument air system.
CV-1451 through CV-1454 cycled within time requirements.
The licensee decided not to perform PET 16-3, a surveillance that verifies the integrity of the valve's pneumatic actuating system, because the air cylinder for CV-1451 had not been worked on.
At 4:58 pm on February 15, 1991, CV-1452 was declared operable.
On February 21, 1991, at 12:52 pm with the facility in Mode 1, an attempt was made to cycle CV-1452 using the safety-relatd accumulator per POT 5-1.
The valve failed to cycle, which indicates that the PMT for the valve maintenance may have been inadequate.
The decision not to perform PET 16-3 may have resulted in CV-1452 being returned to service without being operable.
On March 5, 1991, during a second attempted valve operation after declaring CV-1452 operable, PET 16-3 was performed and failed.
After returning to Mode 5 on March 5, 1991, licensee craftsmen replaced CV-1452's air cylinder, per MR 90-6833 A design change that machined a groove into the air piston and installed a third seal was performed.
Licensee craftsmen noted that a light grease (Dow Corning 111) was used in the air cylinder.
When maintenance supervision contacted the vendor, the vendor noted that Dow 111 was not appropriate for use on CV-1452's actuator. Of the four similar air cylinders in the system, the licensee verified that CV-1452 was the only air cylinder where that grease was used as a lubricant.
The vendor recommended using a light mineral oil.
The lubricant was changed, seals installed, the valve reassembled and retested satisfactorily.
_ _ _ _ _ _ _ _ _ - _ _ - _ -
-.
- - - - - -
_ _.. _
'
.
..
.
,
'
'
One violation was identified.
5.
Surveillance (6172,6]
As part of inspector followup on accurate tank level indication,
>
identified in inspection report 50-344/91-03, the inspector looked at the licensee's chemistry analysis of the sodium hydroxide concentration
'
(Specific Gravity 1.30) in the spray additive tank which atlects the calibrationandthustheaccuracyofthetanklevelindication.
The inspector observed the licensee perform a determination of the concentration of sodium hydroxide in the Spray Additive Tank, as required
'
byTechnicalSpecification(TS)4.6.2.2.b.
The licensee determines sodium hydroxide concentration from specific gravity and temperature, and by use of a table relating NaOH concentration to specific gravity and
. temperature. The table was provided from a chemical company information pamphlet.
The inspector questioned the appropriateness and accuracy of such an approach and was told by the Chemistry Branch Manager that this approach was standard in the industry.
The inspector checked various other plants and discovered that other plants with a spray additive tank perform a titration to determine concentration.
The initial chemistry program (specified by the vendor)
also specified a titration, as do subsequent industry standards (i.e.
The titration method can be used since NaOH is a strong base, and accurate results can be obtained by titrating with a known concentration of a strong acid.
The licensee also contacted other plants and verified that the standard method was to perform a titration for this measurement.
After observing the licensee perform a concentration measurement, the inspector had the following concern's with the licensee's procedural guidance, accuracy, and data collection.
'
The directions given in the procedure were to simultaneously measure temperature and specific gravity, locate the appropriate temperature column on the graph, and find where the specific gravity is equal to the-measured parameters.
The percent concentration can be then be estimated.
,
The technician placed the sodium hydroxide solution in a clear cylinder and then raised the temperature to 80 degrees F in a temperature bath.
The technician made.sure the hydrometer was not touching the sides of the cylinder to assure an accurate measurement.
The raising of the temperature and assuring that the hydrometer was not touching the sides were not in the procedure. -The chemistry technician did measure the specific gravity and temperature, but did not rer.ord these values.
No calculation was performed to extrapolate the measured values to the values on the graph, but a value was estimated by the technician.
The licensee recorded the concentration as 31 weight percent NaOH.
If the tables were correct, the inspector performed an extrapolation and determined the value as 30.7%.. CMP-13_also referred to a backu)
gravimetric procedure to be used if a hydrometer was not availasle.
The licensee does not have a gravimetric procedure.
-
The inspector questioned the licensee about the accuracy of this approach.
The licensee had no information about how accurate their
.
.
..
.
...
.
-
__
...
-.
. i
-
.__ - - _ _ _ _ _ _ _.
._
.
.
e
-
approach was, except that the quality checks performed with blind samples were always accurate.
The inspector questioned if the vendor supplied table was accurate.
The licensee checked the values on the table with accepted references, which assured the accuracy of the table.
The inspector asked if the licensee had any basis for deviating from the Westinghouse recommendation to conduct a titration to determine the NaOH concentration.
The licensee was not aware of any basis documents.
The licensee determines the results and puts them on log sheets which are separate from the procedure.
The acceptance criteria are specified on the log sheets.
This data sheet is considered a Quality Assurance (QA)
record to document the results of the surveillance test and is performed every six months per Technical Specification 4.6.2.2.b.
This is the only place where the results are recorded.
Since the licensee had blanks for over 20 results, the licensee could have had over 10 years of data on one sheet.
The inspector's concern was that QA records are normally stored in the records vault to provide assurance of retention.
The licensee stated that the data sheets would be changed to resolve this problem.
The inspector had concerns about the documented review of the results of theinistry analyses.
10 CFR 50, Appendix B, criteria XI states that test results shall be documented and evaluated, and the licensee's chemistry program require periodic reviews of results.
The results are documented as being reviewed by initials on the data sheets.
The inspector reviewed the results for the Reactor Coolant System (RCS) analyses and various boron analyses.
The RCS analyses for boron, chloride, fluoride and oxygen are required to be done three times every seven days, and the boron concentration in the accumulators is required to be analyzed once every seven days (Modes 1 thorough 4).
The documented, initialled verification for these technical specification recuired analyses was not performed for periods up to five RCS analyses, anc three accumulator analyses.
The inspector review of these results indicated that none of the results exceeded the technical specifications, but there was no documented evidence that they were reviewed, nor did there appear to be a requirement regarding how soon the results should be reviewed by the Chemistry Foreman.
With respect to the accuracy of the boron method compared to industry standards, the licensee does their analysis in accordance with industry standards, except with respect to the effect of lithium on the boron concentration.
Industry standards account for the effect of lithium on the boren concentration.
The licensee stated that the effect of lithium is small, (2-3 ppm) and is ignored.
In summary the inspector concluded that:
The licensee meets the minimum requirements for a procedure in some of their chemical analyses.
The licensee deviated from industry standards for chemical analyses and did not always have the basis for the deviation.
The licensee's cecord keeping of the chemical analyses could be improved since there is no documentation that some analyses had been
_
_ _ _ _ _ - _ _ _ - _ _ _ - _ _ - _
_
- _ _ _ - _ _ _ _ _ _
,
o reviewed, and the amount of data on one sheet of paper could be several years worth of data.
No violations or deviations were identified.
6.
Event Follow-up (62703, 92701, 93702)
a.
Automatic Reactor Shutdown _ Maintenance Technician Error At 10:20 am, February 12, 1991, while at 100% power, the reactor automatically shutdown due to a main turbine trip signal that then caused a reactor trip via the Reactor Protection System (RPS).
During the reactor shutdown, all equipment functioned as expected with the exception of the steam dumps (failed to control in auto while in the steam pressure Mode), the A Steam Generator (5/G) power operated relief valve (PORV) (failed to fully rescat after opening)
and the main feedpump (MFP) suction relief valves (failed to rescat).
5)ecifically, at 10:12 am on February 12, 1991, maintenance request (iR) 90-12401 was released to repcir damaged (dented) electrical conduit for a portion of the facility's access lighting.
The work instructions of the MR required the installation of a jumper; however, the HR's work instructions did not identity the jumper termination points. When the craftsman inttalled the jumper in panel Q-27, he failed to fully understand the configuration of the components in the panel and installed the jumper across the incorrect points.
This resulted in impressing a 120 VAC signal on the 125 VDC system, and the generation of " electrical noise in the electro-hydraulic control (EHC) system.
The licensee concluded that the " electric noise" activated three relays (high exhaust hood temperature, low hydraulic pressure and shaft pump discharge low pressure), which then closed electric contacts that completed the EHC system's turbine trip logic, following the reactor trip, the licensee, to determine the root cause of the event, conducted a post trip review and convened an event review team (ERT).
The licensee, using event and causal factor analysis, identified one root cause and eight contributing causes.
The root cause of the event was "the lack of proper panel layout and component level drawings [that] prevented the ur.it supervisor from identifying specific terminals and prevented the worker from correctly identifying the intended terminal" for maintenance that was being conducted to repair damaged electrical conduit.
The more significant contributing causes included:
failure of the work process to accurately assess the risk; failure of the work process to prevent the changing of work instructions without an independent review; failure of the planning process to require a comprehensive walkdown prior to developing work instructions; and failure of the peer verification process to prevent termination errors.
To meet 10 CFR 50.73 reporting requirements, the licensee submitted Licensee Event Report (LER)
50-344/91-04.
.__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ -
_
.-
-
..
.
l
.
.
The licensee performed an engineering evaluation to assure that components were not damaged by incorrect installation of the jumper, lhe licensee concluded no A.C. or D.C. components were damaged since only minimal voltage, and current and power increases of short duration (about 5 millisec), were experienced.
The licensee also conducted a visual verification (walkdown) of all potentially affected equipment and no unusual conditions were noted.
the licensee evaluated the equipment history of the Additionally,lhey found that in 1983 the EHC vendor recommended to EHC system, PGE management, via a technical bul ktin and subsequent telephone conversation, a modification to replace existing relay boards with relay boards that were not sensitive to electrical noise.
PGE management chose not to perform the modification, but rather to install, contrary to the vendors recommendation, diodes to suppress electrical noise.
PGE plans to reevaluate the need to modify the EHC system to better suppress electrical noise.
Additionally, the performance of the EHC system was monitored during the February 15 reactor restart.
During the restart, the facility, when between 45%-55% power, experienced several turbine load oscillations of up to 400 Ws that were believed to be due to EHC system malfunctions.
Because the load oscillations were intermittent, the licensee was unable to determine their cause.
PGE plans to perform detailed diagnostic troubleshooting of the EHC system during the 1991 Refueling Outage.
With respect to the equipment malfunctions that occurred during the reactor shutdown, two of these, the SG-PORV and MFP suction relief valve malfunctions, were repeat malfunctions that appear not to have been previously effectively addressed.
The licensee took the following actions to address equipment malfunctions experienced during the February 12 reactor trip.
The main feedpump suction relief valves, that in 1976 were determined to have a relief setpoint that was too low (due to increasing discharge pressure of the condensate pumps), but were not replaced due to an inadequate design change (RDC 76-139), were gagged closed and will be replaced during the 1991 Refueling Outage.
The SG-PORV reset setpoint was found to have drifted and was readjusted.
The manual-auto station for the steam dump controller was found to have a marginal negative
'
15 volt power supply and the controller was replaced.
The resident inspectors observed the reactor shutdown and cooldown following the reactor trip, attended the ERT meetings, reviewed corrective action request (CAR) 91-0032, discussed the event with selected personnel involved in the event and reviewed the event's root cause analysis.
The inspectors concluded that the licensee's evaluation failed to fully develop the causes of this event.
For example, the analysis did not identify that the job analyst f ailed to comply with procedures in the planning of MR 90-12401.
In discussions with personnel involved with this event, the inspector learned that the job analyst had not walked the job down per step 4.6.d.1 of Administrative Procedure (AO) 3-9, " Maintenance Requests," Revision 35.
By this procedure, the job analyst was required to " assess the
_.
_-_-- -.
__ -_-_
'
.
-
.
.
reported condition [ job task) and develop work instructions with sufficient detail to enable qualified individuals to perform the specified actions without direct supervision.
IF during the performance of the job assessment activities, peFmanent plant equipment / components are found NOT to be properly identified, THEN initiate immediate action to ensilie the installation of identification BEFORE the requested work / maintenance is ptrformed."
This procedural noncompliance led to the initial issuance of inadequate work instructions.
This is an apparent violation (50-344/91-07-02).
Another example of the event analysis's failure to fully develop the causes of the event was that the work control center's review of MR 90-12401 failed to establish work instructions to ensure that acceptable plant conditions were established.
For this maintenance, the initial work instructions could have resulted in potentially draining down of the plant batteries.
The on-duty shift's review of MR 90-12401 questioned the acceptability of potentially draining the station batteries and shift management requested the MR be rewritten. The original MR work instructions were discarded and could not be found during the event reconstruction efforts.
In the development of the new work instructions, the need to install a jumper was identified.
Because Trojan procedures did not require termination points for jumpers, the revised work instructions did not identify the specific termination points.
Because the licensee's event analysis did not identify this discrepancy, and in view of the fact that several similar errors had recently occurred, the inspector considered this an apparent violation of 10 CFR 50, Appendix B, Criterion V (50-344/91-07-03).
In review of CAR 91-0032, the residents' inspection found that the licensee identified twelve corrective actions to prevent event recurrence.
All corrective actions with the exception of 91-32-04 appeared appropriate.
91-32-04 changed A0 3-9 to require specific identification of jumper termination points for all future work packages; however, the work packages for the upcoming 1991 Refueling Outage were not reexamined to require termination points for jumpers or lifted leads.
Licensee management, during discussions with the residents on this corrective action, agreed to review the 1991 outage work packages to ensure the packages that installed jumpers would identify their termination points.
Additionally, the committed completion dates for a number of the open corrective actions appeared untimely.
For example, by September 30, 1991, NPE electrical branch would evaluate improvements for labeling and drawings on contactor enclosures even though work was planned during the 1991 Refueling Outage in those enclosures.
Also by August 31, 1991, Work Planning would evaluate training needs for work planners prior to being permitted to planning work packages even thcugh planners were planning packages until August 31,1991.
The Plant Manager committed to reevaluate the timeliness of these corrective actions.
The Quality Assurance review of the corrective actions associated with CAR 91-0032 appeared weak with respect to assuring the actions were timely and thorough.
The inspector, in discussions with a
_ _ _ _ __ _ _ _ _ - _ _ _ - _ _
.-
.
.
. _ -. -.
- -.
_.
_
'
.
.
,
'..
,
'
Quality Assurance supervisor, learned that the CAR had not been properly routed through the QA organization and, therefore, had not received the appropriate reviews prior to accepting the corrective actions, b.
Manual Reactor Shutdown Due to Inverter failure At 7:20 am on March 3 1991, licensed operators noted that the number 4 Preferred AC Instrument Bus had shif ted to its alternate power source-station battery via inverter.
Technical Specification , 3.8.2.1 required emergency busses to be OPERABLE, which includes all
'
four of the 120 Volt AC Preferred Instrument busses.
The Action required that "with less than the required complement of AC busses j
OPERABLE, restore the inoperable bus to OPERABLE status within 8 i
hours or be in COLD SHUTDOWN within the next 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />." The licensee initiated an urgent MR to investigate the cause of the inverter shifting power sources.
Initially, licensee craftsmen found no reason for the power source shift.
At 2:51 pm, plant personnel attempted to restore the inverter's preferred power source.
After the inverter was energized, craftsmen discovered an overheated 1500 ohm resistor.
The licensee searched the spare parts supply and found no replacement resistors in stock.
At 4:50 pm, plant operators initiated a reactor shutdown.
PGE also declared an Unusual Event, as required by facility Emergency Procedures.
Further parts research identified an equivalent resistor with a 2 watts vice 1 watt), in an inverter slightly higher power rating (ing building.
that was located in the train The licensee cannibalized the resistor, initiated temporary modification 91-004, and installed the resistor.
In the process of installing the resistor, the craftsmen visually checked the four major circuit cards of the inverter.
They discovered the unit sensing board had a failed 22 ohm resistor and a failed diode.
(The unit sensing board is the card that causes the inverter to transfer to the alternate powersource).
That circuit board was not carried in the spare parts inventory.
The facility achieved hot shutdown (Mode 3) at 4:55 am on March 4,1991.
Repair of the damaged circuit board continued.
The facility entered Mode 4 at 9:25 pm.
At 3:10 am on March 5, the plant entered Mode 5 and exited the Unusual Event.
In addition, PGE contracted the vendor to assist with inverter trcubleshooting and repair.
The vendor determined that the cause of the inverter failure was that components on the static switch board
,
had failed due to the Y-28 inverter having a 12 degree phase shift between what the inverter alternating current output and the inverter alternating current supply.
The nominal value for phase shift should be not greater than 6 to 10 degrees.
PGE engineers concluded that over time, the excessive angle phase may have detrimentally affected the components on the static switch card.
Craftsmen replaced the defective components and tuned the inverter to match the required phase angle.
The inspectors observed selected aspects of craftsmen and vendor troubleshooting and repair of the inverter, attended the ERT for the event and discussed the shutdown with plant management.
The Vice
&
_
t
.
<
.
..
!
,
'
President Nuclear stated the spare parts inventory for the inverters would be reviewed and augmented to minimize the probability of forced outages caused by inverter failures.
c.
Pressurizer Cooldown Limit Exceeded
.
'
At approximately 5:30 pm on March 25, 1991, with the facility shutdown (Mode 5) and the-reactor coolant system (RCS)-temperature
'
at 129 degrees F., plant operators began manually borating (20 gpm)
the RCS to acnieve refueling boron concentration and to increase pressurizer level in preparation.for transitioning to solid plant operations. -Other plant conditions important to this event include pressurizer level at 25%, pressurizer vapor and water temperature at 435 degrees F., pressurizer heaters deenergized, the C Reactor Coolant Pump (RCP) running and the C RCS loop providing pressurizer spray.
At approximately 6:10 pm, because plant evolutions associated with PET 15-1, "CCW Heat Exchanger and CW Supply Test,"
,
were to complete by 6:30 pm, the shift supervisor directed plant operators to discontinue manual boration and begin emergency
'
borating (100 g)m) to more rapidly increase RCS boron concentration.
At 6:40 pm, wit 1 pressurizer level at_43%, plant operators recognized that pressurizer water temperature was at 140 degrees F.,
while the pressurizer vapor. space temperature was still at 435 degrees F.
The operators stopped emergency borating, energized pressurizer heaters and obtained data to assess the pressurizer cooldown.
Plant operators determined that Technical Specification (T.S.) 3.4.9.2.b., which reqJires that "the pressurizer temperature shall be limited to a maximum cooldown of 200 degrees F. in.any one hour," had been exceeded with a maximum cooldown of approximately
.275 degrees F. in one hour.
Pressurizer water tem)erature was returned to within Technical Specification limits ay 7:30 pm.
lechnical Specifications require the licensee to perform an analysis to determine the effects of the out-of-limit condition on the fracture toughness properties of the pressurizer.
The licensee will complete the evaluation prior to the scheduled June 7, 1991 reactor-restart and has placed a Mode 4 hold on the ready-for-restart list.
Licensee post event followup consisted of contacting the vendor, reviewing plant history for similar events, and assessing operator actions during the event.
In discussions with the vendor, the licensee learned that over-the past two years, twelve other nuclear facilities of Westinghouse design experienced over-cooling of the pressurizer.
Westinghouse believed pressurizer over-cooling occurs-due to reduced pressurizer spray flow when using the C reactor coolant loop spray and charging to the RCS at relatively high rates.
Westinghouse noted that the concern of over-cooling the pressurizer was thermal-shock to the pressurizer's lower hemispherical head and support skirt.
Westinghouse will be conducting an-analysis for PGE
=to determine the effects of the over-cooling event on the fracture toughness properties of the pressurizer.
In checking historical data,_PGE found that on February 15, 1983, pressurizer cooldown limits were exceeded during the performance of in-service testing that verified proper ECCS valve operation and flows.
PGE changed procedures to test the valves under different plant conditions.
.
.
.
.
. -
__ -_ _ _. _. _.. _ _ _ _ _ _ _ _. _ - - _ _ _. _ _
~
-
.
14
,
,
-
.
,
.
With respect to evaluating the March 25, 1991 event, the licensee determined that the significant contributing factors for the event were being in an abnormal lineup (B RCP off), using C RCS -loop spray (which appears less effective than B loop pressurizer spray) and an inadequate pre-briefing for plant evolutions. The_ licensee, at the l
conclusion of the inspection period, was conducting a root cause evaluation.
The resident inspectors attended the licensee's event review team (ERT) meeting, evaluated pre and post event plant conditions, discussed the event with the on-shift operating crew, reviewed plant procedures and reviewed the 1983 pressurizer over-cooling event.
Theinspectorsconcludedthatthemajorcontributorstothisevent were complacency by Operations management to fully research the effects of an apparent design deficiency and lack of involvement by
_
Operations management in on going plant activities.
Normally, the RCS is taken solid using the B RCP and B RCS loop pressurizer spray.
Because the B RCP had a suspected failed lower motor bearing, it was not available for use.
0)erations management failed to evaluate the impact of not having the LRCP for cooldown.
Furthermore, plant procedures state that when going solid or filling the pressurizer, RCP pump combinations that include the B RCP should be used.
The event was exacerbated because the licensed operators did not consider themselves to be bounded by Operating Instruction (01'3-3),
-
" Drawing or Collapsing a Pressurizer Bubble," Revision 23.
01 3-3 describes.the operation of the Pressurizer System during drawing or
<
collapsing of a pressurizer bubble, the first major step of which is
'
to fill the pressuriter.
The procedure, which the operators reviewed prior to beginning to fill the pressurizer, provided precautions to prevent over-cooling, and instructions to prepare and to take the RCS solid.
It appeared to the inspectors that bad shift
_ management examined 01 3-3 and researched the importance of the B RCP, the event could have been prevented.
The licensee committed to submit a licensee event report (LER). -NRC followup of this event
- will.be performed through review of the LER.
Two violations were identified.
7.
Followup on Unresolved Items (92701)
Unresolved Item 50-344/89-10-04, (Closed), " Containment Spray Header Structural Support Elements." This unresolved item concerned missing-structural support elements underneath the Containment Air Coolers (CACs).
The missing supports also affected the containment spray header structural integrity.- -The licensee identified the item during a walkdown that-was conducted due to previous concerns raised over the adequacy of pipe supports.
The inspector verified that PGE installed additional supports _for the CACs and containment spray header.
The inspector also reviewed a Bechtel calculation that evaluate.1 a seismic II/I concern.
Bechtel concluded that, even without the modification, the support system would have withstood a seismic event
.
..
.
.
-
.
-.
_ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - _ - - _ _ _ _.___ -______-_____--. _ - _ _ _ _ _ _ _ - _ - - _ _ _ _ _ _ _ _ _ _
_ _ _ _
'
j
.
.
.
,
..
,
'
since the loads would be redistributed to other supports.
Based on the calculation and addition of structural support this item is closed.
Unresolved Item 50-344/90-06-03, (0 pen), " Evaluation of the Steam Generator Pressure Decrease on Steam Generator Flow Instrumentation."
During a review of Engineered Safety features (E5F) setpoints on steam generator flow, the inspector identified a concern regarding the effects that low steam generator pressure could have on steam generator flew indication.
The original 100% power design pressure for the steam generators was 895 psig.
The current 100% flow indication is with a
'i steam generator pressure of approximately 850 psig.
Since the configurationatTrojanhasaverylowdPacrosstheflowelement, steam can be considered as a compressible fluid.
Because the flow indication relies on a differential pressere (dP) across the flow element, the inspector requested the licensee to evaluate the effect on flow for the pressure decrease.
PGE, with assistanct from Westinghouse, is evaluating the inspector's concern.
The results of the evaluation should be
'
completed within approximately six weeks.
This item remains open pending further licensee evaluation.
,
Unresolved item 50-344/90-06-06, (Closed), "Undervoltage Relay Time Response Testing." This unresolved item concerned the method of testing the licensee used to test the time response of ESF bus undervultage relays.
The relays are supposed to be calibrated each refueling cycle.
.The degraded voltage relay trip setpoint is 3580 + 80 volts for 55 + 5 seconds.
This 55 second delay consists of a four second relay (whi~dh also verifies the voltage) and a 51 second agastat relay.
If the degraded voltage is in-for more that four seconds, the agastat then actuates.
PGE calibrated the agastat relay every three years, which is beyond the_ refueling cycle calibration.
Subsequently, PGE determined that a technical specification compliance issue did not exist because the four second relay was calibrated each refueling.
The licensee used the definition of a instrument channel in
.IEEE 279 which states, "a. channel-loses its identity where single action signals are combined." Based on the electrical drawings, the four second relay comprises the entire instrument channel.
The 51 second relay is in the logic part of the circuit, therefore, the technical specification does not apply to the agastat relay.
To resolve this item, the licensee plans to revise the technical specification.
The licensee revised the procedures to calibrate the agastat relay on a refueling basis.
This item is closed.
Unresolved item 50-344/91-11-01, (Closed), " Component Cooling Water (CCW)
to Residual Heat Removal (RHR) Heat Exchanger Flow.".This unresolved-
-item identified proc 6 dural weaknesses that could allow RHR/CCW flow to be-less than the required:5000 gpm. -The licensee revised-these two procedures to more explicitly require the proper flow to the CCW/RHR heat exchangers.
Based on the procedure revisions, this item is closed.
..
.
._
_
.
_,__
_
_ _.
. _ - _ _ _ _ _ _ _ _ - _ - _ _ _
. _ _ _ _.
.
-
.
-
.
,
.
Open Item 50-344/91 03-06, (Closed) " Qualification of Electrical Containment Penetration 5plices."
Questionable splice configurations (16 splices) were tested at Wylie Laboratories.
This activity was completed in March, 1991.
During the week of March 4, 1991, Wylie verbally reported that the spilces passed the testing.
Howeser, when the Wylie test chamber was opened on March 8, 1991, it was noted that the splice sleeves exhibited linear indications in the vicinity of the bolting pad.
The responsible PGE engineer, when informed of this condition, requested that Wylie return all tested splices to Trojan for inspection.
-
On March 9,1991, the tested splices arrived at the Trojan site.
PGE inspection identified that two-thirds or more of the splices displayed
"
splitting on the exterior of the Raychem heat shrink tubing surrounding
,
the fiberglass cover and bolting pad, extending through the heat shrink tubing thickness.
failedconfIguratIon.Emadethedecisiontoreplaceallspliceswiththe On March 10 1991 PG The replacement splice was one for which all applicable Environmental Qualification documentation was clearly on file.
The replacement configuration / method involves removing the fiberglass sleeve covering the bolting pad and reinstalling new Raychem heatshrink tubing in accordance with the configuration originally qualified by Raychem.
,,
The licensee planned to have Wylie test 24 splice specimens of the failed configuration to establish when the discrepant splice would have failed.
The licensee stated that the reportability of this situation would be evaluated.
This item is cloud.
No violations or deviations were identified.
8..
Engineered Safety Features (ESF) System Walkdoyn (71710)
The Emergency Diesel Generators (EDGs) are the safety-related source of
'
,,
power to the 4.16 kilovolt (kV) ESF Buses following loss of preferred (offsite) power. As the source of safety-related power, the diesels can provide electric power to the ESF equipment following reactor trips or safety-related subsystems including Diesel Fuel Oil (DF0)y a number of safetyinjections.
The diesel generators are supported b
'
, Diesel Generator Air Start, and Diesel Generator Lube Oil systems.
Prior to walking down this system and subsystems, the inspector collected
-
information on selected EDG instrumentation.to verify that it was within its current calibration cycle.- During the review on-this instrumentation, the--inspector noted that Level-Transmitter (LT) 4900A was almost at 1.25 times its normal calibration interval (2 years).
This transmitter is used to verify that the DF0 storage tank contains greater r
than 33,000 gallons of fuel oil, which is a technical specification surveillance requirement (4.8.1.1.2.a.2).
The inspector brought this to the attention of the Instrumentation and Control (I&C) Department (2/21/91) and was told by the 1&C supervisor and manager that the
-
veH e
,%m+,e---m.a w
w.
5--
...--n..-w=--.
.-em
--m m--w,
-.srw-
-
s-r.---.smm
..:,e e--wc-E -. e%.
r*
---
v mw.m-
- m.w,e
mm-wei m e ~ ~ -
_ _ _ - _ _ - - - _ _ _
_ _ _
'
.
.
.
licensee's program allowed the Technical Specification (TS) priority 2 instruments to go up to one calibration interval beyond its normal due date.
The inspector questioned what a TS priority 2 instrument is and received several different answers from different groups.
The surveillance group showed definitions to the inspector stating that the 1.25 criteria applied to the TS priority 2 instruments.
The licensee's calibration program is defined in Maintenance Procedure (MP) 2-0, " Installed Plant Instrumentation." Step 6.1.1.c of MP 2-0 states in part, that "The calibration due date for equipment covered by or meeting technical specification requirements (safety related) may be extended by a factor of 1.25 times the calibration interval."
Surveillance requirement 4.8.1.1.2.a.2 states that it will be verified that the DF0 tank has greater than 33,000 gallons at least every 31 days (on a staggered test basis).
LT 4900A is defined as a TS priority 2 instrument and the indicator fed by this transmitter is called out in Periodic 0)erating Test (POT) 12-1, " Monthly Idle-Start and Loading of Emergency )iesel Generators " to verify if the transmitter is reading greater than 81% (33,000 gallons).
As of March 31, 1991, the transmitter had not been calibrated.
Since this transmitter is used to meet the technical specification requirement, the licensee's failure to calibrate the transmitter within 1.25 times its normal calibration frequency is an apparent violation (50-344/91-07-04) of the licensee's program.
The inspector was informed that the licensee deferred this surveillance interval tn reduce the amount of time technical specification equipment is out of service.
The reasoning used for not calibrating this instrument was that the calibration of LT4900 involves the physical removal of the transmitter off the tank, which trips the low level switch off the level transmitter prohibiting the DF0 pump from starting.
Since the calibration of the transmitter takes more than the four hours allowed by the technical specification, this renders the system inoperable.
The inspector reviewed other priority 2 instruments in the licensee's calibration. program.
The inspector noted that the A train EDG day tank level is verified by the level indicator (LI-4904A) on panel C-17 by POT 12-1.
This indicator is used, in part, to meet technical specification 4.8.1.1.2.a.1.
The level transmitter that feeds this indicator also went beyond its 1.25 frequency allowance on March 24, 1991.
The inspector walked down the accessible portions of the diesel and of the diesel support systems.
In the walkdown, the inspector noted that most of the instrumentation was within its calibration cycle (except as noted above).
The diesels were determined to be operable.
The areas were generally clean, and no significant leaks in the piping systems were noted.
The inspector did find some trash in the following locations:
in the local start panels: fuses, instrument light bulbs, a paper clip, and a metal strip; under the base plates of the diesels:
cigarette butts and an old donut box.
The items under the diesels were heavily oated with dust and appeared to have been there a long time.
The inspector informed the licensee about these items and they were
.
'
'
.
.
.
.
,
.
I removed.
The inspector noted a piece of wood wedged in on top of the power and current transformer connection box on the West diesel, and the spaces where the power cables went out of the transformer boxes (West i
EDG) and into both (east and west) of the local start panels were not fully sealed.
Water from the sprinkler system in the EDG rooms could get in the local start panels if the sprinkler system actudted.
I
,
The licensee was finishing an electrical Design Basis Document (Db0)
walkdown of the diesels and had noted some of the items.
The DBD walkdown did not look in the local control cabinets as part of the DBD effort and the licensee did not identify that the power cables were not fully sealed.
The inspector related his concerne to the licensee about the seals where the power cables are.
The licensee restored the seals (MR 91-2407 and 91-2408) and initiated a Corrective Action Request (CAR) 091-616.
The inspector will follow the resolution to the CAR.
One violation was identified.
9.
Followup of Notices of Violations (92702)
Enforcement Item 50-344/90-29-01, (Closed), "IST Gauge Accuracy Recuirements." This enforcement action documented a violation of the A5FE code that established range and accuracy requirements of the temperature instrument (pyrometer) used to measure pump bearing temperature on the Component Cooling Water (CCW) pump.
Part IWP 4120 of the ASME code states that, "the full-scale range of each instrument shall
!
be three times the reference value or less.' The historical value for the reference value for this temperature is approximately between 50 and 70 degrees F.
The pyrometer used had a 0 to 500 degree F. scale in 10
degree increments.
The reference temperature was approximately 70 degrees F.
The full-scale instrument range to use in this case would have been between 150 and 210 degrees F.
The inspector also verified i
that there was no program to train plant operators on gauge accura:y requirements.
As corrective actions, the licensee put a dedicated temperature instrument with an acceptable range in the control room and plans to reperform the test during the next quarterly surveillance.
Based on planned test performance with an acceptable lastrument, this item is closed.
i No violations or deviations were identified.
10.
TI 2515/65, Three Mile Island (TMI) Action Plan Eclowup (92701)
TMI Item III.D.3.4.3, (Closed), " Control Room Ha3itability Requirements -
l Implementation."
This TMI action plan item was to assure that control room operators will
'
be adequately protected against the effects of an accidental release of I
toxic or radioactive gases, and that the control building heating,
,
l ventilation and air conditioning (HVAC) system was OPERABLE.
I i
-
. _.
.. -
- -
-.
- - -.
. -. -.
.
..
- - -. - - -
-
--
- - - -
--
.= _ - -.
-. -
'
'
.
.
- ..
,
.
Based on an escalated enforcement action, a previous NRC inspection l
50-344/86-29 left open several issues regarding this TMI item.
The control room HVAC was found at that time to have several deficiencies
!
which made it susceptible to certain common Mode failures, further, the
'
TMI action plan required several inspection activities to close out a TM1
action item.
These items can be summarized as follows:
-
a.
Verify the installation and operability of the CR HVAC b.
Verify that design makeup flow to the control room has t'een increased to 500 Cubic feet per Minute (CFM)
i c.
Verify the licensee turns off the control building HVAC systems (CB-5,6, & 10) in the event of a CB-2 trip d.
Verify that addition of information identified by the NRC has been included in a supplement to LER 86-02 e.
Assess the status of the ammonia and sulfur dioxide technical specifications
-
f.
-keview the licensee's design analysis on the ventilation system, r
subsequent to the escalated enforcement action, including technical i
specification and FSAR changes The inspector walked down the system with the system engineers to
'
-determine its status.
The system was considered OPERABLE.
The inspector i'
found one bolt on the support for CB1 that did not have adequate thread engagement.
The system engineer determined that the operability of the
system was not impacted by the finding.
-
,
The inspector verified that the design makeup flow had been increased to 500 CFM and that CB 5 6 and 10 automatically deenergized if normal controlroomventilation,waslost.
The additional information requested in Inspection Report 50-344/86-29 was included in the supplement to LER
86-02.
The ammonia detector's status was resolved by an NRC letter of August 7, 1990, from R. Bevan-to J. Cross, that stated that NRC staff
-
review determined that the ammonia detectors were no longer required.
The Sulfur Dioxide detectors were operational, but the licensee has
,
submitted a License Change Amendment to-delete these detectors.
This issue h currently under NRC staff review. -Based on the walkdown and design repairs to the control room HVAC, this item is closed.
,
No violations or deviations were identified.
11.
Evaluation of Licensee Self Assessment Capability (40500)
This inspection evaluated the-effectiveness of the licensee's self assessment programs --the Trojan Nuclear Operations Board (TNOB) and the Plant Review Board (PRB). - The evaluation focused on the ability of the self-assessment program to contribute to-the prevention of problems by s
monitoring and evaluating plant performance.
The inspectors attended i
selected committee meetings, observed TNOB and PRB_ meetings, reviewed committee members' qualifications, reviewed audits conducted under the cognizance of the.TH0B, selectively reviewed the followup of committee action items and evaluated the adequacy of the committees' open item tracking systems.
_
.
.
... _ _. _ _ _ _ _.. _ _ _ _ _, _ _. - _ _ _
_ _ _, -
_ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.
'
-
..
...
.
a.
Tro.jan Nuclear Operations Board Trojan Technical Specification 6.5.2 describes the function, composition, review and auditing requirements of the TN08.
PGE implementsTS6.5.2withNuclearDivisionProcedure(NDP)500-1,
"Tro1an Nuclear Operations Board." The TNOB, a nine member committee consisting of five Nuclear Division General Managers, one contractor, one non-Nuclear Division member, and two Nuclear Division Managers, appeared to have limited effectiveness.
The TN00 generally recognized weaknesses in facility programs but did not effectively develop and communicate their concerns to management.
Specific examples included ineffective followup of the floor drain preventive masntenance program, ineffective followup of fire protection program deficiencies and weaknesses, and ineffective followup of outage preparation and execution weaknesses.
The Board's ineffective followup generally resulted from not verifying the actions or the effectiveness of the actions that the line organization committed to perform.
TH s, in part, was exacerbated by the Board's limited use of the Quality' Assurance organization to probe suspected weak areas and the Board s limited access to plant vitalareastoindependentlyassessperformance.
The effectiveness of TNOB communications with the Vice President, Nuclear appeared limited.
The TNOB communicated with the Vice-President, Nuclear via the TNOB meeting minutes.
The TNOB meeting minutes (approximstely fifty pages in length for each meeting) documented the discussions that occurred during the TN0B meetings.
The minutes, produced by the TNOB staff, and reviewed and approved by the TNOB, did not effectively summarize or present the Board's concerns to the Vice President.
The concerns were often in the form of TNOB open items tnat were interspersed throughout the meeting minutes.
The TNOB appeared to meet the requirements of the Technical Specifications, but was not fully effective in focusing on assessing facility performance.
As an example, Technical Saecification 6.5.2.7 establishes the review requirements for tie TN0B.
The TNOB had no formal document that described how specifications 6.5.2.7.e, 6.5.2.7.f and 6.5.2.7.h were met.
The TN0B staff supervisor believed the intent of the specifications were being met.
For example he noted the intent of TS 6,5.2.7.f that required review ofsignlficantoperatingabnormalitiesordeviationsfromnormaland expected performance of plant equipment that offect safety, was being met by the TN0Bs review of LERs, NCARs, NCRs and CARS.
However, the inspector found that the TNOB actually reviewed only severity level I and 11 CARS; and the QA subcommittee, not the TNOB, was provided with a listing of the severity level !!! CARS; and that the TNOB staff recommended only a selected subset of the Severity Level !!I CARS for review.
The inspector also noted that TNOB open item 204-08, which had been open since January 18, 1990, had raised concerns over the TNOB's compliance with TS 6.5.2.7.h, which requires in part review review "all recognized indications of an unanticipated deficiency in some aspect of design or operation of safety related structures, systems or components," and had been
._
,
. -.
.
.
--
..
...-
-
. -
___ - _ ____ _ _ _ -
'
'
..
-
,.
,
'
l continuously deferred.
This, to the inspectors, appeared to be an inordinate time to have a concern over compliance with technical specifications.
To address the inspectors' concerns over compliance with the above TS6, the TNOB chairman committed to formalize the j
method in which the TNOB meets the TS review requirements.
In discussions with the VP Nuclear and the newly appointed TNOB Chairman, the ins 3ector learned that in the near future the constituency of tio TNOB would be changed to include more outside representation and fewer members of line management.
Additionally, the Chairman plans to review and revise the TN0B charter.
b.
Plant Review Board (PRB)
TrojanTechnicalSpecification6.5.1describesthefunction, composition, review and auditing requirements of the PRB.
'
-
implements T.S. 6.5.1 with Administrative Order (AO) 2-1, " Plant Review Board Charter."- The PRB, a five member committee consisting of four supervisors of the Trojan Plant Organization, a supervisor of the Quality Assurance organization, and alternates appeared to begenerallyeffectiveincontributingtothepreventIonofplant problems.
With respect to PRB performance, the inspectors identified two weaknesses.
The first weakness, the PRB's review of plant safety issues, as required by T.S. 6.5.1.6.g, was not always thorough.
The safety significance, root cause and corrective actions for events were not always well discussed or understood.
For example, the PkB during discussions on the failure of the Y-28 inverter, did not discuss tethnical specification requirements, design basis or root cause of the failure. During discussion of the impact of low temperature on NAMC0 limit switches, the PRB did not fully investigate the impact of cold weather on other limit switches.
The PRB could be more effective if their reviews of events were more intrusiveness and thorough.
When plant events were evaluated, the applicable T.S., design basis, impact to safety, and previous
- similar events were not always discussed.
During the PRB meetings the inspectors attended, only the event, issues leading up to the event and the actions taken to correct the problem were generally discussed.
The second weakness was that the PRB appeared not to review all potential safety events prior to reactor restart from a forced outage.
P1 t Administrative Orders (AO) A0 3-7, " Post Reactor Trip /Shutdowti or Safety Injection Reviews," and A0 3-25,
" Ready for Startup," requires that the PRB review all potential safety events that resulted frem and occurred during a reactor trip or forced shutdown. The PRB did not have a method for determining which potential safety events would be reviewed prior to restart.
For example, for the intended restart from the March 4,1991 reactor shutdown, the PRB would not have reviewed the failure of auxiliary feedwater flow switch FIS 3004D2, had it not been questioned by the resident inspector.
The PRB Chairman, in discussions with the inspector, ackncwledged these weaknesses.
He committed to establish a format for event
.
,....r.._
,,-,
.-...,,,<s.
,..
-
,,,...
,
, _.. -
- -.
. _, -, ~,.
,, _
-.,,, -.
--
-,
-
<
.
.
.
t
.
,
review that would include technical specification requirements, design requirements, design basis, past history and root cause.
He also committed to develop a criteria and format to ensure the PRB reviewed all outage safety issues.
No violations or deviations were identified.
12.
Follow-up of Licensee Event Reports (92700)
LER 90-22, Revision 0, (Closed), " Degraded Fire Penetration Seals As a Result of Inadequate Installation Practices." This LER described the licensee's discovery, while conducting fire seal inspections per the criteria recommended in NRC Information Notice (IN) 88-56, of 44 degraded fire seals for through wall electrical and pipe penetrations.
The licensee concluded the cause of the deficient fire seals was lack of proper inspection criteria wf installation technique during original construction.
As correct" lons, the licensee posted compensatory watches, inspected simi'
rations, developed a comprehensive program for the insper I fire seal penetrations and repaired the fire seals found inac
, licensee concluded the safety significance of this hat 11 of the 44 degraded fire penetration seals cou, ected the safe shutdown of the reactor had there both a fire in iccted areas.
The inspectors verified the following licensee corrective actions:
posting of compenntory etches for degraded seals; installation, using revised practices as recommended by HRC IN 88-56, of a repaired fire seal; and review of the February 28, 1991, schedule for restoration of other degraded fire seals.
The licensee attico appeared appropriate; however, in some respects not timely.
This LER is c!osed based 4.n licensee completed and proposed corrective actions.
LER 90-27, Revision 0 and Revision 1, (Closed), " Containment Isolation Valves Missed Surveillance Due to Inadequate Procedures." This LER described the licensee's nonperformance of Technical Specification (TS)
Surveillance 4.6.1.1, " Primary Containment-Contairment Integrity," for a manual containnent isolation valve for the steam generator blowdown
,
system (SG-159).
The licensee concluded the cause of the missed surveillance was inadequate implementation of a programmatic changt in how surveillance requirements were met.
As corrective actions, the licensee added SG-159 to their surveillance procedure reviewed drawings toensureallcontainmentisolationvalvesweresurvellledandplansto conduct in-containment walkdowns during the 1991 Refueling Outage to verify all drawings for containment isolation valves reflect actual plant configuration.
The licensee concluded this event had minimal safety significance since SG-159 was normally closed and found closed when the deficiency was identified.
The resident inspectors verified that the licensee revised the surveillance procedure (POT 3-3) to include the valves identified in the LER.
The inspectors also verified the licensee reviewed facility drawings to determine if other valves had been properly identified and surveilled.
During the inspection, the residents noted that the licensee had not performed a formal root cause for this event It appeared to the
_,
_. _
_
-_
__
+
.
.
.
,
s t
inspectors, the root cause of this event was an inadequate design change process.
The required configuration for containment isolation of closed systems has not changed since the construction of Trojan and the design
,
change process should have identified SG 159 as a containment isolation i
valve.
In reviewing the design chan e package the ins ectors found that j
SG-159wasnotidentifiedasacontanmentisolationvave.
The licensee agreed with the inspectors' findings.
The licensee play to establish a i
design review group to improve the quality of their design changes.
This
'
LER is closed based on completed and proposed licensee corrective actions.
LER 90-36, Revision 0, (Closed), " Inappropriate Testing of Containment Isolation valves." This LER described in service testing of six containment isolation valves when plant conditions as described in TechnicalSpecification4.7.1.2.1,werenotestab1Ished.
The licensee concluded that the cause of this event was their inconsistent interpretation of Technical Specifications.
Licensee corrective actions included revising in-service testing procedures and submitting Technical Specification chan The licensee concluded the testing of these valves, with inap)ges.ropriate plant conditions established, had no adverse impact on safety )ecause the fully redundant opposite train for these systems was always OPERABLE.
The inspectors verified licensee procedures were revised to comply with Trojan Technical Specification requirements.
This LER is closed based on licensee corrective actions.
LER 90-37 Revision 0, (Closed), " Lightning Strike Causes Auto-Start of
GeWe~nTyDits41 Generators." This LER descrified the actuation of an engineered safety feature (FSF), the starting of both emergency diesel
~
generators, due to a lightning strike causing a momentary reduced voltage on the facility startup transformers.
The licensee evaluated the grid voltage and startup transformer voltage oscillation that occurred during the transient and concluded all plant systems functioned as designed.
No licensee corrective actions were necessary.
The licensee concluded this event had no adverse safety it. tact.
,
The resident inspector evaluated the grid voltage oscillation and the licensed operators' response to the event.
The plant safety systems functioned as designed, This LER is closed.
LER 90-39, Revision 0, (Closed), " Inoperable Control Room Eme_rgency Ventilation System Due to Excessive Centrifugal Charging Pump (CUT Seal Leakage." This LER described, as the result of CCP mechanical seal leakage exceeding a leakage rate of 1772 cc/hr, the inoperability of the control room emergency ventilation system.
The licensee concluded the CCP seal failed due to boric acid buildup on components of the CCP mechanical seal.
As corrective action, licensed operators isolated the CCP thereby stopping the seal leakage.
Additionally, because the CCP mechanical seals appear to fail more frequently than desired, PGE is evaluating an alternate design.
The licensee concluded this event had no adverse impact on safety due to the short duration of the leak and due to alternate means available for licensed operators to reduce their exposure in the event of a design basis accident.
-.
-.
- - - - -. - -
--
-
-
--
-.-
- _ _ - _ _ - _ - - _ _ - _ _ _
'
..
o
.._
,
The inspectors observed selected portions of the maintenance ar,sociated with the CCP seal replacement.
The inspectors concurred with the licensee that there were no immediately visible indicators for the seal's failure.
Based on the licensee's evaluation of alternate seal designs, this LER is closed.
No violations or deviations were identified.
13.
1991 Refueling Outage Preparations (62700, 71707)
Trojan, due to being located in a region that is predominately supplied from hydroelectric power, conducts annual refueling outages.
The 1991 Refueling Outage is scheduled to be conducted from March 27, 1991 through June 7, 1991.
Majoractivitiesplannedforthe1991RefuelingOutage include reactor refueling, steam generator tube ins)ection, low pressure turbine inspection and containment air cooler refuraishment.
The Trojan Outage Program is described in Nuclear Division Procedure (NDP)800-4,"TrojanOutageProgram." This procedure, nich was approved 1990, established outage by the Vice President Nuclear on September 12,iet, and preoutage program milestoneguide1Ines.ageprogramresponsibilit program policies out Preoutage milestones are target dates for activities that are to be time phased prior to the outage that, if accomplished, should ensure a well-planned and executed outage.
Administrative Order (AO) 3-30, " Outage Preparation ana Execution,"
establishes for the Plant Organization, the process for outage scope, development, preplanning and execution of outages.
This inspection focused on PGE's preparation for the 1991 Outage as compared to the preoutage milestone guidelines established in NDP 800-4
'
and A0 3-30.
In addition, where departure from the guidelines occurred, the inspectors assessed the potential impact on the remainder of the planning process.
In conducting the inspection, the inspectors attended outage planning meetings, reviewed outage schedules, reviewed selected outage statusing documents and discussed outage preparations with members of the Trojan Planning and Control organization.
The inspectors identified two major improvements over the 1990 Refueling Outage preparations.
First, the Planning and Control organization developed a detailed schedule that included networking all proposed maintenance activities.
The schedule appeared to properly schedule maintenance activities for appropriate plant conditions.
The inspectors particularly recognized PGE's efforts to minimize plant operation at recognized that approximately 90% of the materials (y, the inspectors reactor coolant system mid-loop conditions.
Secondl parts) for the Outage-had been ordered and received.
This represents a substantial improvement over the 1990 Refueling Outage where many maintenance tasks were cancelled due to non-availability of parts.
The inspectors noted several recurring weaknesses from the 1990 Refueling Outage.
First, the implementation of the planning process appeared not guidelines to support the schedule.
NDP 800-4 established the following/ Design for work package completion:
Request for Design Change (ROC)
Change Packages (DCP) issued six months prior to the outage, Outage
.
- -.
.
.
.
-- _
.-
-
-
.
-
--
. - _ _ _ _ _ _ _ _ _ - _ _.
2b
.o
,
.
.
,
.
Maintenance Requests for all Preventive Maintenance Requests (PMR)
written seven months prior to the Outage, Outage Maintenance Requests for all Corrective Maintenance identified seven months prior to the outage, l
and all maintenance work packages written three months prior to the outage.
In nearly every respect, the inspectors identified significant deviations from these preoutage milestones.
For example, of the approximately 3200 maintenance requests scheduled to work during the outage, only approximately 2050 were fully 31anned and ready to work on March 27, 1991 - the beginning of the 1991 Refueling Outage.
Of the 36 scheduled DCPs, PCCs or PMRs, only seven were ready to work on March 27, 1991.
Approximately four days into the Outage, as a result of unavailability of some work packages, maintenance activities were being rescheduled.
To date, the inspectors had not identified where rescheduling activities resulted in safety issues or outage delays.
Second, it appeared that all levels of PGE management failed to focus on
indicators that pointed to large departures from the s tablished preoutage milestones. With respect to planning for PMRs, DCPs, and RDCs, the Manager of Plant Modifications did not start to hire personnel to plan maintenance requests (MRs) until mid-January 1991, even though the MRs were scheduled to be completed by December 27, 1990.
With respect to the planning of MRs for outage preventive and corrective maintenance, PGE recognized in November 1990 that meeting preoutage milestones would be dif ficult.
To address this potential problem, in December 1990, PGE planners began working in excess of 60 hour6.944444e-4 days <br />0.0167 hours <br />9.920635e-5 weeks <br />2.283e-5 months <br /> work weeks.
As noted earlier, even with this action, only approximately 65% of the work packages were available at the start M the Outage.
On January 17, 1991, at a PGE-NRC management meeting, the 1991 Refueling Outage was discussed.
Atthatmeetinglewithrespecttotheplanningofworkpackages,NPC Regional man overly optimist particularly design packages.
The inspectors could identify no additional actions PGE took to address potential Mk planning shortfalls.
Finally, it appeared that due to the lack of adequate resources planning, PGE may have to work planners in excess of 60 hours6.944444e-4 days <br />0.0167 hours <br />9.920635e-5 weeks <br />2.283e-5 months <br /> per week for six months or more.
The inspectors were concerned that this may result in presture for the planners to product packages too quickly.
The inspectors noted that the February 12, 1991, reactor trip was largely due to a poorly planned package.
PGE should closely monitor MR quality throughout the Outage.
The inspectors concluded that although improvements in this area had clearly been achieved, the full implementation of the licensee's new procedures was not yet effective and warrants additional management atteqtion.
No violations or deviations were identified.
14.
Exit Interview (30703)
The inspectors met with the licensee representatives denoted in paragraph 1 on March 29, 1991, and with licensee management throughout the inspection period.
In these meetings the inspectors summarized the scope and findings of the inspection activities.
.
.
.