IR 05000295/1982010
| ML20054K504 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 06/14/1982 |
| From: | Connauhgton K, Jackiw I, Ring M NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML20054K500 | List: |
| References | |
| 50-295-82-10, 50-304-82-08, 50-304-82-8, NUDOCS 8207020262 | |
| Download: ML20054K504 (9) | |
Text
f
.
.
U.S. NUCLEAR REGULATORY COMMISSION
REGION III
Report No..50-295/82-10;~50-304/82-08 Docket No. 50-295; 50-304 License No. DPR-39; DPR-48 Licensee: Commonwealth Edison. Company P. O. Box 767 Chicago, IL 60690 Facility Name: Zion Nuclear Power Station, Units 1 and 2 Inspection At: Zion Site, Zion, IL Inspection Conducted: March 4, 5, 8-10, April 7, 14, 15, May 17 and 27, 1982 Inspectors:
0 /WF2, gg"-
K. A. Cdanaughton ( -///- 6 l (March 8-10 and May 27, 1982)
A V'
[' /t/-f L Approved By: J.
. J ckiw Chief Test P/ogram,Section p
Inspection Summary Inspection on March 4, 5, 8-10, April 7, 14, 15, May 17 and 27, 1982 (Report No. 50-205/82-10; 50-304/82-08)
Areas Inspected:
Special inspection to review circumstances surrounding foreign material in IB and ID steam generators during eddy current inspec-tion on Unit 1 and RTD bypass loop low flow occurring during Unit 2 cycle 6.
Review consisted of interviews with operating, maintenance and technical staff personnel, examinations'of maintenance procedures, outage logs, Westinghouse drawings and evaluations, a Babcock and Wilcox report and licensee investigation reports. The inspection included 77 inspector-hours onsite by two NRC inspectors.
Results: No items of noncompliance were identified during the inspection.
8207020262 820616 PDR ADOCK 05000295
r
.
.
DETAILS 1.
Persons Contacted
- K.
Graesser, Station Superintendent
- T. Lukens, QC Supervisor
- P. LeBlond, Assistant Technical Staff Supervisor
- P. Hull, Quality Assurance
- Denotes those attending the exit interview of May 27, 1982.
The inspector also contacted other members of the licensee's admin-istrative and technical staff during the course of the inspection.
2.
Synopsis of Events Relating to Hinge Pieces and Hardware Found in Zion Unit 1 Band D Steam Generators On February 25, 1981, during the cycle 6 refueling outage, the Zion Unit 1B and D steam generator primary side manway covers were removed initially for purpose of eddy current examination of the tubes. Upon entry into the hot leg channel head areas, a stainless steel hinge approximately 30 inches long by 2 inches wide was found in the ID steam generator and 3 pieces (2", 2", and 24") of a similar hinge were found in the 1B steam generator. Additionally, five bolt-nut-washer combina-tions were found in the ID steam generator channel head or hot-leg region and damage to the tube ends was noted. One bolt-nut-washer combination was found in the ID cold leg. The licensee began an in-vestigation which resulted in the discovery that one of eight aluminum nozzle covers was missing. These covers are normally used to block the nozzle area in the steam generators to prevent loose equipment from falling into the reactor coolant hot or cold leg piping during eddy current testing. Since the hinges found in the steam generators were similar to those on the remaining nozzle covers, the licensee concluded that one aluminum nozzle cover was left in the hot leg channel head of the ID steam generator during the outage between cycle 5 and 6 in the spring of 1981. The nozzle cover consisted of three sections of 1/4 inch thick aluminum hinged together with two 30" long
stainless steel hinges and 36 nut-washer-bolt combinations. The licensee determined that the aluminum plates dissolved in an aluminum and water reaction and left the hinges and nuts, bolts and washers behind.
I Much of this information was initially documented in PNO-III-82-25
dated February 25, 1981 and updated on March 1 and 4, 1982. Via a
'
Task Interface Agreement with NRR dated March 10, 1981, Region III was requested to perform the following:
a.
Monitor steam generator inspection and repair activities.
- '
Monitor licensee's investigation of the presence of missing b.
material (hinge fragments and screws) in the primary coolant j
system.
I
!
k
r
.
.
c.
Review adequacy of licensee's QA/QC programs and management controls.
The results of the inspection of item a. (steam generator inspection and repair) and c. (adequacy of QA/QC and management controls) are primarily documented in Inspection _ Report No. 50-295/82-08. This report documents the results of item b.
(investigation of missing material) and additional subjects related to the cause and effects of the nozzle cover left in the 1D steam generator.
No items of noncompliance were identified.
3.
Missing Parts
Upon opening the manways to the 1B and ID steam generators, the licensee located one intact hinge approximately 30 inches long and pieces of the other hinge which when fitted together indicated approximately six small pieces of hinge were still missing. Of the original 36 nut / bolt / washer combinations which held the hinges to
-
the aluminum plates, one combination was found on the cold leg side of the 1D steam generator indicating the combinations could pass through the steam generator tubes. The licensee subsequently re-covered or located eleven other combinations in the hot leg channel head, the hot leg piping, or stuck in the steam generator tubes.
Tubes with bolt / nut / washer combinations stuck in them were later
i plugged.
During fuel moves, two more combinations and one bolt minus the nut and washer were located stuck in the bottom of fuel assemblies which were transferred to the spent fuel pit. The bolts had been peened over to prevent the nut and washer from coming loose
.
and the fuel assembly bolt was the only one found without the nut and washer attached. No portions of the aluminum plates were located and the licensee concluded it all dissolved. Further discussion of aluminum is contained in paragraph 6.
In total, the licensee believes there to be 21 bolts, 22 nuts and washers and six hinge pieces to
,
'
be missing.
It is not known exactly how the hinge pieces migrated from the D j
steam generator to the B steam generator, however, the most probable explanation is that the pieces were swept from the D steam generator to the B steam generator by reverse flow when D loop reactor coolant pump was shutdown with B still running at the end of cycle 6.
The physical configuration of the loops is such that the B hot leg is adjacent to the D hot leg.
Reactor coolant pump starting and stopping logs confirm that the reverse flow situation existed.
Efforts to locate the missing pieces included; inspecting the channel
,
heads of all four steam generators during eddy current testing, lower-
!
ing a sled mounted camera down the ID hot leg toward the reactor vessel, fiberoptic examination of blocked tubes to confirm nut / bolt / washer i
combinations, visual inspection of the upper internals, visual inspec-tion of the top of the fuel assemblies, visual inspection of the gap between the lower internals core barrel and the vessel nozzle, visual
.i
e
.
.
inspection of approximately one-third of the fuel assemblies during refueling, and radiographic inspection of several drain lines. Based on the results of these efforts, the licensee believes the best prob-able location for the remaining parts is in the bottom of the reactor vessel. The licensee requested Westinghouse to evaluate the effects of these remaining loose parts on the operation of Zion Unit 1.
Westinghouse's evaluation was documented in two reports entitled,
" Evaluation of the Effects of Foreign Objects in the Zion Unit 1 Reactor Coolant System," dated May 1982 and " Analytical Assessment for Effects of Loose Parts, Zion Plant No. 1," dated May 10, 1982.
The inspector reviewed the reports and noted that they were submitted to NRR via letter, F. Lentine to D. Eisenhut, dated May 17, 1982.
In summary, the reports all conclude that the effects on the past core (cycle 6) are no more than already documented and that there is no safety concern for the operation of Zion Unit 1 for cycle 7.
Part of the analysis deals with the sizes of the remaining loose parts preventing most of them from passing up from the bottom of the vessel past the bottom nozzle plate into the fuel region. Those pieces that are small enough to pass the nozzle plate would be stopped at the bottom grid strap. The evaluation also analyzes the effect of all the pieces lumping together to block flow, clad wear due to pieces becoming lodged in an assembly, mechanical loadings on the vessel and internals, and damage to the flux thimble assemblies.
The summary of the analysis states that the presence of the foreign objects in the reactor coolant system will not affect normal plant operation.
It should be noted that the evaluation did not consider the remaining loose parts to be distorted from the initial shape or reduced in size.
It should also be noted that the ten year in-service inspection for Zion Unit 1 is due at the end of the next cycle at which tim the fuel will be off loaded and a detailed inspection of the lower internals and the vessel bottom will be conducted.
The inspector reviewed the licensee's investigation into the numbers and sizes of missing parts and the evaluation of their effect on continued operation.
In addition, at NRR request, the inspector reviewed Westinghouse drawings of the vessel internals to evaluate the licensee's determination of probable and possible location of the remaining parts.
No items of noncompliance were identified.
4.
Loose Parts Monitoring
!
During the time period the aluminum nozzle cover was thought to have been in the ID steam generator, the licensee received alarms on the loose parts monitoring system which are now believed to have been caused by the cover and its parts during the dissolution of the cover.
,
The Unit 1 Loose Parts Monitoring System (LPMS) was made operational
'
around April 13, 1981 during restart operations following the refueling outage. No LPMS alarms were noted initially. The reactor coolant pumps had been running since four or five days earlier and the nozzle
L
e
.
.
cover is thought to have been' intact, pressed up against the steam generator tube sheet, and held there by the reactor coolant flow.
It is normal practice not to energize LPMS until normal operating temperatures and pressures are reached in order to avoid the excess noise and alarms which would be caused by pump operations, heatup, and flow noises occurring while the plant proceeds from refueling through cold shutdown to hot shutdown. Unit 1 was taken critical on April 16 and returned to power operation on April 22 with no LPMS alarms noted during this period. The first LPMS alarm occurred on April 25, 1981.
Initial determination by the licensee was that there was a loose part on the ID steam generator primary hot leg side due to the impacts being localized to the sensors in that area. Babcock and Wilcox (LPMS vendor) was requested to assist in the evaluation.
Subsequently, impacts were observed on sensors located on the steam generator secondary side, the reactor coolant pumps and the reactor vessel head. The LPMS continued alarming, but on April 29, 1981, the
-
evaluation had reached the stage where thc licensee, Westinghouse and B&W concluded that a loose part probably existed on the primary side of the 1D steam generator. Using the impact noises coupled with reactor coolant flow, B&W was able to infer a probable mass of the part.
Since the noises were of low energy content, the part was thought to be small (approximately 1/2 pound).
Since the alarms were intermittent, the part was thought to be sometimes captured.
During this period, the plant was also investigating a steam genera-tor power mismatch (further discussed in paragraph 5), however, the April 29, 1981 meeting with B&W and Westinghouse concluded that the LPMS alarms and the reduction in power in the 1D steam generator were not related.
Babcock and Wilcox documented its investigation in a 12 port No. M-81-267 dated May 18, 1981 with recommendations to continue monitoring LPMS and that the part should be removed at the best opportunity. The alarms decreased in frequency and were last logged on May 18, 20, and June 2, 1981 It is now felt that the nozzle cover was intact and probably held in place by flow until April 25, 1981 when it began disintegrating in pieces which produced the LPMS alarms. The LPMS alarms continued until the aluminum dissolved and the parts were captured or migrated to the reactor vessel bottom.
The inspector reviewed the operation and design of the Loose Parts Monitoring System, the B&W report, the observed alarm history, and Appendix A to the May 17 letter from F. Lentine to D. Eisenhut on Steam Generator Inspection and Repair.
No items of noncompliance were identified.
5.
Other Operating Indications Unit i reached operating pressure and temperature on April 13, 1981 and the twelve reactor coolant flow transmitters (three per loop)
were all adjusted to read 100% flow on that date. The flow trans-mitters had been calibrated during cold shutdown using a deadweight
--
7--
.
.
tester with the transmitters not connected to the loops. Final calibration (the April 13 adjustment) is done in hot shutdown to compensate for a known shift in the instrument readout which occurs when going from a cold, no flow, low pressure condition, to a hot, full flow, normal pressure. While this shift is not fully understood, it is known to occur and normal plant procedures compensated for it by making the transmitters read 100%.
The nozzle cover is thought to have been pressed against the hot leg tube sheet of the D steam generator during this adjustment. The cover was producing a 4% actual flow reduction at this time.
Since the adjustment described above assumes normal 100% flow ccnditions to exist during the adjustment, the three loop D transmitters were set to read 100% flow. Therefore, the adjustment masked the flow reduc-tion and produced completely normal control room flow indications as of April 13, 1981. The inspector reviewed maintenance records of the adjustments made to the D loop flow transmitters as well as to the transmitters of the other three loops. There was no signi-ficant difference between the adjustments made to loop D and those made to the other loops.
It appears as if the blockage, while large enough to produce a 4% flow reduction, was not large enough to produce an abnormal hot shutdown flow adjustment and therefore did not attract the attention of the maintenance personnel performing the adjustment.
Flow indication remained at 100% until April 30, 1981.
By indicating 100%, no correlation could be drawn between the power mismatch which was being observed during same time frame (discussed later in this section). Reactor coolant loop D indicated flow gradually began in-creasing from 100% on April 30 to a max of 104% on May 8,.1981.
At this point, loop D coolant flow was measured locally at the instru-ment taps to be roughly equivalent to 100% from initial startup data and the transmitters were reset to 100% indicated flow on May 9.
The other three loops remained constant or decreased very slightly during this time period. The increase in loop D flow was caused by the gradual dissolution of the aluminum cover, thus removing the flow blockage and restoring the 4% flow reduction noted earlier.
Since the transmitters were already set to 100%, the result was 104%.
Since the flow blockage was not known at this time, a cause was never determined for the reactor coolant flow increase.
Later analysis of the amount of tube sheet blockage and the resultant loop flow reduction which an unfolded nozzle cover could cause agrees roughly with the
'
observed loop flow increase.
Following shutdown at the end of cycle, and discovery of the hinges l
l l
and associated parts in the B and D steam generators, the licensee I
determined the setting of the flow transmitters to 100%, while actual flow was probably 4% less, constituted a nonconservative, loop D, 90%
flow trip setpoint.
In accordance with Zion Technical Specifications, the licensee reported this on February 26, 1982.
The inspector reviewed the maintenance procedures, the recorded data for the flow adjustments, and a plot of the observed flow indications
i
_
_
_
r
.
.
with time. The licensee has performed investigations and experiments to quantify and better determine the nature of the flow transmitter shifts on change of state from cold shutdown to hot shutdown. These investigations were discussed with the inspector and the licensee is changing the hot shutdown adjustment procedure to account _for some of these effects. This change should decrease the amount of flow blockage which could get masked by the adjustment.
In addition, the licensee intends to conduct a special test on restart of Unit 1 to compare loop flow indication with other methods of flow determina-tion such as use of reactor coolant pump power requirements and mass flow balance. The results of this test are intended to te used to develop permanent plant procedures for determining flow.
Across the same time period as the nozzle cover is thought to have been in the steam generator, steam generator B appeared to be producing 8-10% more power than D steam generator as determined by calorimetric calculations. Steam generators A and C showed little or no change from normal 100%. Conversations with plant personnel indicate that initially there was greater concern with B steam gener-ator indicating a higher output than with D indicating lower than normal. The licensee investigated the problem performing additional calorimetrics and trending several of the parameters. However, as with flow, the power mismatch condition appeared to gradually correct itself across the same time frame.
Subsequent analysis of the effect on steam generator power if a flow blockage were present, coupled with the physical configuration of Zion's steam piping, indicates that a decrease in steam generator flow in D, due to a blockage similar in size to a nozzle cover, should produce an equivalent increase in B to compensate since they share the same main steam header.
No items of noncompliance were noted.
6.
Chemistry Following the discovery of the hinges and associated parts in the steam generators, the licensee commenced an investigation to determine what happened to the aluminum of the nozzle covers. Crud samples were taken from the fuel during refueling operations and various locations in the primary system were sampled. Westinghouse analysis of these samples showed no unusual quantity of aluminum present in the crud. Sampics taken of the CVCS hold up tanks, reactor coolant system, reactor coolant filter, CVCS letdown demineralizer effluent showed essentially no aluminum.
During operation for Cycle 6, the licensee had experienced and investigated indications of loss of heet transfer capability in the letdown heat exchanger. Later, the heat exchanger was opened during the outage and lancing of the tubes showed crud with 14 wt.% aluminum (about 4.5 pounds of aluminum was found). The letdown demineralizer used during most of cycle 6 had been discharged and shipped out prior to discovery of the problem and was therefore unavailable for analysis. The licensee believes that essentially all of the aluminum has been removed from the system.
.
.
.
The licensee's Technical Services Department performed an evaluation of chemical mechanisms involved in the dissolution, transport, removal, plate out and wear effects of the aluminum as well as probable future I
'
effects. No significant safety problems were noted or are expected in the next cycle. This evaluation was documented in Appendix D to the May 17, 1982 letter F. Lentine to D. Eisenhut. The inspector reviewed the licensee's evaluation, the sampling results and results of samples taken during the time the cover was in the steam generator.
These last samples indicated no abnormalities because they were not specifically analyzed for aluminum due to no one realizing the cover was in the generator.
No items of noncompliance were noted.
7.
Unit 2 Flow Reduction in Loop "A" RTD Bypass Loop After Unit 2 started up from Cycle 6 refueling in 1981, the unit experienced RTD bypass low flow alarm in loop A.
According to procedures, the bistable inputs into the reactor protection system were tripped and an investigation was commenced. The concern was that an actual low flow in the RTD bypass loop could result in longer RTD response times in transient situations and subsequent reductions in DNB margins. The RTD bypass flow preoperational test was duplicated to determine the extent of the low flow condition.
The results of the RTD flow test showed a decreased flow of 88.1 gpm hot leg and 135.9 cold leg when compared with original preopera-tional data of 94.3 gpm hot leg and 155.7 gpm cold leg. The hot leg flow for loop A corresponding to the limit on RTD response is 65.07 gpm.
The actual flow calculations were reviewed by Westinghouse and found to be acceptable. Efforts undertaken to locate the cause of the low flow consisted of the following:
a.
The flow indicator was recalibrated two separate times using two different methods and found to be correct.
b.
All isolation valves in the RTD bypass line were x-rayed to verify their condition. One valve appeared to not be fully open.
It was fully backseated with no change in RTD bypass flow seen.
c.
Each leg of the RTD bypass was isolated allowing flow data on each leg to be taken, d.
Flow indicator equalizing valve lines were checked for warmth to see if equalizing valve was leaking.
One of the few possible chacks not done was to remove and check the orifice. This was not done because other conditions in the plant made it undesireable to secure flow in the loop; and because a problem with the orifice would mean that actual bypass loop flow was normal and only indicated flow was low.
Subsequently, problems with RTD's caused securing of the loop and allowed removal of the
-
.
.
.
orifice. When the orifice was checked, it was found to be installed backwards.
The orifice was reinstalled correctly and on restoration of the loop, normal RTD bypass loop flow was observed. Subsequent investigation traced the incorrect installation to a maintenance repair item and the licensee's maintenance department revised their procedures to provide better controls on the installation of orifice plates.
The inspector reviewed the onsite -review evaluation of the problem, as well as the adequacy of efforts taken to locate the cause of low flow prior to discovery of the incorrect orifice installation.
No items of noncompliance were noted.
8.
Exit Interview The inspectors met with licensee representatives denoted in paragraph 1 at the conclusion of the inspection on May 27, 1982. The inspectors summarized the purpose and scope of the inspection and the findings.
,
m 9