ML20070M565
| ML20070M565 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/22/1994 |
| From: | Broughton T GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| C311-94-2055, NUDOCS 9405020071 | |
| Download: ML20070M565 (5) | |
Text
~
GPU Nuclear Corporation
. Nuclear "rg w*
Middletown, Pennsylvania 17057-0480 (717)944 7621 Writer's Direct Dial Number:
(717) 948-8005 April 22,1994 C311-94-2055 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D.C.
20555 Gentlemen:
Subject:
Three Mile Island Nuclear Station, Unit I (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Control Rod Drive Testing Contingency Plan
Reference:
1.
March 26, 1994, GPU Nuclear letter to NRC 2.
March 29, 1994, NRC Confirmatory Action Letter to GPUN In our letter of March 26, 1994, GPU Nuclear committed to performing additional ~ control rod drop time testing within three months of-reactor startup. We also committed to submitting our evaluation criteria for the testing and a contingency plan.
The purpose of this letter is to submit our evaluation criteria and contingency plan and to notify you that we currently plan to shut down TMI-l on June 1, 1994, to perform this testing.
The purpose of the testing is to determine whether the actions we have taken to increase the pH of the reactor coolant system and to periodically exercise the rods to a greater length of travel have been effective in improving the parformance of the control rods.
The criteria for determining this are contained in the enclosure.
We request prior approval of our plan to perform additional drops of control rods that show improvement.in as-found drop times but nevertheless still-exceed the 1.66 second time limit specified in the Technical Specifications.
If_ all control rod as-found drop times meet the TMI-l Technical Specification limit, we plan to restart without delay.
In no case would we start up without NRC approval if any as-left drop times exceed 1.66 seconds.
nn c u v a,.,'
9405020071 940422 PDR ADOCK'05000289
'l P
PDR-
- GPU Nuclear Corporation is a subsidiary of General Public Utihties Corporabon.
M l
Document' Control Desk C311-94-2055 Page'2 of 2 We request your review and approval of our plan and evaluation criteria prior to May 27, 1994, to support testing on June 1, 1994.
If you have any questions, please let us know.
Sincerely, khw T. G. Brough n Vice President and Director, TMI 4.
JSS/ emf
. ]
Attachment cc:
M. Evans - Senior Resident inspector R. Hernan - Senior Project Manager T. Martin - Administrator, Region I
. 1
)
l
. i l
s 1
1 Y-
ATTACHMENT C311-94-2055 Page.1 of 3 CONTROL R0D DRIVE TESTING EVALUATION CRITERIA AND CONTINGENCY PLAN 1.
BACKGROUND TMI-l Technical Specification (T.S.) 4.7.1 specifies that the. maximum control rod trip insertion time fr.m the fully withdrawn position to 3/4 insertion.(104 inches travel) shall no en eed 1.66 seconds.
Control Rod Drive (CRD) drop times (" trip time" and "drep time" are used interchangeably) for hot, full flow conditions have historically been less than 1.35 seconds at Three Mile Island (TMI).
Post-refueling testing in the 10R outage (October 1993) was the first occurrence of a TMI control rod exceeding the T.S. limit.
Three rods had trip times of 1.83, 1.72, and 1,81 seconds.
This was unexpected because a mechanism examined during 10R as part of the B&W Owners' Group life extension program was found to be in very good condition.
As a result of finding the three slow-rods, a decision was made to repeat rod drop tests at any shutdown following_
r four months of operation.
On March 17, 1994, during a shutdown for valve repair (10U2), rod drop time testing found 12 control rods that exceeded the TMI-1 T.S. limit. Details of these results are contained in LER 94-02 dated April 18, 1994.
The slow drop times,are believed to have been caused by corrosion product (crud) build-up in the CRDM thermal barriers. The thermal barrier is a large machined component which essentially separates the base of-the CRD motor tube from the upper reactor vessel fluid region at the top ~of the reactor vessel head and is intended to keep the CRD motor tube components at. temperatures well below the reactor coolant fluid temperature.
Our conclusion that crud deposits are the most likely cause is based upon previous internal examinations of three CRDMs which are:
a 1.78 second Oconee-2 CRDM with significant crud deposition, a 1,32 second Oconee-1 CRDM with minor crud derosition, and a 1.26 second TMI' CRDM (10R exam) with no evidence of crud deposition.
The crud is believed to-restrict water flow into the CRDM motor tube during trip motion principally by not allowing the ball check valves ~ to function (either to not open at all or to-partially open) and/or by restricting the clearance between the lead screw and the thermal barrier bushing, thereby causing hydraulic drag.
Dynamic modeling of a dropping rod with restricted water flow results in predicted drop times that' match TM1 rod velocity profiles.
This supports the applicability of this known problem to TMI.
GPU Nuclear has taken the following corrective actions to minimize crud buildup within the CRDM:
A.
The lithium concentration of.the reactor coolant system (RCS) was increased to raise pH to reduce the rate of corrosion.
B.
Bi-weekly control rod movement testing was changed to move the rods to a greater length of travel to promote a better interchange of-RCS water inside the thermal barrier bushing.
The purpose of the drop time testing to be performed'an or about June 1, 1994, is to determine if the condition of the CRDMs has improved since March 1994-when the above corrective actions were instituted.
ATTACHMENT C311-94-2055 Page,2 of 3 II.
TESTING PLAN All 61 non-APSR rods will be drop tested on or about June 1, 1994, to determine as-found drop times. These rese'ts will then be compared to the as-found data obtained in March 1994 to determine further action.
III. EVALUATION CRITERIA AND PLANNED ACTIONS A.
As-Found Drop Times For All Rods <l.66 Seconds If the as-found drop time of all rods is <l.66 seconds, it will indicate that the corrective actions taken have been effective.
In this case, GPU Nuclear plans to restart THI-l without delay.
There is some variability in rod drop times from test to test even without any physical changes in the CRDM.
This variability is estimated to be no greater than 0.1 seconds.
Therefore, if the average as-found drop time of all rods that were <l.66 seconds in March 1994 has not increased >.1 second, TMI will operate through the remainder of cycle 10 without conducting a shutdown for the express purpose of testing rod drop times.
If this additional requirement is not met, TMI will retest all rods in the fall of 1994.
Furthermore, if during the remainder of cycle 10 TMI is shut down for any reason at a point greater than four months after startup, rod drop time testing will be conducted.
B.
As-Found Drop Times For One Or More Rods >l.66 Seconds 1.
Evaluation Criteria If the as-found drop times for one or more rods are greater than 1.66 seconds, GPU Nuclear will take action based upon a comparison of drop time data taken in March 1994 and June 1994.
The comparison will determine if the drop times have improved using the evaluation criteria defined below:
a, no rod which had an as-found drop time <l.66 seconds in March 1994 has an as-found drop time >l.66 seconds in June 1994, and b.
no rod which had an as-found drop time of >1.66 seconds in March 1994 has a drop time >2.14 seconds in June 1994.
An improvement in rod drop times indicates that GPU Nuclear has identified' the cause of the slower drop times and is taking the proper action to correct them.
The criterion of 2.14 seconds was chosen based upon the results_ of dynamic modeling discussed above which calculated a drop time of 2.14 seconds for a mechanism with stuck ball check valves but with nominal clearances -in the leadscrew-thermal barrier bushing ~ area.
t' ATTACHMENT-C311-94-2055 Page,3 of 3 2.
GPU Nuclear Action if As-Found Rod Drop Times For One or More Rods Are >l.66 Seconds But Are Improved From March 1994 Given that rod drop times are showing improvement as defined above, GPU Nuclear will perform multiple drops of these rods to free ball check valves to reduce drop times to less than 1.66 seconds.
If multiple drops result in all rod drop times <l.66 seconds, THI-l will return to power, and the next drop time test will be scheduled for the fall of 1994.
If it is not possible to obtain a_,-left rod drop times less than 1.66 s
seconds for all rods, GPU Nuclear will not restart TMI-l without obtaining NRC approval.
3.
GPU Nuclear Action if As-Found Rod Drop Times For One Or More Rods Are >1.66 Seconds And Do Not Show improvement From March 1994 If as-found drop times are not improved as defined above, as a minimum, one CRDM for a slow rod will be disassembled and inspected.
Based on the results of the inspection, specific remedial actions will be developed which may include replacement of the thermal barrier (s).
Contingency plans are in place for what is believed to be the problem component, the thermal barriers.
GPU Nuclear will have replacement barriers available.
These replacement thermal barriers have larger ball check valve clearances which would be expected-to provide less sensitivity to crud deposition.
i
-