ML18057A749
| ML18057A749 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 02/11/1991 |
| From: | Slade G CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9102210221 | |
| Download: ML18057A749 (9) | |
Text
consumers Power GB Slade General Manager POWERiNii MICHlliAN'S PROliRESS r~b~u ~ryar{ia,nt: I ~910 Blue Star Memorial Highway, Covert, Ml 49043 Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 DOCKET 50-255 - LICENSE DPR PALISADES PLANT -
REQUEST FOR RELIEF FROM SPECIFIC ASME B&PV CODE REQUIREMENT - CONTAINMENT AIR COOLER LEAKAGE - REVISION I Enclosed is a revision to our January 16, 1991 "Request for Relief From Specific ASME B&PV Code Requirement - Containment Air Cooler Leakage." This revision contains an added description of the containment leak detection instruments and methods used to identify small containment air cooler service water leaks and a revised description of how containment air cooler leakage will affect the containment flood level and associated containment equipment.
-~J*)J~
Gerald B Slade General Manager CC Administrator Region III, USNRC NRC Resident Inspector, Palisades Attachment r--- -91<)£'.210221- -91 <J:2 :LT~-- -- .
PDR p
ADOCK 05000255 p~~
/f {}'f7 A OMS ENERGY COMPANY
'{I
ATTACHMENT Consumers Power Company Palisades Plant Docket 50-255 PROPOSED REQUEST FOR RELIEF FROM THE PROVISIONS OF ASME B&PV CODE SECTION XI -
CONTAINMENT AIR COOLER LEAKAGE REVISION 1 Pursuant to 10 CFR 50.55a(g)
February 11, 1991 7 Pages
1 HISTORICAL PERSPECTIVE.
In June of 1986, containment air coolers VHX-1 & 2 were found to have minor leakage. As a result of this discovery a special repair process was qualified (June through August 1986) to repair the cooler tubes. Qualification of this repair process proved very difficult since, due to joint configuration, the code required original joint strength could not be achieved. An air pressure drop test was also develpped (August 1986), with acceptance criteria designed to accept a minor amount of coil leakage due to the difficulties associated with assuring zero leakage due to coil design configuration and materials.
The leaks on these two coolers were then repaired in August and September of 1986.
New leaks were discovered on all containment air coolers in October of 1986.
These leaks were repaired in October and November of 1986. A section of one of the coils from VHX-4 was removed at this time for evaluation and destructive examination. The results showed that the overall poor workmanship of the coils was the primary cause of the joint leakage and is reported in the referenced letter "Containment Air Cooler Repair Report" dated February 17, 1987.
In December of 1986 additional leaks were repaired on VHX-4.
During the 1988 refueling outage a number of leaks on VHX-4 were repaired. It is estimated that this repair effort required approximately 625 man-hours due to the fact that when a braze repair of the tubes was attempted, the heat applied to the repair area would affect adjacent tube connections and create new leak locations.
During the 1990 refueling outage the coils on VHX-4 were replaced with a modified set of coils designed to alleviate, as much as possible, the cause of coil leakage. We will be monitoring this new coil configuration to see if it alleviates the problems experienced with coil leakage and repairs, and given satisfactory experience we intend to proceed with coil replacement for the other coolers during future outages.
Also, during the 1990 refueling outage, leaks on VHX-1 were identified and repairs attempted. After several repair attempts, the leaks on the cooler piping greater than one inch were successfully repaired. The leaks identified on the 5/8 inch tube to header connections could not be totally repaired. The 5/8 inch tube to header connections are brazed connections in a nearly inaccessible location. Attempts at brazing repair have frequently led to adjacent non-leaking connections being affected by the heat applied in brazing resulting in additional leaks. All leaks were repaired until the air pressure drop test could be successfully passed per the pre-established acceptance criteria.
BASIS STATEMENT The Palisades plant containment air coolers (VHX 1-4) are air handling and cooling units located entirely in the containment building. Plant service water from the critical service water header is circulated through the units cooling coils. The units are designed to control containment temperature during normal operations and condense steam during certain accident scenarios to help limit peak containment pressure. The coolers are classified as safety class III per Regulatory Guide 1.26. The ASME Boiler and Pressure Vessel Code
2 (B&PV),Section XI, Table IWD-2500-1 Examination Category D-B, Item No. 02.10, and IWA-5250 requires a Class III system to be examined and to exhibit zero leakage during post-repair hydrostatic testing. Presently, containment air cooler VHX-1 does not meet this requirement and due to the coil design, configuration and materials, this requirement i~ considered impractical for the Palisades containment air coolers.
Each cooler consists of eight sets of coils piped to manifolds for supply and return connections to the critical service water system. Some of the coils' header to tube connections have exhibited leakage and attempts have been made to repair these leaks. While some of the attempts have been successful, limitations of the cooler design, geometry, and materials of the construction of the tube to header, connection make achievement of a complete zero leakage repair of the coils impractical. 10 CFR 50.55a(g)(4) states that components which are classified as ASME Code Class 1, Class 2 and Class 3 shall meet the requirements of Section XI of the Code, to the extent practical within the limitations of design, geometry and materials of construction of the components. 10 CFR 50.55a(g)(5 & 6) goes on to explain what the Licensee does once it has determined that conformance with a certain code requirement is impractical and how the commission may grant relief. The 5/8 inch cooler tube to header connections are the brazed connections in a nearly inaccessible location where attempts at a brazing repair have frequently lead to adjacent non-leaking connections being affected by the heat applied in brazing and subsequently developing leaks. Any connection greater than one inch is more external to the coil itself and experience has shown that repairs can be successfully completed on these pipes without heat affecting any adjacent parts.
In addition to the design limitations, adequate lead time does not exist to redesign, fabricate and procure new cooler coils and install them during the present refueling outage.
We have completed an analysis, and safety review which-show how much leakage can be tolerated from each coil and cooler, and have established a program to periodically inspect the coolers in order to assure compliance with these leakage limits.
As described in FSAR Section 6.3.2.1, each cooler has a sump with a drain, a liquid level switch and an overflow valve. Normally, very little water will be condensed from the air and the small amount of condensation will easily flow out through the drain. If a cooling coil leak or steam leak occurs to cause a flow through the drain greater than 20 gpm, the level in the sump will rise to the liquid level switch and initiate an alarm in the control room.
RELIEF REQUESTED It is requested that relief be granted from the zero leakage requirement during pressure test as specified in ASME Boiler and Pressure Vessel (B&PV)
Code,Section XI, Table IWD-2500-1 Examination Category D-8, Item No. 02.10, and IWA-5250 as clarified by Inquiry Number (IN)89-017, for the 5/8 inch tubing sections of the containment air coolers. As an alternative to maintaining zero leakage, Palisades will or has performed the following actions:
- Small amounts of cooler leakage will be monitored and identified as follows: Two containment sump high level alarm switches (LS-0358 and LS-
3 0360) are located approximately 4" from the bottom of the sump. The sump contains about 950 gallons at this level. Operations Alarm Response Procedure (ARP) 8 presently requires the operators to evaluate the cause for the alarm and determine if the alarm is due to normal leakage or unexpected or abnormal leakage. This alarm response procedure will be modified to readily notice the operators that abnormal leakage may be due to the containment air coolers and that they should be sensitive to the coolers as a contributor to the sumps level. The revised procedure will require the operators to contact the Duty and Call Superintendent if containment air cooler leaks greater than 4 gpm are identified. Cooler leakage as high as 4 gpm and the subsequent increase in containment flood level has been shown to have no adverse affect on plant response to any analyzed accident.
Normally the containment sump is drained before the high level alarm comes in. The control room operators are sensitized to containment sump level as a potential indicator of other in containment system leaks. An increase in the number of times the sump is drained would also be an indication to the operators that some additional leakage is contributing to the sump level.
When the sump is drained it enters the dirty waste drain tank which holds approximately 3800 gallons. From there the water is pumped to one of three 20,000 gallon miscellaneous waste hold tanks. When a miscellaneous waste tank is full it is sampled and processed through the radwaste system.
As part of the plant programs to minimize radioactive waste during the course of a normal work week the operations Plant Support Supervisor for Radwaste, routinely inventories the radwaste system's total volume. Part of this inventory is to trend the dirty waste drain tank level to assure that the plant's normal daily waste water generation is kept to a minimum.
Any noticeable trend towards an increase in normal daily waste water generation would be reported to the shift operations staff to determine the source of the leakage.
Recent experience exists to show that leaks in the containment on the order of one gallon per minute can be readily identified. As an example the plants shield cooling system began leaking at about this rate and it was through the above process that the leak was detected which then resulted in an on-line containment inspection which determined the leakage source. All four containment air coolers are located on the 590' elevation of the containment building and are accessible for inspection during power operation.
- All cooler leaks on piping greater than one inch shall be repaired and tested during refueling shutdowns, in accordance with Section XI, 1983S83 requirements.
- All leaks on cooler piping one inch or less shall be reduced to the lowest possible level within the requirements of the safety evaluation.
- Further minimization of 5/8 inch coil leakage will be pursued by investigating other cold type repair techniques (epoxy, liquid steel).
- A periodic inspection for leakage by visual observation shall be performed
4 on a refueling outage basis in accordance with the Palisades Preventative Maintenance Program. Additionally, the ASME Section XI required pressure tests at each period and interval shall be completed per the Palisades Inservice Inspection Pressure Test Program. For coolers discovered to be leaking, an air pressure drop test shall be performed during refueling shutdowns to quantify leakage. Total containment air cooler leakage will be maintained at less than 4 gpm.
- An actuation of the air coolers sump float switch requires Operations to check the cooler for coil leakage. Should any cooler become inoperable due to leakage the Limiting Conditions for Operation (LCO) and action statements of Technical Specification 3.4, "Containment Cooling", would apply. This LCO allows one cooler to be out of service for up to 7 days before placing the reactor in hot shutdown.
- The replacement of existing cooler coils utilizing a modified tube to header joint configuration appears to be the best long term solution.
Failure studies were performed on containment air cooler coils in 1986 and 1987. Recommendations from these studies led to design engineering and modification of available spare cooling coils. These modified coils were installed in containment air cooler VHX-4 during the present 1990 refueling outage. Palisades will monitor the operation of these coils, and given satisfactory experience with the modified coils, we intend to continue modifying and replacing coils in future outages for the remaining containment air coolers (VHX-1,2&3). Until replacement coils are available, the current repair and testing methods (brazing, hydro test, air pressure test), provide appropriate leakage control.
- The coolers have been verified as being capable of performing their safety function with a minimal amount of leakage without adversely affecting the outcome of any analyzed event or accident. This verification has been documented in the 10 CFR 50.59 safety evaluations for Palisades deviation report D-PAL-86-218 and permanent maintenance procedure SWS-M-6. This verification was repeated in the safety evaluation for this relief request.
This safety evaluation shows that the minimal amount of leakage from the containment air coolers has been determined to not be an unreviewed safety questions as follows:
- 1. The leakage does not increase the probability of an accident previously evaluated in the FSAR.
Loss of service water pressure boundary (ie leakage) is not the initiating event for any of the previously evaluated accidents discussed in FSAR Chapter 14. Therefore, analyzed accident probabilities are not increased.
- 2. The leakage does not increase the consequences of an accident previously evaluated in the FSAR.
An administrative maximum allowable leakrate from the Containment Service Water Pressure boundary of 4 gpm will be established (Following repairs during the current outage, the as left leakrate was less than .2 gpm which is significantly less than the above limit.). While this leakrate is less than that which can be detected
5 by existing service water alarm setpoints, leakrates as low as 1 gpm from other systems in containment have been previously detected by operations monitoring of containment sump and dirty radwaste instrumentation. Operation with a leakrate of 4 gpm would approach radwaste processing limitations and would not support indefinite continued plant operation.
The containment flood level following a large break LOCA (most limiting accident with regard to flooding of containment environmentally qualified electrical equipment) will be increased 3 inches over a period of approximately 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> with a leakrate of 4 gpm. This increase in flood level will result in the submergence of the following equipment (reference EA-GCP-91-02 and EA-PJS-91-01:
PT-0105A PCS wide range pressure transmitter PT-01058 PCS wide range pressure transmitter M0-3010 LPSI injection isolation valve M0-3012 LPSI injection isolation valve It has been verified by analysis that the above equipment is not required to operate once they have reached their accident position on SIS and that their partial flooding poses no new consequences as a result of the accident (reference EA-PJS-91-01).
Analysis has also been performed to support that Containment Air Coolers are no longer required for containment cooling 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> following LOCA and that they may indeed be isolated.
Operation with a leakrate of 4 gpm would result in the addition of approximately 17,220 gallons of water to the containment building over a 70 hour8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> period. This amount is considered insignificant with regard to dilution of boron concentrations considering; the length of time over which the dilution would occur, our ability to inject concentrated boric acid, and the fact that control rods would be inserted.
Therefore, based on the preceding arguments, the consequences of previously analyzed accidents is not increased.
Permitting leakage from the Service Water pressure boundary in containment raises the question of accident leakage from the containment to service water and hence to the environment. This issue has been repeatedly evaluated in the past, most recently in PRC meeting 88-081 on Nov 17, 1988. In that meeting, PRC concluded that some leakage was acceptable and that leakage should not be included in ILRT/LLRT totals for the following reasons:
a) The path from containment to service water and to the environment is torturous.
b) 10 CFR 50 Appendix J, III A.1.6 states that "Systems normally filled with water and operating under post-accident condition, such as the containment heat removal system, need not be vented,"
during ILRT testing.
6 c) The Service Water System outlet piping is a monitored pathway.
- 3. The leakage does not increase the probability of malfunctions of equipment important to safety.
Operation with minor leakage at brazed joints may accelerate loss of braze material at the affected joints; however, due to the relatively stiff design of the coils the leaks would be self limiting and would not increase the probability of catastrophic cooler failure.
This position is further supported by the following:
Past operating history has demonstrated the ability to repair containment air cooler coils to leakrates of approximately .2 gpm and that leaks of this size will not result in significant increase or catastrophic coil failure during power operation.
Coil leakage appears to be induced by pressure transients in the Service Water system which generally occur as a result of system testing, during plant shutdown.
On January 31, 1991 ASME B&PV code hydrostatic testing of the containment service water system (@125 psig) did not increase coil leakage.
- 4. The leakage does not increase the consequences of a malfunction of equipment important to safety.
Implementation of this relief request will have no effect on the consequences of air cooler malfunction. That is, failure of the cooler after operating with leakage will be the same as if the cooler operated leak-free. Therefore the consequences of equipment malfunction are not altered.
- 5. The leakage does not create an accident of different type than any previously evaluated in the FSAR.
Operation with the code relief requested will not introduce any new failure modes and hence no new accident types. therefore the possibility of a new accident is not created.
- 6. The leakage does not create the possibility of a malfunction of a different type than any previously evaluated in the FSAR.
Operation with the Code relief requested will not introduce any new failure modes and hence no new malfunction types. Therefore the possibility of a new malfunction is not created.
- 7. The leakage does not reduce the margin of safety as defined by the plant Licensing Basis.
The change in predicted consequences of an accident are that the flood level in containment following a large break LOCA is raised 3 inches.
Since this increase in flood level does not flood any required equipment, the margin of safety is not increased.
~-- --------- - - - - - - - -
7 REMARKS This request for relief applies to the containment air coolers VHX 1-4.
REFERENCES ASME B&PV Code Inquiry Number IN89-017, "Section XI, 1983 Edition; IWA-1400, IWX-3000, IWA-6000 and IWA-7000, through wall leakage - Repair and Replacement Requirements". This inquiry asked for code clarification on repair of through wall leakage. The Code Committee indicated that it was the intent of the Code that through wall leakage be repaired.
ASME B&PV Code,Section XI, Table IWD-2500-1, "Test and Examination Categories" and IWA-5250. This code table outlines the test and examination requirements for the different examination categories.
Palisades is requesting relief from the "No leakage" Acceptance Standard of Examination Category D-B, Item No. 02.10.
Palisades Periodic Activities Control Sheet X-SYENG046. This is periodic activity that initiates the periodic inspection of the Containment Air Coolers for leakage.
Palisades FSAR 6.3, "Containment Air Coolers". This FSAR section describes the function of the containment air coolers and their sump level switch which actuates to an alarm in the control room if there is a cooling coil tube is leaking or steam is condensing at a rate greater than 20 gpm.
- Technical Specification 3.4, "Containment Cooling". This section outlines the Limiting Conditions of Operation (LCO) and Action Statements for containment cooling systems. The containment air coolers are included within its scope.
Safety evaluation for relief request, PS&L Log No 91-228. This evaluation documents that containment air cooler leakage does not affect plant or public safety.
Plant Review Committee Meeting No 88-081. These meeting minutes document the review of containment air cooler leakage by the Plant Review Committee. In that meeting, PRC concluded that some leakage was acceptable and that leakage should not be included in Integrated Leak Rate Testing (ILRT)/Local Leak Rate Testing (LLRT) totals for the following reasons: 1) The path from containment to service water and to the environment is torturous 2) 10 CFR 50 Appendix J, III A.1.6 states that "systems normally filled with water and operating under post-accident conditions such as the containment heat removal system, need not be vented", during ILRT Testing. 3) the service water outlet piping is a monitored pathway.
Letter PDFlenner to DDCrabtree, Containment Air Cooler Repair Report, February 17, 1987 (FLNR 34-87). This letter describes the results of the investigation by Consumers Power Company and its contractors on the condition of the containment air coolers and gives recommendations for future activities.