ML18066A387

From kanterella
Jump to navigation Jump to search
Safety Evaluation Accepting License Request for Staff to Authorize Consumers Energy Proposed Alternative to Requirements of ASME Section XI Article IWA-5250 for Degraded Primary Coolant Pump Casing Bolts at Plant
ML18066A387
Person / Time
Site: Palisades Entergy icon.png
Issue date: 01/28/1999
From:
NRC (Affiliation Not Assigned)
To:
Shared Package
ML18066A386 List:
References
NUDOCS 9902090066
Download: ML18066A387 (5)


Text

-.. ----. c=-=-=========---c-=====~------

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO INSERVICE INSPECTION PROGRAM REQUEST FOR RELIEFRR-13 CONSUMERS ENERGY COMPANY PALISADES PLANT DOCKET NO. 50-255

1.0 INTRODUCTION

During a forced outage on December 15, 1998, active leakage was detected past pump casing gaskets on primary coolant pump (PCP) P-50A. The boric acid leakage had caused wastage on two casing bolts. Wastage had previously been observed on the two casing bolts during the 1998 refueling outage, but the licensee believed the source of the boric acid was a leaking instrument line above the casing bolts.Section XI Article IWA-5250 of the American Society of Mechanical Engineering's (ASME) Boiler & Pressure Vessel Code specifies that when leakage is detected during a system pressure test, the bolting shall be removed, visually examined for corrosion, and evaluated in accordance with IWA-3100. The licensee at the Palisades Plant has requested relief from the requirement to remove the bolting in accordance with 10 CFR 50.55a(a)(3) as a requirement for which hardship or unusual difficulty would be incurred without a compensating increase in the level of quality and safety, and for which the proposed alternatives provide an acceptable level of quality and safety.

2.0 DISCUSSION During the 1998 refueling outage (June 1998), boric acid crystals were detected in the vicinity of two PCP pump casing bolts. At that time, the licensee concluded that the leakage came from an instrument line above the casing bolts. The licensee conducted an analysis and concluded that sufficient margin existed in the closure for continued operation.

Active leakage was detected in the vicinity of the casing bolts during a December 1998 forced outage. In both cases, the casing bolts were cleaned and the remaining diameter measured.

The licensee reported that the bolts have a nominal diameter of 4.500 inches; however, they are actually 4.575 to 4.590 il"!ches in diameter according to the design drawing. The diameters measured during the 1998 refueling outage were 3.970 inches and 3.920 inches. An analysis was conducted and the licensee concluded that the structural integrity of the joint had been maintained. During the forced outage, the diameters were 3. 770 and 3.81 O inches. On December 16, 1998, all 16 closure studs on PCP P-50A were examined using ultrasonic (UT) techniques. The.UT examinations were made to look for rejectable indications according to ASME Section XI, 1989 Edition, Category B-G-1, Item Number 86.180. Acceptance criteria are provided by IWB-3515, "Standards for Examination Category B-G-1, Pressure Retaining Bolting Greater than 2 in. in Diameter." The UT results confirm that the visually identified corrosion on the two bolts is a flaw connected to the outside surface of the bolt. The physical measurements 9902090066 990128 PDR ADOCK 05000255 P

PDR

\\

' i

2 indicate that the bolting has exceeded the acceptance criteria provided in IWB-3535.1 for non-axial flaws greater than 'UJ inch in length.

The licensee used a: linear extrapolation of the wastage rate to predict the diameter of the bolts when the 1999 refueling outage is reached. The licensee then calculated the structural integrity assuming the two bolts continue to corrode at the same rate anc:l determined that the design requirements will continue to be met until the 1999 refueling outage.

The licensee has proposed operation with known leakage and degraded bolts until the refueling outage. The basis for its determination is the acceptable visual and ultrasonic examination results and the engineering analysis performed to IWB-361 O to evaluate identified flaws.

Although the bolting will remain structurally acceptable until the next refueling, the bolts will ultimately require replacement per IWA-5252(a)(2).

In an effort to minimize wastage, the licensee has applied Carbozinc 11 coating to the exposed surface of the degraded studs and to the adjacent studs. The bolts were then encapsulated with stainless steel flashing sealed at the casing surface to protect the bolts and coating material. The licensee has concluded that there will be no adverse reaction from contact between the coating material and the boric acid.

The licensee considered two options to address the problem: the first is to replace the two corroded bolts completely and the second is to rebuild the pump including bolting and gasket replacement. The licensee provided the following reasons for not implementing these options before the 1999 refueling outage. For the first option, replacement of the bolts alone would not stop the leakage; based on vendor experience, replacement of the two corroded bolts could easily increase the existing leak rate requiring additional maintenance; and an estimated 15 person-rem exposure would be incurred to replace the bolting that would have to be repeated during the 1999 refueling outage due to continuing flange leakage. For the second option, this would result in an estimated additional 3-10 weeks of down time for repair crew mobilization, parts procurement, and planning; a dose estimated to be 50 person-rem would be incurred, compared to an estimated 35 person-rem if the work is deferred to the 1999 refueling outage; and additional vendor-recommended pump reliability improvements would not be implemented

. due to insufficient lead time.

The licensee has stated that delaying the repair to the 1999 refueling outage results in a slightly higher chance of a controlled shutdown due to failure of both bolts. This would be a controlled shutdown and not an accident condition. The licensee concluded that delaying the repair does not represent an increased risk to the plant. In fact, the licensee states that minimizing mid-loop operation by only doing so at the refueling would be a reduction in risk. The licensee did not provide a shutdown risk analysis to support its judgement.

The licensee stated that slow degradation of the casing gasket is highly likely to result in a leak that exceeds the Technical Specification leak rate before the joint could fail catastrophically.

Also, leak rates of 0.2 gpm have been experienced at other plants without significant sealing surface degradation.

3 The licensee stated that from discussions with the pump vendor, failure of both bolts would be expected to produce a leak rate of less than 1 O gpm. This is well within the charging system capacity for PCS makeup, but would necessitate a controlled plant shutdown.

The licensee has committed to the following compensatory measures. The licensee stated that it will visually inspect the pump flange area at each forced shutdown prior to the 1999 refueling outage which requires the PCS to be in Hot Shutdown conditions or below. The licensee will perform UT inspection of the degraded bolts at each forced shutdown requiring the PCS to be placed in Cold Shutdown. PCS unidentified leakage will be administratively limited to 0.5 gpm versus the 1.0 gpm allowance provided by plant Technical Specification 3.2.5a. The licensee also committed to perform a confirmatory leak rate determination if the leak rate calculation indicates an unidentified leak rate in excess of 0.3 gpm or if the sump inleakage rate increases by 0.2 gpm in any 24-hour period. If the confirmatory calculation verifies that the unidentified leak rate is greater than 0.3 gpm and the reason for the leak rate is not understood, action will be initiated within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to place the plant in Hot Standby and it will be verified that the indication of leakage was not a result of significant additional degradation of P-50A before the plant is returned to power operation.

The licensee committed to replace the bolting when data indicates that degradation will exceed the limits established by analysis. In any case, the degraded bolting will be replaced no later than the next refueling outage scheduled to begin in October 1999.

3~0 STAFF EVALUATION The NRC staff has written six information notices, a bulletin, and a generic letter on boric acid corrosion of carbon or alloy steel components in pressurized water reactors. Information Notice No. 80-27, "Degradation of Reactor Coolant Pump Studs," was issued on June 11, 1980, and describes a situation that is similar to the current situation at Palisades. It describes the severe corrosion damage found on a number of closure studs in two of four Byron Jackson reactor coolant pumps at Fort Calhoun Unit 1. During a system pressure leak test, visual inspection prior to plant restart indicated saturated and dripping insulation was observed at one of the reactor coolant pump flange regions. Upon removal of the insulation, leakage was found coming from the seating surfaces. After the insulation was removed from the three pumps, wastage was identified on three studs on two different pumps. The most severe wastage occurred on a stud that was initially 3.5 inches in diam~ter, but the stud had wasted to 1.0 and 1.5 inches.

Information Notice No. 82-06, "Failure of Steam Generator Primary Side Manway Closure Studs," was issued on March 12, 1982, followed by Information Notice No.86-108, "Degradation of Reactor Coolant System Pressure Boundary Resulting from Boric Acid Corrosion," issued on December 29, 1986, followed by Information Notice No.86-108, Supplement 1, "Degradation of Reactor Coolant System Pressure Boundary Resulting from Boric Acid Corrosion," issued on April 20, 1987, followed by Information Notice No.86-108, Supplement 2, "Degradation of Reactor Coolant System Pressure Boundary Resulting from Boric Acid Corrosion," issued on November 19, 1987, followed by Information Notice No.86-108, Supplement 3, "Degradation of Reactor Coolant System Pressure Boundary Resulting from Boric Acid Corrosion," issued on January 5, 1995. All of these information notices describe instances where leakage of boric acid solution resulted in significant degradation of carbon or alloy steel components.

--~---*--*-----

-*--~-"'--~*-~ --~*-*-----**

4 On June 2, 1982, the staff issued Bulletin No. 82-02, "Degradation of Threaded Fasteners in the Reactor Coolant Pressure Boundary of PWR Plants." This bulletin requested that licensees develop maintenance procedures, conduct certain inspections, identify bolted closures where leakage has been identified, and to identify any bolted closures where leak sealants have been used to control leakage.

Generic Letter 88-05," "Boric Acid Corrosion of Carbon Steel Reactor Pressure Boundary Components in PWR Plants," was issued on March 17, 1998. This generic letter requested that licensees identify principal locations where leaks smaller than allowable technical specification limits can cause wastage, procedures for locating small leaks, methods for conducting examinations and performing engineering evaluations when leakage is located, and corrective actions to prevent recurrences of this type of corrosion.

The staff has reviewed Attachment 2 of Consumers Energy Company's submittal of December 20, 1998, for relief request No. RR-13 for the Palisades Plant. In this attachment (EA-C-PAL-98-1939-01 ), the licensee performed an engineering evaluation of the structural integrity of the primary coolant pump P-50A joint between the casing and cover with local degradation of 2 of the 16 bolts. In order to seek a relief from the requirement of removing bolts for visual examination per IWA-5250(a)(2) until the 1999 refueling outage, the licE!nsee performed an engineering analysis considering an anticipated corrosion of the degraded studs by the 1999 refueling outage. The licensee calculated the stresses in the degraded studs for loads including stud reload and design pressure using reduced sectional properties of the stud.

The combined stress due to design loads was compared against the design basis allowable loads. The effect of differential thermal expansion and seismic loads on the bolted joint was not significant. The allowable stresses were based on the licensing basis code (ASME Section Ill, 1965). The combined stress in the degraded bolts is within the design basis allowable stress.

Based on its engineering analysis, the licensee concluded that sufficient preload still exists in the degraded bolts, and projects that the preload will be maintained until such time when the bolts are replaced in the 1999 refueling outage, thus ensuring the structural integrity of the joint between the casing and cover of pump P-50A.

The staff has reviewed the licensee's engineering analysis EA-C-PAL-98-1939-01 and finds it reasonable and acceptable and concludes that the structural integrity of the joint between the casing and cover of pump P-50A with two degraded bolts will be maintained until the 1999 refueling outage.

Given the evidence of boric.acid crystals at the vicinity of the two PCP pump casing bolts during the 1998 refueling outage, and that the leakage was coming from the bolted connection, IWA-5250 in the 1989 Edition of the code specifies that the bolting shall be removed, visually inspected, and evaluated in accordance with IWA-3100. Several licensees have submitted relief requests to the NRC staff to use a more recent edition of the code which states that at least one bolt shall be removed rather than requiring that all of the bolts be removed and examined. The later editions of the code state that if any degradation of the removed casing bolt is visually detected, then all of the bolts must be removed and visually examined for degradation. The licensee at Palisades is requesting that due to the accessability of the casing bolts in the two PCP pumps (approximately 4 inches of the casing bolt is visible in the area where the degradation is occurring), the increased person-rem involved with replacing the bolts at this time, and the potential to increase the leakage from the joint if the bolts are removed and inspected at this time, that the removal of the bolts be postponed until the 1999 refueling

I

- - *-~--~------*-

---~*-~~-"---"~------~.... --

5 outage. Any additional degradation or failure of the two bolts may increase the leakage to a rate higher than the administrative technical specification unidentified leak rate.

  • During the closure of Generic Issue 29, "Bolting Degradation or Failure in Nuclear Power Plants," (Generic Letter 91-17) issued on October 17, 1991, considerable discussion took place on the possibility of total failure of a bolted connection. The Advisory Committee on Reactor Safeguards reviewed* the issue of bolting degradation with the staff and industry experts.

General agreement was reached based on a number of years of industry experience that it is extremely unlikely that a bolted connection would catastrophically fail rather than leaking to a

. detectable extent.

Although the staff has not specifically endorsed its guidance, the Electric Power Research Institute (EPRI) has suggested a basis for allowing a joint to leak in EPRI NP-5769, at page 3-2.

The bases are (1) the leakage from the closure is assured before catastrophic failure of the joint under the design basis for the plant, (2) the safety consequences of closure leakage are acceptable, and (3) the margin against break at the point when leakage becomes detectable exceeds an acceptance level. Based on the licensee's analysis, the EPRI requirements will be satisfied.* Leakage will exceed the administrative limits committed to by the licensee before the closure fails when considering design-basis events. The leakage will not impact any other carbon or alloy steel components in the vicinity of the leakage. If further degradation of the flange seals occurs and the leakage rate increases, it will result in detectable leakage before the code margins are exceeded.

Further, the staff has verified that the licensee has no othe~ carbon steel bolting in the reactor coolant system that is degraded and accepted by analysis.

4.0 CONCLUSION

The licensee at the Palisades Plant has presented a basis for delaying the replacement of degraded studs in the P-50A PCP until the 1999 refueling outage. Palisades is requesting that due to the accessability of the studs in the two PCP pumps (approximately 4 inches of the stud is visible in the area where the degradation is occurring), the increased person-rem involved*

with replacing the studs at this time, and the potential to increase the leakage from the joint if the studs are removed and inspected at this time, that the removal of the studs be postponed until the 1999 refueling outage. Any additional degradation or failure of the two studs may increase the leakage to a rate higher than the administrative technical specification unidentified leak rate.

Based on the information and commitments submitted by the licensee, the NRC staff finds the proposal to defer the replacement of the studs until the 1999 refueling outage to be acceptable.

Relief is granted in accordance with 1 O CFR 50.55a(a)(3)(ii) since compliance with the specified requirements of this section would result in hardship or unusual difficulty without a compensating increase in the level of quality and safety as discussed in Section 2.0 of. this safety evaluation, and the proposed alternative provides reasonable assurance of operational readiness.

Principal Contributors: J. Davis B. Jain Date: January 28, 1999