ML20214Q199

Forwards Response to 860616 & 0710 Ltrs Re Tube Failure in once-through Steam Generator on Reinjection of Emergency Feedwater During TMI-2 Accident Possibly Causing Primary to Secondary Leaks & Large Release of Iodine

ML20214Q199
Person / Time
Site:	Three Mile Island
Issue date:	09/19/1986
From:	Blough A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:	Davenport D AFFILIATION NOT ASSIGNED
References
NUDOCS 8609240208
Download: ML20214Q199 (7)

Text

4

-Q sg SEP 101986 Docket No. 50-289 Ms. Deborah Davenport 1802 Market Street Camp Hill, Pennsylvania 17011

Dear Ms. Davenport:

This is in response to your letters, dated June 16 and July 10, 1986, to R. Conte of my staff and to me. We answered your questions verbally at a meeting with you and Mr. Eric Epstein in Middletown, Pennsylvania, on August' 27, 1986. As you agreed at the end of that meeting, the meeting was beneficial in clarifying many of the questions you had in your letters.

We have summarized our responses to your letters in the attachment. The docu-ments referenced in the attachment were provided to you separately at the meeting.

Thank you for your interest in our activities.

Sincerely, Original Sisacd By[

Alle.n R. Blough, Chief Reactor Projects Section 1A Division of Reactor Projects

Attachment:

As Stated cc: PDR LPDR Senior Resident Inspector bcc: R. Jones, NRR:PBRS J. Thoma, NRR:PBD6 y

h RI:DRP 9 RI:DRP I: RS M RConte/cpn A81ough HKister 9//6/86 9//6/86 9//[/86 0FFICIAL RECORD COPY LTR DEBRA DAVENPORT - 0001.0.0 09/11/86 .rcoi I

)

hhh Nb P

a 4-ATTACHMENT RESPONSES TO D. DAVENPORT LETTERS Letter dated June 16, 1986 Re: B&W Trial Record, B&W Exhibit 407, Kunder Exhibit

1. The reference describes tube failure in Once-Through Steam Generator (OTSG)

.on re-injection of EFW during TMI-2 accident possibly causing primary to secondary leaks and large release of iodine. What are 1979 and current definitions of tube . failure (size of leak, amoun't of contaminants in such leak)?

Response

There has been no change in the tube rupture definition since 1979. The TMI-1 Safety Analysis Report (SAR) Section 14.1.2.10 (provided separately) assumes a tube rupture to be the double ended break of one tube for ac-cident analysis purposes. However, to provide reasonable assurance of continued integrity of the OTSG tubes, the licensee is required by Techni-cal Specifications (TS) 4.19 (provided separately).to perform in-service testing (IST) of OTSG tubes primarily using the eddy current techniques (ECT). Section 4.19.4 of TS defines various official terminology as to the status of an OTSG tube from ECT data results. It should be noted that the TS is oriented toward detecting degradation before an actual break.

It is unlikely that a tube break would occur before leak detection; it is even more unlikely that there would be multiple tube failures.

To complement the IST on OTSG tubes, certain operational limits are imposed by TS 3.1.4 and 3.1.6 to assure the licensee stays within the bounds or assumptions of the SAR. TS 3.1.4 limits the amount of radioactivity in the Reactor Coolant System (RCS) to assure that 10 CFR 100 limits are not exceeded. It is recognized that releases will occur on a tube rupture via either the atmospheric dump valves or the steam generator cafety valves such as was experienced in the Ginna event. Further, TS 3.1.6 limits the amount of primary to secondary leakage (1 gpm), which is an assumption for the design basis tube rupture event. Excerpts on these applicable TS were provided separately.

The 10 CFR 100 is NRC regulations providing nuclear power plant siting criteria to assure the dose to the public at the site boundary is within limits.

2. OTSG's need more water to operate effectively. TS address dry steam gen-erator situations during an accident / transient and in reference to tube thermal shock, can feedwater always be adequately regulated in flow and thus temr9rature?

I

O i 2

Response

Your statement that the OTSG needs more water to operate needs clarifica-tion. It appears you reached this understanding from discussions with the NRC staff. The amount of secondary plant water flow (feedwater or main steam flow) is determined by the design power at which the main turbine generator is designated to operate. During steady-state full power condition, the amount of water (steam and liquid) in a B&W OTSG is less than that of the other vendor type because it is of a relatively smaller size for that size generation capacity. For a main steam line break, this is somewhat beneficial because the lesser water inventory in the OTSG results in a less severe temperature transient (cooldown) on the primary system. These types of events are rare. On the other hand, for a loss of feedwater transient, there is a more noticeable effect on primary temperatures. The loss of feedwater events are more frequent and this is part of the reason the B&W reassessment is occurring.

For TMI-1 since restart, there have not been significant feedwater regula-ting problems in controlling OTSG 1evel. There have been no events in which a loss of feedwater caused the automatic initiation of emergency feedwater. The licensee has not been in a dry steam generator situation.

Please note, at TMI-1, the main feedwater and emergency feedwater systems are separate systems with separate injection points into the OTSG. Level control is being enhanced with the completion of additional modifications to the EFW system during the next refueling outage.

The licensee has detailed operating procedures to restore a dry steam generator to service, primarily to avoid a tube thermal shock situation.

There are no TS which specifically address dry steam generator situations.

3. Since there are EF and water use problems associated with B&W plants, would those factors affect OTSG integrity under certain conditions (e.g.,

on rapid injection of EFW would thermal shock be inevitable)?

Response

Thermal shock to the OTSG tube is a concern primarily when feeding a dry steam generator, which is a post-event recovery concern. Procedures have been written to alleviate this concern. The NRC staff, in the B&W reassessment, is focusing its attention on the initiation event, loss of feedwater and the resulting challenge to safety systems from the ensuing operational transient.

4 3

4. Wouldn't repeated . transients and feedwater problems increase the risk [of tube failure] particularly in a B&W plant?

Response

The repeated thermal cycles on the plant is one of many factors that is considered in the plant design regarding aging of a nuclear power plant and the resultant degradation of components. Consequently design and operational limits have been established to minimize the number of thermal cycles and to keep track of the cycles. Periodic ins'rvicee testing is also a key factor in monitoring for degradation of various components such as nozzles, OTSG tubes, etc. This permits replacement or repairs before actual failures of the component.

5. If a B&W OTSG has to use more water, wouldn't the problem with thermal shock be more likely to occur resulting in tube failure or other faults at B&W plants?

-Response The B&W OTSG use of water is addressed in question No. 2. The thermal shock concern is primarily when feeding a dry steam generator and is not unique to the B&W design. As with other reactor types and fossil fueled plants, the secondary plant is designed to heat feedwater to assure steam generating components are not thermally shocked or excessively stressed during steady-state power operation. Operation limits have been- esta-blished to assure the feedwater is sufficiently hot enough to enter the OTSG.

6. Because of numerous TMI-2 sequence of events, it appears that the "B" 0TSG 1eak at TMI-2 was never fully dealt with as of this date. Will Region I be looking at the steam generators at TMI-2 in reference to the B&W reassessment?

Response

Region I will not be considering the TMI-2 "B" 0TSG 1eakage in the B&W reassessment. The handling of a steam generator tube rupture event is being addressed by NRC staff under a separate multi plant action item.

At our 8/27/86 meeting we provided you an excerpt from the NRC's TMI-2 Accident Investigation Report. You will note confusion on the part of the operators early in the accident as to whether or not the shell of the "B" OTSG was leaking. The p'rimary to secondary leakage was never quantified and it was not the focal point of the numerous technical problems identi-fied in that report. Later licensee chemistry ^results of the "B" 0TSG secondary side water confirmed the presence of trace radioactive contami-nants, indicating that slight leakage occurred. The investigation report indicated that the secondary side of the "B" 0TSG was isolated. It has

I 4 essentially remained that way ever since. The NRC staff at TMI-2 reports that precautions have been taken by the licensee to assure any leakage will be in-leakage from the secondary side to the primary side of the "B" OTSG.

Any attempt by the licensee to restore the TMI-2 plant to service, along

~w ith its OTSG's, would have to be fully evaluated by NRC staff, which would have to include a review of the "B" 0TSG 1eakage problem.

7. Will Region I be looking at the concerns I raised in reference to the B&W reassessment?

Response

The Loss of Feedwater event as it relates to overall plant response and challenge to safety systems will be considered in the B&W Reassessment.

The OTSG tube rupture accident response procedures are being re-reviewed separately as a multi plant action item by the NRC staff. The lead office in the B&W reassessment is the Office of Nuclear Reactor Regulation (NRR).

Region I is in a supportive role.

Letter dated July 10, 1986, to R. Conte Re: B&W Trial Transcript - Hartman

1. In reference to OTSG safety valve problem, this was one of the biggest design deficiencies that cost the licensee millions of dollars for the job.

Response

It is our understanding that the TMI-2 safety valve problem (1978) was indeed a design deficiency with a valve designed by a specific vendor.

The current TMI-1 safety valve is from a different vendor and the problem noted during restart appears to be unrelated. For TMI-1, there appears to be a valve coordination problem on plant trips between turbine bypass valves, atmospheric dump valves, and safety valves with respect to their.

respective staggered lift settings from 1025 psig to 1100 psig. This issue is being reviewed in the B&W reassessment. In the interim, plant operators are aware of the problem and they can and have taken manual control of the turbine bypass valves to lower OTSG pressure to prevent the repeated lifting of the safety valves (re: NRC Inspection Report Item No.

50-289/85-25-05).

2. Reference is made to the . lack of automatic bypass in the condensate polishing system and design deficiency for condensate polishing isolation valves to fail close on loss of instrument air.

f

I 5

Response

At TMI-1 there is an automatic bypass to the condensate polishing system.

The loss of condensate or feedwater is an analyzed event. Since the TMI-2 accident, improvements have been and will be made to detect the loss of main feedwater and automatically initiate emergency feedwater.

Letter dated July 10, 1986, to A. Blough

1. Are feedwater flow problems (in particular excessive FW with potential for OTSG tube rupture) being addressed by NRC (re: B&W reassessment)?

Response

See response to letter dated June 16, 1986, question No. 7.

2. Re: B&W Trial Record - Hartman Exhibit

" The reference describes TMI-2 condensate hotwell fill design problems resulting in a loss of vacuum, operation of the atmospheric dump (AD) valve, and subsequent failure of an operating bellows for the ADV.

Concern is expressed at TMI-1 on repetitive problems: (a) hotwell level operational problems; (b) recent 1986 bellows failures in TMI-1 main con-denser; and, (c) atmospheric dump operating with tube rupture at TMI-1.

(a) Since the restart, the licensee has not experienced significant problems on hotwell level control resulting in a loss of vacuum.

(b) The use of " bellows" in any industrial complex, including nuclear power plants, varies widely with all types of designs, material uses and' functions. The TMI-1 condenser bellows is a metallic expansion joint used for the extraction steam system which penetrates the con-denser. This is a different type bellows than a diaphragm arrangement frequently used with air operated valve operations applications (as was the apparent case for the TMI-2 event). These two events are not related.

(c) The operation of the atmospheric dump valve or safety valve during an OTSG tube rupture is an analyzed event (see response to letter, dated June 18, 1986, question No. 1).

3. Re: See No. 2 above.

Hartman described steam entering the TMI-2 control building because of an apparent ADV tailure. Concern was expressed on risk to workers and public due to steam release inside the plant along with (radwastes) contaminants. Has this been considered for TMI-1 operations?

__%e -. , _ , _ . . - . _ _ _ . - ,_.--_ ___ _ _ . , _ . , _ _ - . _ _

\.

8 4,

F 6

Response

As noted above, the OTSG tube rupture is an analyzed event, which assures not exceeding 10 CFR 100 limits. Likewise is the main steam line rupture event.

Procedures are in place to combat internal plant release with respect to

worker protection. Although.unlikely, industrial accidents with the main steam system could occur in any plant using steam.

. Releases to the off-site boundary are governed by 10 CFR 20 and 100 requirements.

NRC staff Control Room Habitability review is near completion, with major licensee modifications and testing completed.

! 4. Have there been assessments of the potential for internal and external releases from TMI-1 in a situation involving steam tube leaks,or failures?

' Response l

4 These are covered by the design basis events described in Section 14 of the Safety Analysis Report. See also the response to question No. 3 above.

Telephone Call of July 30, 1986, to F. Young

1. Are there filters on the condenser offgas system?

Response

There are no filters on the condenser offgas system. The primary type of radioactive material passing this point is noble gas. Filters are effec-tive only for particulates and iodine type materials, not noble gas.

2. Are there problems with GPU measuring releases / recording releases?

Response

You apparently referred to a Licensee Event Report (LER) No. 86-04, dated March 27, 1986, in which the condenser offgas radiation monitor was inad-

vertently valved out of service for one shift. There was backup instru-

. mentation available. The LER was reviewed in NRC Inspection Report Nos. 50-289/86-05 and 86-06.

+vt , --i- g gem----w----,y,y + e-wg.--,_, -g+----------+--*e--- +e*--+++e e-----rv'i e-*v--- -* ---e* . ' -

• r----w--'-------'--*--~=+------w----*--e----*="w-w'--*----w'