Text

Response Opposing Licensee & Staff Motions for Summary Disposition of TMI Alert Contentions 1.a,1.b,1.c,1.d,2.a, 2.b.1 & 2.b.2.Related Correspondence
	ML20087N667
Person / Time
Site:	Crane
Issue date:	04/03/1984
From:	Doroshow J; THREE MILE ISLAND ALERT
To:	; NRC COMMISSION (OCM)
Shared Package
ML20087N665	List: ML20087N664; ML20087N667; ML20087N671;
References
	NUDOCS 8404040164
	Download: ML20087N667 (86)
	v • d • e

u b;t

.,UENCE

,a

. e*

1 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION ED In the Matter of

)

04 fp2-3 P2:15 METROPOLITAN EDISION COMPANY

)

Docket No. 50-289

)

(Steam GeneratoriRepair )

(Three Mile Island Nuclear

)

Station, Unit 1)

)

TMIA RESPONSE TO LICENSEE AND STAFF MOTIONS FOR

SUMMARY

DISPOSITION TMIA hereby responds to Licensee and Staff motions for summary disposition on TMIA Contentions 1.a, 1.b, 1.c, l.d, 2.a, 2.b.1, and 2.b.2, received by TMIA February 27, 1984, purcuant to 10 C.F.R. 5 2.749.

None of these contentions warrant dismissal as alleged by Licensee and the Staff.

TMIA submits statements constituting genuine material issues of fact with regard to each contention, in support thereof.

This response is based upon discovery material received by TMIA, as well as Licensee and NRC documents which can be found in the public document room.

In particular, TMIA recently came upon an ACRS transcript of a combined subcommittee meeting held January 28, 1983 on the subject steam generator tube repairs.

In addition, TMIA consulted Dr. George C.

Sih, Director of Fracture Mechanics, Lehigh University during the week of March 19, i

1984.

Dr. Sih was able to provide comments on certain aspects of l

Licensee's motion by March.28, (Attachment 2), but due to time limitations, these comments could not be converted into af fidavit form.

l is a'very brief statement by Dr. Sih as presented in l

hearings before U.S.

Senator Arlen Specter in Harrisburg last December.

l B404040164 840403 PDR ADOCK 05000289 9

PDR

r i

. \\

In TMIA's reponse to discovery requests, TMIA broadly maintained i

that there was no reasonable assurance that the kinetic expansion repair program could insure safe plant operation, but that without technical assistance, the precise technical aspects of the program could not be competently refuted.

Until this time, TMIA has tried to locate voluntary technical assistance on all relevant aspects of this case, but has been unable to do so.

Thus, TMIA has found it necessary to respond to the Licensee and Staff motions without technical assistance, save what invaluable help Dr. Sih has been able to contribute in his particulary area of expertise.

However, even without competent technical help, TMIA has uncovered a sufficient number of genuine material issues of fact to withstand these motions.

Meanwhile, although TMIA suspects many of the more technical aspects of Licensee and Staff doguments contain additional issues, TMIA has of necessity been forced to address only those aspects of Licensee and Staff documents which raise the types of obvious questions which lay people can understand, in supporting these contentions.

In conclusion, there are genuine material issues of fact requiring adjudication of the above referenced TMIA Contentions.

Summary disposition should not be granted.

Respectfully submitted, Three Mile Island Alert D%M CD LCh (Jpanne Doroshow April 3, 1984 Youise Bradford

UNITED STATES OF AMERICA o

NUCLEAR REGULATORY COMMISSION In the Matter of

)

METROPOLITAN EDISION COMPANY

)

Docket No. 50-289

)

(Steam Generator Repair)

(Three Mile Island Nuclear

)

Station, Unit 1)

)

TMIA STATEMENTS OF MATERIAL FACTS AS TO WHICH THERE ARE GENUINE ISSUES TO BE HEARD A. TMIA CONTENTION 1.a

1. TMIA contention 1.a alleges that Licensee' post-repair and i

plant performance testing are inadequate to provide reasonable assurance that tube ruptures will be prevented.

2. Since Licensee and the Staff both insist that Licensee' post-repair and plant performance testing are simply meant to provide

" additional assurance" that ruptures will be prevented, but that primarily reliance is being placed upon Licensee's qualification program and in-process inspection of the kinetic expansions, see Licensee's Statement of Material Facts

(" Licensee Facts") 1144, 56-57, the adequacy of Licensee's qualification program and in-process j

testing must first be determined.

See also, NRC Staff Motion for Summary Disposition, p.

4.

l The kinetic expansion repair.

I

3. Contrary to Licensee's implication at Licensee Facts 1 8, the repair program which Licensee has undertaken in this case if far from routine, and there is no evidence these types of repairs have ever been i

conducted at a nuclear power plant in the large scale manner as has l

been done at TMI-1.

Indeed, the Staff has always considered the process unigpe and experimental. Attachment 1.

l r

-~

-, ~ ~,-- -~, --

,- +

. 4. Further, the fact that the kinetic expansion repair may have been used in other steam generators, Licensee Facts 1 8, is irrelevant without come evidence of its previous success rate.

Qualification testing.

5. There are additional genuine material issues of fact concerning the accuracy and sufficiency of the qualification testing done by Licensee.

6. According to Licensee, the qualification program was to demonstrate that "the expansion joint meets licensing basis," and is "at least as effective as the original rolled and welded joint."

Licensee Facts 1 15.

Thus, by definition, the program itself was not meant to be a comprehensive test to determine compliance with GDC 14, 10 CFR Part 50, App. A, i.e.,

that the steam generator tubes have "an extremely low probability of abnormal leakage, of rapidly propagating failure, and of gross rupture.

Licensee Facts 1 12.

The program was only designed to see if the actual repaired joint will meet the GDC requirement, but was apparently not designed to analyze the propagation of fatigue cracks in the tubes.

7. Through qualification testing, a qualification criteria was established for kinetic expansion as follows: tubes to be expanded must have a 6" ECT defect free expansion length, a minimum of two inches lef t unexpanded between the tube and the lower face of the tubesheet, a

defect free 1/8" to 1/4" transition zone, and a < 40% throughwall crack.

Licensee Facts 1 9-11.

It was determined that tubes which met this criteria could be safely expanded.

Indeed, except to the extent that a 6" expanded joint and the transition zone are directly affected, potential problems arising from the type and extent of tube failures

+-

v v

. due to IGA and the corrosive environment, were expressly not part of the qualification program.

Quality and Quantity of Samples Tested

8. The qualification, as far as it goes in relation to GDC 14 requirement, is seriously flawed.

First, there is a serious issue regarding the reliability of archive tubes for pullout and leak tests.

Licensee Facts 1 17, 36; See also Reference Document 19 at p.

26.

9.

The accurate simulation of actual TMI-l conditions is an obvious prerequisite to establishing the validity of tests run on archive tubes.

10. The question of whether laboratory conditions can be accurately simulated is at least an open question, since the actual chemistry and sequence of the orginal tube cracking, and the chemical agent which caused the cracking, are not known and to some extent are in dispute.

11. As most clearly put by Licensee during discussion before the ACRS concerning the issue of system cleaning, There are a number of unknowns which actually increase the uncertainty of predicting this potential for further corrosion.

And these unknowns are addressed here as the next four points.

We really don't know what the-total amount of sulfur is in the reactor coolant system.

Sure, we have done sampling to try to indicate how much is there, but we really don't have a definitive value for the total amount of sulfur.

Even if we did have that information, we don't know what the threshold value for deposited sulfur in corrosion film, what the threshold value is for that to cause corrosion of sensitized Inconel 600 in PWR environments.

We are sure that that threshold is very high during operating conditions with lithium chemistry control.

However, we have not done the testing to investigate every possible condition that one could get into going to power operation and then back to an oxidizing shutdown condition, but even if we knew this, if we knew the threshold value,.we knew the total quantity of sulfur in the system, we still would not

. have all the answers we need.

We really do not know the detailed conditions that can in fact produce metastable sulfur states from nickel sulfide that is already on the tube surfaces.

Furthermore, we do not know the conditions that produce that.

We don't know what the lifetime of those states are during operating conditions.

Even with that lithium control. The lifetime may be so short that we don't get corrosion, but we have no testing and no data in the literature that can really give us a good handle on this factor.

ACRS Tr. at 255-256.

See also, TPR 5/19/83 at p.8.

55. Licensee's fracture mechanics methodology was criticized for is failure to address accumulative damage, by Dr. George C. Sih, Director of the Institute for Fracture and Solid Mechanics, Lehigh University, in testimony presented before Senator Arlen Specter during recent hearings held in Harrisburg..

Dr. Sih further stated, with regard to the, fracture mechanics theory used, that it is not justifiable to claim as the state-of-the-art which is irrelvant to the safe eveluation of nuclear reactor components, other means and approaches for evaluating the damage caused by surface flaw have been available in the open literature for many years.

p.4.

56. See also, the report of Staff consultant Dillon, at p.

7,

("I understand the fracture mechanics calculations of residual tube

... ~...

=

. properties in circumferentially cracked tubes are presently unsupported by experimental data).

57 Further, the TPR refers to the "large extrapolation of a 1mited crack-propagation-rate data base," making a firm conclusion on crack propagation "hard to substantiate " TPR 2/18/83 at 10.

By its May revision, the TPR still felt extrapolations were necessary, despite additional data supplied to it, prompting it to suggest additional i

tests be done which Licensee has said it will not do. TPR May report at 5.

See Licensee Reference Documents 63, 64.

This problem is of obvious concern to the TPR.

The Staff simply ignores this recommendation.

58 It is possible that a crack could propagate throughwall in one day.

ACRS Tr. at 78-79 Futher, plugged tubes which are not stabilized may continue to degrade and severe.

ACRS Tr. at 33.

59.

ECT randomly conducted after kinetic expansion repair revealed 15 tubes with small pits and scratches below the sensitivity of the.540" standard differential high gain probe.

SER at 14.

The Staff decided to rely on mid-cycle ECT to " confirm the decision that they are acceptable," but was clearly unable to confirm this at the present time, providing no explanation as to why this has no safety significance.

2 60 If small undetected defects which exist close to the lower portion of the tubesheet propagate during a transient and the tube shrinks or the joint slips, an actual guillotine break could occur.

i Further, the same could happen to tubes with free span cracks which have been plugged but not stabilized.

Thus, an "in-sheet" break can not be guarenteed.

i i

f

=

y y

y,,

v

-owy vm.

,,.9.,y--.

..-6.sme

-s.

r.

e

..-,,y,

. 61. Further, the TPR orginally recommended that because the particular cracks at TMI-l are stress ccrrosion cracks, tubes with ID indications <40% throughwall be plugged.

TPR 2/18/83 at p.

6.

Three months later, the TPR withdrew this recommendation, indicating it felt assured that cracks of critical size would be detected in enough time to avoid breakage.

The Staff seems to go along with this.

Yet as has already been discussed, supra, Licensee's criteria and method for determining critical crack size is a major open question.

Further, the TPR's original point was that the stress corrosion nature of the cracks warranted this precaution.

This point is simply dropped by the TPR, and not addressed at all in the SER.

62. The TPR also recommends that tubes within three rows of the lane region and the wedge-shaped region at the periphery which have OD indications at the 15th support plant or higher, be plugged as has been done in other plants.

TPR 2/18/83 at p.

6.

By May, the TPR withdrew this on this recommendation also, insisting it is sufficient that GPU only plug the first row.

The SER does not discuss the issue.

But the original point was that this is standard in all steam generators.

There is no explanation rrom the TPR or the Staff why the TMI-l steam generators warrant an exception.

63. Finally, the decision of Licensee and the Staff to ignore or discount the significance of potential ECT errors in qualifying tubes for kinetic expansion repair, and in determining which tubes can safely remain in operation without plugging, belies such concerns as articulated by staff consultant Dillon, "the policy which dictates kinetic expansion and resealing of qualified tubes and plugging of those tubes not suceptible to successful resealing infers a crack detection system of considerable reliability.

. The presence of even a few tubes capable of open-ended breaks is a matter of serious safety co.ncerns.

Dillon at 9.

In-Process and Post-Repair Testing 64.' Licensee maintains that meeting the " design basis " through the qualification program and in-process repair testing program "provides the same reasonable assurance that tube ruptures will not occur during any postulated transient."

Licensee Facts 1 57.

65. It has previously been demonstrated that the qualification program is deficient, see, supra.

The random in-process testing which was done is additionally inadequate to provide any confindence repairs will insure tube integrity.

Thus, the design besis has not been met through the qualification and in-process testing program.

66. Moreover, by failing to run the system through some hard transients in post repair testing, there is no technical basis to conclude the repairs can safely withstand such transients.

67. Clearly, no amount of qualification testing can insure that each individual tube has been properly expanded, or that it meets the qualification criteria.

Each tube is unique and the only way to be sure each was properly expanded is to examine each of the 29,000 tubes individually, or to run the system through the actual loads which it was qualified to withstand.

68. Licensee would probably argue that the sheer number of tubes which had to be reparied made it impractical for Licensee to conduct 100% profilometry verification or post expansion diameter guag~ing and depth check samplings.

Licensee Facts 1 41, 43.

69. Yet clearly, the individuality of each tube and its physical i

properties, as well as its surrouding environment, raises questions

)

<i

. concerning potential problems.

And while Licensee may have tried to take precautions to avoid minimize the risk of such problems, they are of sufficient concern to have at least demanded more extensive post repair testing.

70. First, while Licensee tried to take precautions to remove, before expansion, impurities other than corrosion in the gap between the tube and tubesheet, the potential for the existence of such impurities was recognized by the Licensee.

ACRS Tr. at 88.

71. Second, tubes in the outer periphery of the tube bundle, where a majority of tube defects are located, may have " caused problems" because of difficulty in injecting the candle in those tubes.

See, p.

1 (inside).

72 Third, not only were there occassional misfires, see ACRS Tr.

at 116, Reference Document 18 at 52. 3. 5. 7, but in at least one instance, a tube misfired and its location could not be more precisely defined than "one of the first 7000 tubes expanded in steam generator B."

Id.

GPU chose not to reexpand those 7000 tubes, stating,

[I]n the unlikely event that the unidentified tube joint slips, the worst case assumption would be for it to lock up in the axial direction and lower the tube under some conditions.... A lowered tube could conceivably wear against a neighbor and cause a rupture of that tube.

An analysis has been made that indicates that plants can safely shut down in the event of a tube rupture.

Id. at 2-22-23

73. Such a rationale is obviously inadequate.

Licensee's stated justification for proceeding with these repairs is the assurance that such a rupture, particularly one which can damage surrounding tubes,

. can not occur.

The safety significance of this is a ger.uine material issue of fact.

74. Given the potential problems, as well as uncertainties regarding qualification testing, it is somewhat remarkable that post-repair testing was so limited.

75. For example, despite Licensee's claim that "the hardening effect on both the inner and outer surfaces of mechanically expanded tubes is more pronounced than in the kinetically expanded tube," and thus " kinetic expansion may be expected to be less susceptible to stress assisted corrosion cracking than the mechanical expansion,"

Licensee Facts 1 22, Licensee did no post repair hardness testing on corroded tubes, claiming such tests are not required to support any conclusion concerning the effectiveness or adequacy of the repair process."

Licensee Response to TMIA Interrogatory 15.

See Attachments 2 and 4, p.

2.

76. In addition, when questioned before an ACRS subcommittee on a major issue of concern, i.e. whether kinetic expansion may have caused the tubes to weaken, thus increasing the risk of tube failure, Licensee responded that they had examined the issue only peripherally, explaining, We have looked at wa'll thinning due to the explosive expansion process compared to hardening due to rolling.

Hardening is a much less wall-thinning operation.

What that says exactly I'm not sure, but that's the kind of a comparative statement between the two, between the ratcheting.

ACRS Tr. at 165. (emphasis added).

77 Second, the only post repair plant performance tests performed were the bubble leak test, Licensee Facts 1 46, and the hot functional test, where the steam generators were put through normal operating

i

. conditions.

Licensee Facts 1 48 et seg.

These tests can not overcome the already demonstrated deficiences in the qualification program for several reasons.

78. First, leak test results may be misleading.

Licensee has claimed that leaks are self-sealing because. corrosion products will deposit in the cracks and seal the leaks. -1

ACRS Tr.

~

at 99, 100.

Further, as the Staff points out, due to the loss of t

pretension, the leakage rate for various threshold cracks may be reduced.

SER at 21.

Thus, decreased leakage may mask cracks which additional compresive loads and bowing could cause to mouth open, or top create new corrosion initiation sites.

See Attachment 1, at p.2.

79. Further, Licensee has remarked that there is inadequate

[

technical data to really know the significance of corrosion sealed cracks as they impact on tube integrity.

ACRS Tr. at 99-100.

80. Second, the repaired system has not been run through any transient conditions, such as those listed in TMIA Contention 1.a or a MSLB.

Licensee asserts that by qualifying tubes to withstand a 3140 lb j

pullout load or 1025 lb. compressive load, there is no need to test the f

system out._Yet there is a clear need to determine if these tubes actually can withstand these loads while maintaining tube integrity.

See, supra.

81. In response to TMIA Contention la, Licensee discusses the impact of one transient: inadvertent actuation of emergency feedwater

.at high_ power.

Licensee Facts 1 62 et s'eq.

Licensee states that if emergency feedwater is injected into the steam generator, which L

Licensee asserts is unlikely, the resulting thermal stresses will not be enough to cause a rupture because the location of any thermal shock F

. would be remote from the repaired portion of the tube.

Licensee Facts 1 65.

TMIA members are not technical experts and thus we do not know the precise location where the EFW may strike the steam generators.

Yet it seems clear that no matter where the direct thermal shock is, the increased load will pull on the entire tube, thus increasing risk

~

of pullout anywhere on the tube, particularly including areas of high residual stress like the transition or HAZ zone.

The effect of this particulary transient has not been adequately explained by Licensee.

82. Further, hot functional testing did not simulate the stresses which would result from a rapid cooldown following a LOCA, which by Licensee's own estimates would be 2641 lbs.

Licensee Facts 1 67.

83. Moreover, at TMI-1, 31,000 tubes failed, and 29,000 tubes were kinetically expanded.

The TMI-l steam generators are is considered by the NRC to be the worst damaged ateam generators in the country.

See, Statement of Harold Denton, Director of Nuclear Reactor Regulation, before the Committee on Interior and Insular Affairs, February 1, 1982.

the amount of damage can not compare to that of any other steam generator in the industry.

Yet the accident consequences of one rupture i.e. what can be expected in a normal, design basis steam generator, was of sufficient concern to cause the Staff to write in 1982:

During postulated accident conditions, such as main steam line break (MSLB), feedwater line break, or LOCA, the S.G.

tubes are subject to increased pressure differentials and possible pressure waves (e.g.,

subcooled decompression phenomena) and vibrational

loadings, These loads increase the potential for failure of degraded S.G.

tubes, which could exacerbate the accident sequence.

In the event of MSLB, failed S.G.

tubes would provide a leakage path from the primary to secondary system and several potential leak paths.for radioacitivity to the environment would then exist.

In the event of a LOCA, the core reflood rate could be retarded by seam binding....S.G. tube failures

l

. would create secondary to primary leak path which aggrabates the steam binding effect and could leak to ineffective reflooding of the core....Large MSLBs and LOCAs are considered extremely low probability events, but are postulated as bounding conditions.

More realistice events might include small and intermediate size MSLBs and LOCAs.

Although these postulated accidents pose a less severe challenge to S.G.

tube integrity, tube ruptures leading to or following such events could have serious consequences.

SECY 82-72, p.

3,.

84 Licensee and the Staff attitudes that qualification testing is sufficient to guarentee that this system can withstand transients, belies such serious concerns.

License Conditions

85. Licensee has assured compliance with certain required " license conditions" to provide assurance against possible tube ruptures.

Licensee Facts 1 51.

86. One condition is a requirement of plant shutdown if increased leakage of.1 gpm is detected.

Licensee Facts 1 52.

While this limit may be only 10% of the technical specification current limits, the tech specs are themselves "the most liberal in the PWR industry.".

87. As has already been discussed, leak rates may indeed be misleading, and may be inadequate to detect cracks which propagate thruwall in one day.

See, 1 78

, supra.

88. Further, as the Staff points out, due to the loss of pretension, the leakage rate for various threshold cracks may be reduced.

SER at 21 Thus, decreased leaks may mask cracks which additional compresive loads and bowing could cause to mouth open.

89. Also, the TPR's support for the established administrative limit of.1 gpm is based in part on a fracture mechnics analysis which

. can be questioned.

See, 1154-5 supra.

Moreover, according to resumes supplied by Licensee, no one on the TPR has any expertise in the fracture mechanics field.

See, Licensee Response to TMIA Interrogatory T-2.

90. Another License Condition is a promise to conduct a special ECT after either the first 90 or 120 of operation.

Licensee Facts 1 53.

Even apart from the problems raised by ECT, (see, 11 supra), a one time ECT can hardly guarantee that as the plant ages, cracks will even be noticed.

The Staff originally believed the

" prudent" approach to be at least on ECT after 30-60 days, followed by one 'after 150-210 days, and then during refueling..

No explanation for this reversal of position is provided.

91. In addition, power ascension will be at staged intervals.

The TPR, however, recommeded " substantial" extended operation at low power, and even suggested operation with one steam generator at a time at high power.

92 Licensee also will rely on long term corrosion tests to simulate operating conditions.

As discussed previously, accurate simulation of actual TMI-1 tube properties is virtually impossible.

See, if, supra.

93. In conclusion, there are genuine material issues of fact regarding TMIA Contention la.

l

. B.

TMIA CONTENTION 1.b

1. Licensee has determined questions concerning simultaneous tubes ruptures to be irrelevant.

Licensee Facts 1 68-71.

Licensee disputes concerns expressed by ACRS subcommittee Chairman Paul Shewmon and similar concerns in SECY 82-72.

2. The Staff asserts that the Paul Shewmon's memo, as further interpreted by Richard Major, is not supportive of the contention because Paul Shewmon was not concerned with the risk of simultaneous ruptures, but rather with tube plugging and soley with free span defects.

Staff Facts 1 2, 3,

8-10

3. The Staff's position clearly misinterprets the clear language of both the Shewmon and Major memos.

Both address two concerns, one being tube plugging, the other being simultaneous ruptures.

4. Significantly, the Staff did not obtain an affidavit from Shewmon himself.

The Staff's twisted interpretation is based on an affidavit of someone who had no first hand knowledge of Paul Shewmon's intent, and whose interpretation should be given no weight.

5. On the other hand, in the absence of an affidavit from Shewmon, his remarks during an ACRS subcommittee meeting, which both Licensee and'the Staff attended, should be entitled to a great deal of weight.

u Shewmon stated there, L

From a person 1 viewpoint, it seems to me that the thing you have to show is the odds are vanishingly small that you're going to have trouble on both steam generators at once because of the excursion of faults you have had heretofore.

And that may be an impossible problem, but it seems to me that that, at least in my mind, is a critical question rather than fatigue cracks.

ACRS Tr. at 159.

. 6. Further, Shewmon's concern was supported by the TPR's first report, which concluded " safe operation of the TMI-l plant after repair of the steam generators will be dependent on several remaining major activities; [ including] completion of analysis including... the contingency of multiple tube ruptures.

TPR 2/18/83 at 4 Clearly, the TPR meant to distinguish the contingency of multiple tube ruptures from

~

other possible situations which were tested and analyzed by Licensee.

Licensee asserts that "these comments were intended merely to flag to the reader that the conclusions drawn were incomplete at that time since Licensee had no completed its analytical or planning efforts."

Licensee Response to TMIA Interrogatory T-8.

7. Three months later, the TPR, for unexplained reasons, withdrew this as an open issue needing resolution before plant start up.

The basis for the TPR's later conclusion that such analysis was no longer required is unexplained and at least raises suspicious questions.

See, TPR 5/18/83 at p.

2.

This is particulary true since the TPR was apparently sufficiently concerned with the possibility of simultaneous ruptures that it suggested running one steam generator at higher power than the other, which would have put this system in so abnormal a configuartion that GPU refused to do it. TPR 2/18/83 p.

12; Reference Document 64.

8 It is clear that during all qualification testing done by Licensee in 1982, the consequences of multiple tube ruptures, which including ruptures in both steam generators, was never treated as a subject warranting special testing.

And in fact, Licensee later asserts with regard to the simultaneous rupture case, that since this is "not a design basis accident for any plant, neither the TPR nor the

. Staff have required analysis for such an event in their respective approval for returning the plant to service."

Motion at p.

26.

9.

Clearly, as Paul Shewmon noted above, the number of failures and unique type of repair used in the TMI-l steam generators demand that the risk and consequences of simultaneous ruptures receive special attention.

The Staff and Licensee can not simply close their eyes to a contingency which the Staff has already considered of major importance with regard to normal steam generators.

(See, SECY 82-72, where the Staff notes that if ruptures occur in both steam generators, unless the plant can be rapidly depressurized and brought onto Residual Heat Removal, there is the potential to continuously lose ECC water outside the containment.)

10 Further, while the Staff maintains such a possibility is unlikely, no probabalistic risk assessment has been done.

Attachment 7.

11. In contradiction to their states position that the simultaneous tube rupture possibility requires no separate analysis, Licensee and the Staff both recognize that this contingency must be considered in operator training and emergency procedures.

See, TDR 406.

12 However, whether emergency procedures will provide adequate guidance to instruct operators in the event of such an accident is a significant open issue.

There is no question that when a simultaneous rupture occurs, no automatic system can cool the plant down.

This type of accident requires the operators to respond with spur of the moment decisions, so that training is crucial.

Further, operator instructions for simultaneous ruptures currently instruct operators to follow the

. steaming, filling, and isolation criteria as written for single tube ruptures, which themselves violate a number of past safety limits.

13.

For example, the procedures reduce the subcooling margin in the RCS, which risks the formation of steam bubbles in the reactor and reactor coolant piping which could block the circulation of cooling water through the core.

The Staff says that this will not occur, because operators have been instructed not to let it occur.

Staff Response to TMIA Interrogatory 92.

But see, 1 19, infra.

14. Further, reducing the subcooling margin may violate the " fuel in compression" limit, which could cause fuel rods to swell or balloon and thus block or reduce the cooling water flow between the fuel rods.

The Staff acknowledges that this could occur, but rationalizes that because steam generator tube ruptures are expected to occur at cooldown, thus involving only moderate to low cladding temperatures, the affect on the fuel will be negligible.

Staff Response to TMIA Interrogatory 93.

This assumption relies entirely on a possibly incorrect interpretation of the original tube failure scenario.

See Contention 2.a, i'nfra.

15. As is indicated in TDR 406, Rev.

1, p.4, the existing tube to shell delta T at TMI-l had been 100 *F.'

But Licensee discovered that before tube /shell delta T exceeded 100 F, the leaking tube was placed under tensile stress and the tube was pulled into a circumferential tear.

Thus, Licensee and'the Staff have required operators to keep the delta T to 70 *.

This is a clear safety measure meant to eliminate the tension or load on the tubes in the event of a transient which could result in tube breakage.

However, the NRC has no instrumentation requirements to measure delta T because the Staff claims such instrumentation is not safety related.

See, Staff

. Response to TMIA Interrogatory 93.

Thus, there is currently no assurance that Licensee's equipment for such measurements, if they have any at all, is reliable.

16. Second, there is no requirement that the plant computer be operable during plant operation.

The Staff asserts that if computer capability does not exist, it is sufficient if the operators rely on estimations of delta T based on past cooldown rates.

Id..

Further, the Staff indicated that it really does not know what the offect maintaining delta T at 70* will be on total cooldown time, but is hoping it will be small. Staff Response to TMIA Interrogatory 94.

17. Third, as the Staff admits, maintaining the delta T at 70 depends squarely on the ability of operators to precisely modulate the controls.

Thus, their training is crucial.

See, Staff Response to TMIA Interrogatory 95

18. Clearly, considering that these particular types of accidents depend upon the operators to precisely respond, they must not only have complete information, which is questionable 1 16, supra.

19. But the operators must be extremely well-trained.

Yet at p.

21 of the most recent TDR 406, a " comment" indicates that operators who were being trained in the use of the revised guidelines found the training to be of " dubious value" and B&W would not endorse the i

l material.

This raises extremely serious safety concerns when considering the environmental contamination which is risked in the event of a simultaneous tube rupture.

20. Thus, TMIA Contention 1.b raises genuine material issues of fact.

. C.

TMIA CONTENTION 1.c

1. Licensee's position regarding Contention 1.c is that the kinetic e.pansion repair process did not in any way " weaken" the tubes or otherwist affect the retention capability or leak tightness of the plugs.

See, Licensee Facts 1 75, regarding roll plugs; Licensee Facts 1 92 regarding weld plugs.

2. The TPR initially raised the question of whether tubes may have i

been weakened.

In discussing stress levels in the steam generators, the TPR asked if they "can be expected to be significantly higher, or the strength of the tubes significantly lower, than those in a normal OTSG?"

TPR 2/18/83 p. 14-18.

It concludes, "from the information received, the answer appears to be negative, with a minor qualification regarding the undetected defects that are left in the tubes." (emphasis added).

The TPR also states that small defects could leave tubes in a weak condition.

Id. at 16.

3. The TPR's reasons that t' e kinetic expansion repair of the tubes could affect the stress levels, if the process would change the strength or dimensions of the tubes, but say further that the kinetic expansion repair process is not expected to affect significantly the stress levels in the tubes.

The question of whether the strength or dimensions of the tubes were changed, to what degree they were changed, the significance of that change, and the question of how stress levels were changed, are simply left hanging by the TPR, or the SER.

4. The question of whether tubes may have been weakened was also raised by the ACRS as follows:

"Perhaps that doesn't have any influence on stresses, but you do, it seems to me, increase the process of ratcheting up there from thermal expansion, and you have also 1

4

~.

[

thinned the wall out where you have gone through this procedure.

Have you looked at this as a possible failure mechanism?

ACRS Tr. at 165-166.

The question is answered in an ambiguous, inconclusive manner.

See, Contention 1.a, p. 22.

l

5. Further, while relying on comparative hardness tests between the kinetically expanded and mechnically rolled tubes to test for tube weakness, Licensee admits / that it conducted no post-repair hardness testing.

Id.

6. It is well-known that increases in hardness results in a reduction in toughness.

Hence, the repaired tubes with increased i -

hardness obviously suffer a reduction in toughness and hence are no long restored to their original state.

See, Attachment 2, p.

3.

Sih raised similar concerns in his statement before Senator Specter's committee..

i

7. Second, regarding rolled plugs, Licensee states that "since most cracking stopped just below the seal weld before the rolled portion of the tubes began, cracks would not be in the area engaged by the plug," Licensee Facts 1 80, and that the cracks which were found within the tube rolled region were circumferential and tight which would not affect the ability of the plug to hold.

Licensee Facts 1 81, 83. Licensee also stated that "there was no general condition of i

i IGSAC identified in the rolled-region."

Licensee Facts 1 82.

8. However, Licensee has stated in other documents that tube ends behind the seal weld have IGSAC cracks near the HAZ and down the inside surface of the tube, approximately 3/8" from the top of the tube end.

Licensee Response to TMIA Interrogatory 33.

Further, Licensee could 4

only state that "at least 50% of the specimen tubes are not affected by i

i

. IGA in the HAZ," leaving open the possibility that close to 50% may indeed be affected.

Id.

9. Moreover, damage to the tube ends was visibly detected after kinetic expansion repair.

Licensee examinations showed cracks above the seal weld to be both axial and circumferential, so that in this area, where the cracks intersected, the kinetic expansion blew parts of the tube away.

TDR 008 at 16.

Further, ductility in those area appeared to be lost.

Id.

10. In addition, some cracks extended through the tubing behing the weld to the tubing below, id., so in situations where circumferential and axial cracks intersect in this area, the risk that chuncks of the tube could loosen and break away similarly exists. Id.

at 17.

This risk can not simply be ignored with a statement that the possibility is "unlikely."

Id.

11 If this should happen in a plugged tube, it seems the plug retention capability would be affected.

It is also_ unclear whether Licensee would then by relying _upon the seal weld in the HAZ as the primary pressure boundary. This is clearly a genuine material issue of fact needing further examination.

12. The fact of 23 leaking plugged tubes found after testing is claimed by the Staff to be " normal."

Staff Facts 1 5.

Yet the Staff's conclusion is devoid of any suporting data or evaluation to determine if in this highly unusual case, leaks may have been related to kinetic expansion repair.

13. Moreover, the Staff incorrectly states that the effect of kinetic expansion repair is irrelevant to the integrity of a plugged tube, since a plugged tube is no longer part of the pressure boundary.

Staff Facts 1 6.

Yet is is well recognized that plugging a tube will

. not arrest degradation, so if not stabilized, a severed plugged tube could indeed damage tubes surrounding it during operation.

12. Thus, there are several genuine material issues of fact concerning TMIA Contention 1.c i

e 1

,w_

. D.

TMIA CONTENTION 1.d

1. Contention 1.d addresses the credibility of the two reviews which attempted comprehensive evaluations of the safety questions raised by Licensee's fix for the TMI-l steam generator problem.

Neither the TPR nor the Staff in its SER conducted separate testing, but instead based evaluations primarily upon Licensee documents.

The Staff also used its own consultants, whose reports, along with the TPR

reports, became attachments to the SER.

See, SER at p.

3, 4 where the Staff describes the TPR as "an independent operation and safety evaluation." See also, Licensee Response to TMIA Interrogatories T-8, et seq.

2. According to the Staff, the SER's purpose is to evalute the specific repair method used by the Licensee and to evaluate subsequent operation using the repaired steam generators.

According to a presentation made by Licensee in October, 1982, the TPR was to determine compliance with NRC rules and regulations, and to determine the adequacy of the steam generator repairs that will allow safe operation of the nuclear unit. -2.

Yet dispite their virtually identical missions, there were safety significant differences in the reports' evaluations.

With regard to those safety issues raised by the TPR, the Staff made a specific finding that those TPR comments were "non-safety significant," SER at p.

4, providing no clear explanation why this was so.

3. And while both review groups ultimately recommended plant start-up, there were a substantial number of concerns raised initially by the TPR which do present clear safety significant questions.

At no time in this proceedings has Licensee provided first hand explanations

_37-from any TPR member as to why certain questions origi'nally raised by the TPR suddenly became non-issues.

(See, Licensee Response to TMIA Interrogatories T-1 et seq.)

4. TMIA did attempt in Interrogatories to discover the bases for TPR findings and conclusions, particularly ones subsequently reversed in later TPR documents.

Despite specific instructions to the contrary, Licensee itself, the subject of the TPR's critical review, responded to all questions asked of the TPR.

The conflict of interest problems are obvious, and Licensee's responses which " speak for" the TPR should be given little or no weight.

Further, these explanations appear to be grounded on questionable assumptions or facts.

See, infra.

5. Clearly, the TPR raised questions worthy of critical attention.

Yet despite the Staff's abolute responsibility to conduct a thorough and credible review of the safety questions raised by plant operation with these steam generators, the Staff fails to even discuss the TPR's points, or explain why certain TPR advice deserves to be arbitrariy rejected.

6. Material differences of opinion are material inconsistencies, and these differences between the TPR and the Staff findings and conclusions raise genuine material issues of fact.

7. Further, the response of TPR and the Staff clearly indicates that their knowledge of fracture mechanics is limited and superficial.

Fundamental misconception prevails throughout the (Licensee Motion for Summary Disposition, pp. 27-32).

What TPR and the Staff have assumed in their analyses are not consistent with the damage as they claimed to have observed in reality. Attachment 2, p.

1.

8. Finally, Licensee repeatedly states in its Motion for Summary Disposition that the analysis for crack resistance, i.e.

for the

-mechnical propagation of fatigue cracks in the tubes, was not part of, I

and is unrelated to, the evaluation of the kinetic expansion repair technLque.

See, e.g.,

Licensee's Motion for Summary Disposition at p.

29.

This is incorrect.

One of the means of evaluating the adequcy of i

the expansion repair technique is in fact to analyze the propagation of fatigue cracks in the tubes.

Indeed, paragraph 3 on p. 29 confirms this and yet the statement referred to above argues against it.

Attention should be focused on the overall techdical aspect of the problem and not whether TPR and the Staff happen to discuss a particular aspect of the tube repair technique., p.

2.

Material Inconsistencies i

9. The TPR concludes that the contingency of multiple tube ruptures needs to be examined. TPR 2/18/83 at p.

4.

Licensee asserts

- that "these' comments were intended merely to flag to the reader that the conclusions drawn were incomplete at that time since Licensee had not completed its. analytical or planning efforts."

Licensee Response to TMIA Interrogatory T-8.

Three months later, the TPR, for unexplained reasons, withdrew this as an open issue needing resolution i

before plant start up.

10..It is clear that during all qualification testing done by Licensee in 1982, the contingency of multiple tube ruptures, which including ruptures in both steam generators, was never treated as a subject warranting special testing.

And in fact, Licensee later asserts with regard to the simultaneous rupture case, that since this is "not a design basis accident for any plant, neither the TPR nor the g

en-,

,,c,+---..,m,w-.

,.r---

.-,,,.em.,,,,,,w,.r_

,,,-<,-,.,---n.,---,w.,..,-

v,,,

~

. Staff have required analysis for such an event in their respective approval for returning the plant to service."

Motion at p.

26.

11 The TPR recommends that because the particular cracks at TMI-l are stress corrosoin cracks, tubes with ID indications <40% throughwall be plugged.

TPR 2/18/83 at p. 6.

Three months later, the TPR withdrew this recommendation, indicating it felt assured that cracks of critical size will be detected in enough time to avoid breakage.

The Staff seems to go along with this.

Yet as has already been discussed, supra, Licensee's criteria and method for determining critical crack size is a major open question.

Further, the TPR's original point was that the stress corrosion nature of the cracks warranted this precaution.

This point is simply dropped by the TPR, and not addressed at all in the SER.

12 The TPR recommends that tubes within three rows of the lane region and the wedge-shaped region at the periphery which have OD indications at the 15th support plant or higher, be plugged as has been done in other plants.

TPR 2/18/83 at p.

6.

By May, the TPR withdrew this on this recommendation insisting it is sufficient that GPU only plug the first row.

The SER does not discuss the issue.

But the original point was that this is standard in all steam generators.

There is no explanation from the TPR or the Staff why the TMI-l steam generators should be treated differently.

13 The TPR suggests the possibility of small cracks in the rest of the RCS, which are difficult to detect. TPR 2/18/83 p.

10.

Three months later, the TPR points out indeed, attacks were found in the waste gas system and the pressurizer.

Thus, while the TPR's concerns are certainly still valid, it makes a specific recommendation against

. more inspections.

No explanation is provided, and the SER does not discuss this.

14 With regard to its conclusion that defects below certain size range will not propagate, the TPR discusses the large extrapolation of data still required.

TPR 2/18/83 at p.

10.

By May, the TPR still feels extrapolation is necessary, and still suggests long-term corossion tests to simulate flow induced vibrations.

It is of obvious concern to the TPR.

The Staff simply ignores this recommendation.

15 In discussing stress levels in the steam generators, the TPR asked if they "can be expected to be significantly higher, or the strength of the tubes significantly lower, than those in a normal OTSG7" TPR 2/18/83 p.

14-18.

It concludes, "from the information received, the answer appears to be negative, with a minor qualification regarding the undetected defects that are left in the tubes."

The TPR also states that small defects could leave tubes in a weak condition.

16. The TPR's reasons that the kinetic expansion repair of the tubes could affect the stress levels, if the process would change the srength or dimensions of the tubes, and states further that the kinetic expansion repair process is not expected to affect significantly the stress levels in the tubes.

The question of whether the strength or dimensions of the tubes were changed, to what degree they were changed, the significance of that change, and the question of how stress levels were changed, are simply not addressed by the TPR, or the SER, leaving a major, genuine material issue of fact unresolved.

Other Analyses

~

it1-

17. Licensce's Motion for Sbmmary Disposition at p.

27, line.two from the bottom, establishes the fa6e that the cracks in the tube do not propagate axisymmetrically See also Licensee Facts 1 99-100.

This implies that a three-dimensional state of stress prevails"around the surface flaw.

The nonuniformiLi'es referred-to in the second

~

paragraph on p.

28 and Licensee Facts 1 100 are not clearly defined and do not necessarily include ths three-demensional effects., p.

1

18. The fact that TPR and the Staff did_not.use the results of the

~

axisymmetric stress analysis for the fracture mechanics fatigue or crack analysis is irrelevant.

What is relevant is the non-axisymmetry character of the local stress field that should be included in a realistic evaluation of the crack failure mode., p.

1.

19. Unless the circumferential cracks are completely around the tube, axi-symmetry is not preserved and the stress state is locally a three-dimensional one.

This implies that even if the load is normal to the crack plane, the crack can grow in a non self-similar manner.

In such a case, the analyses performed by TPR and the Staff are not valid.

See, Licensee's Motion for Summary Disposition at p.

28, line three from the bottom. Attachment 2, p.l.

20. In Licensee's Motion for Summary Disposition at p.

29, lines 2 to 4 on top, the statement,

...the data were developed to characterize the material properties of Inconcel-600, and are independent of material or loading geometry", is self-contradictory.

It is well-known that all data collected from specimen tests are sensitive to changes in specimen size and loading rates.

Therefore, it is necessary to i

simulate-the conditions experienced in the structural components when

. collecting material data.

No justification along this line has been given by TPR and the Staff on Inconel-600., p.2

21. If the increase in hardness during repair is claimed to be beneficial, then the simultanious decrease in fracture toughness should also be pointed out and evaluated accordingly.

See Licensee's Motion for summary Disposition at p.

29; Licensee Facts 1 101.

This relation was not discussed.

In this respect, the so-called " toughness" factored into the fatigue model used by the TPR and the Staff may not be valid, particularly when yielding occurs as implied on p.

30.

Attachment, p.

2.

22. The concept of stress intensity depends on homogeneity of the crack tip stress field which prevails only when the material is predominantly in the linear elastic range.

See Licensee's Motion for Summary Disposition at p.

30. top; Licensee Facts 1 102.

When yielding or plastic flow occurs, the local stress field becomes non-homogeneous and the concept of a stress intensity ceases to apply.

Therefore, the argument outlined on top of p.

30 is irrelvant., p.3.

23. A fundamental misconception appears in line 10 on p.

30.

SUBJECT:

staff EVALUATI0tl 0F TM1 #1 STEAM GEf1ERATOR CORROSI0ti PROBl.EM

'We have determined that. the f.ubject problun t..:i n r t i t u i. 5. a.

u.re/ir.w !

La fe ty que" ti on a nd rea:miend t ha t f orma l t a l : n. view be r.,;uir"J Our detennination i ?. iia:.ed on four pr ima ry f a. ' ore.,.c.

'"Ilics; 1.

Uniqueness and e.t ent of the S. G. corroe.iun danuge.

1 2.

Potential for thi2 type of corrosion to af f ect other prisn y pres,ure boundary ma t erials.

3.

Uniqueness of the repiir method which is propos.d by GPUtic.

4.

Unpredictabili ty of ECT in detecting and quantifying t his type of corrosion.

In all likelihood the r.PilflC prmiram will an'.wcr the qm"u i.m which are nece*.sary to en.ure that a significant safety -ha7ard does not exi r.t.

Enclosed is a more detailed discussion of our ratirinale in support.

of this ret maandation.

.y r.

\\u'..( l.n'.,_ o.s ' ',

William V. J,hne.t on, A.. is t ar.t Director ibterials 5 Qualifications Engineuring Divie. inn of En<tineerin :

Cent.i t:

t.. & Cr.nien j

Iyt. 2Hb95 CC-See lieyt p., g o

,mv r-

-e

.rr-

,+,

-,,,,,,y.---.e,

,,,,w-.----.---w,,

--~-----,-,,---,~.,e--.

+:

s

, m........

Pa ti.....

The po.na1e far Deternination..

o.

of

."... c. a.- ;.. : -

. r t.. a...: c-l-at..=...;.

.2 tential for TF.I fio.1 S.G.

~. - - -.

..:. m, :

'~ c :c.

2 Corroston Prob.lems to. Constitute an

~

Un. revie..<cd. Safety Question

~.....'.,.,.-(:..... -

f..,...t,,

,s

.,,(.

.w., ;

s.c -... :

s.9

s......s.n.. >.... v:Fe,p...:....alf by mI-1. ~ i.o~ resol've th'etr st'cas generator pro

. ~

'The prop 53ed.

appears to be reasonable.

B1 sic lly, they are in the proecss of identifying:,

1.

The c(Lent of degradation in the 5.G. 's.

2., The extent,.of 'degradatinn[~:if any, in the remaindee of the reactor l

.;. c.co ant system,.s...... ~.,i..

. s.

3 The causative agent (s) and their source,

4...

,...;....e..

,4.. Cleanup techniques to remove the causative agent (s) and

..e.

t 5.

Optimma S.G. repair techniques.

Ikr.tever, we 1"iieve ti" re are a number of issue <. i coant irq t he pr ogr :r.i which.hauld he fot mally i cviewed by the staf f.

f:ar r 4.cning in 6 i idi 0 that Iomal s;at f review i: required is based on three major tattors.

A.)

To the extent that we have not experienced this type of t chavior

efore, the corrosion mechanism is unique, thus t ho r.taff has not reviewed the potential cons. quences of additional operatiens sub-sequent. to repair of (num.de fects. Partic'ularly, the patential for thi", type'of corrusica to rapidly progress upon r: start ani adversely affect the S.b. prit,as y pre. : ore Laundary.

' 3 ). The potential far thi; type of corrosion to adversely a f fect other prim.ny pressure boundtry materials.

l C)

The proposed S.G. tube repair technique, alk. hough having a similarity

~to nor..e part rcp.iir inchni<;ues is in itself unique.

l',.'e con <.ider the existance of a type of corrosion which har exte.sively degraded l the steam genri itors, to also have the potential to degarde et ber rnacter I coolant syste;a materials.

In addition, the liconsee praposes to employ a

unique repair *.chnique.

'le helieve that the cc::.bination - f all these tecnr

'an un. viewed safety que<. tion.

/r..t.itod at the beqinnirig, the p oqrau whit h GPt:NC is ianct :t i:.J c ppears Fa on.d.1o.

In aiI ) iL1 d. n,d. t he hiIIE pregra:a ai!I an :u i ih

.;:enLicns that a signit icant sai c* y h ic.o d &cs not 5thith.uo n m...ey io i r.u: e i

' exist.

I

.e f.'ve tha fol!v.riou.<

.-.uM/pr om.cros :

P. is ed 01 t.or i i..il e.!ge t o d it e.

l.

I ddy c u r ron t t.e. t i n.;u h.. -r.. in tha plant ta hnit..I

..f' itions end in ti.2 /.M'J cod" ie by the O. :s ir.adm u

...ure that

+-

p ci m ;. d, !..

the n' ur' J the i

.: ::.1 anina'ul ire.p

! u. will 1 I

,. _ _... ~, _,

m.._.,__._...,.~,_,_

.y.

. ~.

L :.: :..:.: : :.. x........:a :. :::= ~= =--

=. =. = -

' ~; * -

-?.. :.-. w : n.: m

- r~.=::s::.. ;.w:: : :y

.*==.=w

r-2-

P

2.. c. :gz.:.;;,

m-

.,.c* u ;.

., g.. ; e.. -

.. . -. +..x.

r. c.

clirrosionTechanist:ilht'ch' exists' at THI-l.

The licen'see's inspection program including' adequacy. " 'the scope of the inspection should.be evalu the staff for its

~~

.. f~ -

...... r....n :e.,.

f.'. Licensee is prop;.....

~

osing to employ an alternative repair technique to plugging"which'.is the ~ required tube repair raethod and specified in the plant Technical Specification.

....7..,,.

Thus, some'in6dification of the plant Tech. Specs. would be necessary.

.,,.y...:

O.

The proposed repair technique involves a leak l'miting rather than a leak free seal against primary to secondary leakage.

Because the expansion joint seal will function as the primary pressure boundary

, for as many as 20,000 tubes, leakage characteristics under nomal and postulated accident conditions should be established by test.

Testing should includa c.:pansion into dirty crevices.

The staff hae, e.ome questions pertaining to the proposed repair For example, will pre-existing cracks on the repair proi e.s it se'.f result in.' significant relaxation of tube prelvad?,1f so, excessive compressive loadini, may result upon bestup of the plant which could lead to bowing or local hockling which could cau'se new corrosion initiation sites.

The corrosion mechanism is unique, apparently very fast acting, and not well understood.

The licensee's recovery program should be closely reviewed by the staff to establish that there is adequate assurance against rapid failures occuring upon plant In addition, some licensing actions may be necessary, restart.

such as (a) more restrictive limits on primary to..econdary leak.

(Note the current 1 gpm limit. is the most liberal in the PW industry) and (b) fre the restart program, quent shutdowns for inspection as par' cf s

onsidering the above listed concerns, we believe that specific staff eview and concurrence is required at least in the following areas:

Review of ECT data and scope of the inspections perfonned to determine that indications outside the tubesheet have been aderpiately characterized and addressed in the repair program.

Review of ECT data and basis for rolling / plugging various tubes.

Including an a sessment of tube relaxation due to cracking or the repair technique (if tubes have been relaxed from +ension due to cracking, excessive compressive loads may ex.ist on restart).

~s 4

. '.. ~ ?' ::.;

..... g...... _.;,

_3_

.....y

,,, ~..

^-

.. ~ -

..-. v. L....

..-~.. ; u.: :-.. - -t.

. :. :~...

6eview of the'~ proposed roll technique, including the supnerting analytical and test verification program -.. : '

l

~

~'

Review l,.T'.:..:..'.'.N..':lfor and materials se'Iccted for primary side

. ~ *.

w....'

.. ~~ -

.. f the basi ;

o exa aination to detect the presence on corr osion on other pressure

'boundry.'ma terial s.

Also, an evaluation of t.he camination

techniques to determine the presence of corrosicn in these materials.

The staff should be present during examination of some pre-selected j

primary system materials.

Review of test data which supports the method selected for suifur removal fro'm system surfaces or conversely, the data which d.:monstrate that removal is not necessary.

Review of the current Tech. Spec. limit (1.0' gpm) for primary toi secondary leakage to determine the impact of operation with up to that volume of leakage and whether it adcouately supports the leak before break objective.

. F.eview of the restart program to ensure 'that' suf ficient check points are included to deter:nine that excersive primary pressure t:oundary At least degradation does not occur during subsequent operations.

the following general type of program would seem to be prudent.

a)

Perform a series of leak' checks ut.ilit.ing nitrogen, helium etc. prior to pressurization.

b)

Conduct a hydro; tat.ic test c)

Per' form a full temp. and press, iiot functional for two to three

~i weeks.

Then, shutdown and FCT a selected number of tuber. to ensure that excessive degradation is not occutring.

d)

Operate for 30 to 60 days.

Then, shutdos.n and ECT to assess the progression of degradation.

j c)

Assuming no excessive degradation in "d" above, operate for 150 to 210 days.

Then ECT again.

j f)

During refueling, ECT and examine additional. primary system materials for evidence of corrosion.

If no nc.; or ev.cessive con o:. ion i.. found, return to noi mal i rg. gu ide t est frequancies.

1 e

I I

i i

I

.. t.

- r-r r r m -.'..

George C. Sih, Director e c <5 '

LEHIGIl UNIVERSITY

+t p (( {1,[hh '

Institute of fracture and Solid Mechanics 7

rackard Lab. Bldg. #19 g

BETHLEHEM. PENNSYLVANIA 18015

~,

gl$ '

3 fn M

3y

, TU TMI Alert J

"'r A

315 Peffer Street

<1 Q ~,

,~

gA Harrisbura, Pennsylvania 17102

]

W ATTENTION:

Joanne Doroshow cc :

w.'

1-EPECIAL

'~

c m-r R t ' ' F 'm'

-- _= _ _

o

$%qc.hmed d COMMENTS ON RESPONSE BY TPR AND THE STAFF:

REF:

pp. 27-32 The response of TPR and the Staff clearly indicates that their knowledge of fracture mechanics is limited and superficial.

Fundamental misconception pre-vails throughout the above reference material. What TPR and the Staff have as-sumed in their analyses are not consistent with the damage as they claimed to have observed in reality.

(1) Reference to Comments on p. 27:

Line two from bottom establishes the fact that the cracks in the tube do not propagate axisymmetrically.

This implies that a three-dimensional state of st ess prevails around the surface flaw.

The nonuniformities referred to in the second paragraph on p. 28 are not clearly defined and do not necessarily include the three-dimensional effects.

(2) Reference to Line 3 from bottom of p. 28:

Unless the circumferential cracks are completely around the tube, axi-symmetry is not preserved and the stress state is locally a three-dimensional one.

This implies that even if the load is normal to the crack plane, the crack can grow in a non self-similar manner.

In such a case, the analyses performed by TPR and the Staff are not valid.

The fact that TPR and the Staff did not use the results of the axisymmetric stress analysis for the fracture mechanics fatigue or crack analysis is irrele-

. hat is relevant is the non-axisymmetry character of the local stress field W

vant.

that should be included in a realistic evaluation of the crack failure mode.

- _ =. _ ~.-.-

l (3) Reference to Lines 2 to 4 on top of p. 29:

The statement (quote) "--- the data were developed to characterize the ma-terial properties of Inconel-600, and are independent of material or loading geometry", is self-contradictory.

It is well-known that all data collected from specimen tests are sensitive to changes in specimen size and loading rates.

Therefore, it is necessary to simulate the conditions experienced in the struc-tural components when collecting material data.

No justification along this line has been given by TPR and the Staff on Inconel-600.

(4) Reference to middle of p. 29:

One of the means of evaluating the adequacy of the expansion repair technique is in fact to analyze the propagation of fatigue cracks in the tubes.

If the increase in hardness during repair is claimed to be beneficial, then the simul-taneous decrease in fracture toughness should also be pointed out and evaluated accordingly.

This relation was not discussed.

In this respect, the so-called

" toughness" factored into the fatigue model used by TPR and the Staff may not be valid, particularly when yielding occurs as implied on p. 30.

(S) Reference to Material on top of p. 30:

The concept of stress intensity depends on homoaeneity of the crack tip stress field which prevails only when the material is predominantly in the linear elastic range. When yielding or plastic flow occurs, the local stress field becomes non-homogeneous and the concept of a stress intensity ceases to apply.

Therefore,

Indeed, paragraph 3 on p. 29 confirms this and yet the statement above argues against it.

Attention should be focused on the overall technical aspect of the problem and-not whether TPR and the staff happen to discuss a particular aspect of the tube repair technique.

the argument outlined on top of p. 30 is irrelevant.

A fundamental misconception appears in line 10 on p. 30.

The fracture toughness of a material does not change when yielding occurs.

The load carrying capacity of the specimen or structural component on the other hand does increase.

(6) Reference to middle of p. 30:

No claims were ever made that hardness was directly associated with crack growth.

Nevertheless, it is well-known that increase in hardness results in a reduction *in toughness.

Hence, the repair tubes with increased hardness obviously

~

suffer a reduction in toughness and hence are no longer restored to their origi-nal state.

(7) Reference to Material on top of p. 31:

The propagation of small and/or large cracks in a thermal environment is im-portant to the kinetic expansion repair technique since the restoration of the system to its original state is at issue.

Hence, it would be relevant to estab-lish the life discrepancy of the repaired tubes as compared with those used in the original design.

(8) Reference to Lines 11 and 12 on p. 31:

The stress intensity approach used by TPR and the Staff cannot accurately determine the state of affairs for partial through-wall cracks.

In fact, the stress intensity factors as defined in the linear elastic fracture mechanics i

f 1 i

i theory are zero at the intersection of the crack and free surface.

However, daniage does occur near the surface.

(9) Reference to Material at bottom of p. 31 and top of p. 32:

TPR and the Staff have apparently failed to understand that the stress inten-sity, when it can be applied, may not be a monotonic function of the crack length under thermal environments.

Global instability can occur for cracks that are much smaller than those estimated by approximated and invalid analyses. The point is that the true nature of the thermal crack behavior may.not be reflected by the analysis made by TPR and the Staff.

To conclude, the relevant issue is the validity of the technical evaluation made by TPR and the Staff on the repaired tubes.

The problem must be viewed in its entirety based on consistency and validity of the technical approach Disposal of the issue on legal grounds at the sacrifice of technical understanding and safe control can further lead to unexpected shutdowns and further repairs, imposing additional burden on an already economically-illed nuclear industry.

It is not justifiable to claim as the state-of-the-art which is irrelevant to the safe evaluation of nuclear reactor components.

Other means and approaches for evaluating the damage caused by surface flaw have been available in the open lit-erature for many years.

- The fracture mechanics discipline is not limited to the views and definitions as conceived by TPR and the Staff.

-. -,.. -. - -. -_.~._.-.. ~. -. _.. -..__-.- -

I 4

i l

]

4 i

r I

t i

,i 5

i I,

I 1

(

s

(

,,,m.-n,--,--._,..--,..--,-n.,n.,---,,...,-...--,,,,-,,,,

b M C.h m e m t ]

A' MEETING REPORT Place:

B&W Plant, Lynchburg, VA Date:

Sep tembe r 20, 1982 Present:

John Pe arson )

Alvin McKim )

Bob Lite

)

Babcock & Wilcox Tracy Saville)

John Phillips)

Mary Jane Graham -CPU Nuclear Vincent Luk )

rRC Charles Davey)

Ihe purpose of this meeting was to discuss the remainder of the items re' quested by,. EEC to cocolete its review and evaluation of the TMI-l OTSG repair process.

2hes?e 8 ite=s (see attached list) were drawn up at the conclusion of the Sept. 15 ecting at SRC at which time the for=al RAI sub=itted by FRC was reviewed, item b y,, i t e m, to identify the still outst rnding information.

In,thfs report, a sumn ry o f t he discussions at Lynchburg I, given, numbe red accor. ding to the attached agenda.

1.

hechanical Dravings w.::

We reviewed the mechanical drawings of tube bundle and tube sheet assembly

~ tog,e ther with tube suppc rt system.

Bob Lite of E5W identified the locatiens of the stes generator where 35W conducted strain gauge measurements at Mt.

Ve rnon.

The two locations are ec follows:

one at the j unction be tween the islet header and the tubesheet, and the other at the veld location underneath the t ube shee t.

The st rain gauge measurements were t aken at the two ends of a diametral row of 132 tubes expanded at the same time.

S5W planned to decunent these findings and to calculate peak stresres and stress intensity for f atigue evaluat ion.

Their p reliminat,c results indi. ated that the usage

.facter at these two locations is less than 0.1.

We req ue s te d S &W Drawin gs

1'31132E and 131112E which describe the structural conifruration at these tgo locations.

The se two d rawings we re promised by Sep te:Se r 24, 1932.

2.

Dyy.anic S tress-S train Cu rves These curves are not available.

Since CPU, B&W and FW are taking an experi-mental approach, they claim there is no need for dynamic data.

3.

Coe f f.i_cien ts o f Exp ansion Handback values are used.

Data supplied to Dr. Luk.

4.

Table of Contents of Report.

GPU vill not release this TOC because of its "c ritical" dature.

FRC will eventually receive a copy of the final report, due October Iste N1 g

wp- - -.. r.

t

.+

5.

Supporting Data for Pressure Vessel Integrity Asse u mnt.

Sete of the statements presented were objectives which require supporting test data.

There is so:e preliminary dats and scre additional undocumented data which will appear in the October 1 report.

We have in hand sem of the preliminary data.

6.

Dw.amie Pressure Profile in Tubes once more, claim was that there is no need for these data.

All work is experimental.

Profileretric data has apparently been done and will be available (to FRC about 10 Oct. 1982).

7.

Critical Data From Mt. Ve rnon.

Pull tests were performd at Mt. Vernon on three expanded tubes to a limit of 36000 lb, in order to preserve the integrity of the generator.

There was no movement in response to loads of this magnitude.

ne forthcoming (Oct. 1) report vill contain all currently availabic data

.,from Penn State, Foster Wheeler and B&W.

3 Ddtails of Strain Gau;;e Instre$entation.

The gauge used was a ros.e t te, Micro Measurecents Ccperation.

It was a film g.mge, two elen.ents in quadrature and one at 45.

Cauge factor was 2.

.The signal conditioner was a Dickey P22.

Output was 1000 microstrain/ volt.

All signals were ni-multiplexed on one tape track.

Frequency response was

". t o 8 KH z.

'*e were shewn signals which ccvered frem -500 s throt.sh event to 3.5 secends.

Overall coverage is I. se con ds.

"'he strain gauge en the tube (axial) gave a ve'ry -small strain sig.al (less than 100 microst rains).

Analy sis of the spec tral ecntent of the strain reeerds sh:ved a level of f at shut 200 Hz.

Little er no signal contribution was.ade at frequences grc. iter than 400 Hz.

V. I.uk C. T. *: 2vey 9/21/82 Om e

j O

O O

w~

o

- ~

y

' y j

4

/

ITD'.S PlQUESTED AT LYNCHEGC MEETING

1) &chanical dravings of tube bundle and tube sheet assenbly together with tube support systen.

2) Dynamic stress-strain curves for tdbe sheet and tubes.

3)

Coefficients of expansion for tube sheet and tubes as a function of te=perature.

4) Table of contents of October qualification program report and critical data now available.

5)

Supporting data for Pressure vessel Integrity Assess:ent (p res ented at 9/15 meting at Bethesda).

. 6). Dynamic pressure profile in tubes during e.xplcsion (requested by NRC).

s%

7)

Critical data fr:m Mr. Vernon tests aircady run.

,8). Details cf inst:unentation used for tests.

Specifically, strain gauge type, signal conditioning equipnent, load configuration, recording methods, calibration techniques.

\\

c.'

\\

t_.

e F

=

e 9

4, O D e

e e

=

" ERD

h h rucbb STATEMENT TO HEARING ON TMI RESTART: STEAM GENERATOR TUBE REPAIR December 16, 1983 e

With regard to the technical aspects of steam generator tube repair, there arises the following concern:

(1) It is questionable that the kinetic expansion repair process could restore the tubes to their original condition.

Should the hardness of the tube be increased then the resistance of the material to cracking measured by the fracture toughness is likely to drop.

(2) The fatigue life of tubes depends on the secuence and magnitude of the combined thermal and mechanical loading. Predictions based on LEFM* (linear elas. tic fracture mechanics) may not be reliable since the methodology does not address accumulative damage.

In my opinion, there is not suffucient technical evidence to reach any conclusion on the safety of the repaired steam generator tubes of TMI.

il-G. C.

Sih I

t i

' Application of LEFM cannot be justified on the ground that it is still the state-of-the art of twenty years ago.

_,. ~,, _ _. _. _. _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _. _ _

1 l

l I

i f

f i

I a.a,,.,

.,,,,,..a.,,

n....

r.v,-,, -.-,

g,... - - -,.:.,.

.c

hk'drnerd 8 TMI-l Steam Generator Tube /Tubesheet Repair Meeting at Parsipanay, N.J.

1.

BACKGROUND Once Through Steam Generator (OTSG) tube leaks have been found at li1I-l.

Failure analysis has resulted in the conclusion that sulfur co'ntamination caused intergranular attack of the TMI-l OTSG tubes.

The atta d '- inside diamater initiated and circumferential in geometry.

The me e the cracking has been located by Eddy Current Testing (ECT) it. 6-s u r end s

of the Inconel 600 tube at or near the weld heat affected zone (HAZ) and the roll transition. Since the failure analysis results continue to indicate that the areas of tubes free of ECT indications can be successfully

' refurne,d to service, a repair process has been selected which kinetically

.~

expands the existing tube against the upper tubesheet hole.

The objective is to expand the OTSG tube with sufficient length below any defects to form a new load carrying and essentially leak tight joint.

2.

E SI TUBE REPAIR 2.~ 1 Sequence of Repair Steps The 'Till-l Ol'SG repair process is as follows.

It should be recognized that Step. 2 may become a part of Step 6 pending the outcome of preliminary l

qualification testing.

l Step Descriotion l

1..

Flush the secondary side tube-to-tube sheet crevice 2.

Clean tube inside diameter in the area of the repair.

3.

Heat crevice to drive out moisture (vaporize water) l 4.~

Maingain crevice in the repair area at a temperature at least 10 F higher than the dewpoint for OTSG secondary side conditions.

i l

S.

Kinetically expand tubes.

6.

Cleanup debris from kinetic expansion.

l 7.

Leak test OTSG 8.

Roll / flush (if required to repair leaking tubes.)

l l

e 4

- g ---

.p..

y-ei-7 y-p w

INFCRv.AL TSC!-21ICAL COMMUNICATION

~

Date k<}$- T, l V 5 To: J'A/ /2A 12 A>

//CR J'Acog.S Spook:./h/?e zgp; From: (EO I

U.S. Nuclear Regulatory Commission FranklinFasearc/

h Center Washington, D.

C.

20555 Philadelphia, PA 19103 (10 BE OPESED BY ADDRESSEE ONLY)

P4ference:

NRC Contract NRC-03-81-130 FRC Project C5506 4 N#8NMT /O NRC TAC No.

FRC Generic Topic 7 M / -/

Plant Th / - /

FRC Task (s) 3 //, _7/2; ?/1 f4, g C,

./

Title:

- ' Attac'hnen:::.

TR /P erpoGr of 3'vL v 2/ Mctmiz Ar

! I') (I)I'.-X re d,

Al T Messaget TC(~UMC / L CD P H C/sTS A /2[ 6/vtAJ,

8A$m OL/

b l5 CM 71 /reA) tT-lurtV -

SUMMk2V OC M CYT rAfS-0[-

~~nLT f) t//> >c C~

Lud AuD 7tPi M o e/2r of" 6fu

/S tU C LLI or'ti,

f*HL:~ Ell' /S S774L Ce A/CC*AM A2>Jr /2rCC,J e A/f~

r iIJ!;;c c a s r,.J UC H C~U O 0C72 C M CJr?-

Ar~W TW C

[t l 1-Un;i set)

/r/J D i

Copy of =essage form only:

NOTE TO SENDER:

Include f

attachment.5 if inforr2. tion is pertinent to program management..

S.

S.- Sa pa NRC Performance Monitor M 6E)@T 8/!A/mL'//

NRC Lead Project Ma nace r

( C///?/ B

>f e C /2ACrft) t~dC Distr ibution:

SPC, SP, Q, /L, b, M,

~,

Revised 4/10/82

~

-g3 f

~_

TRI? REPORT:

PROJECT C5506, Assignment 10.

PLACE VISITED:

Foster Wheeler, Livingston, NJ DATE:

July 21,1982 Th0SE.PRESEST:

Representatives from:

SRC FRC GFl!

EV B&W Present from FRC:

L. Leonard C. T. Davey V. Luk T. A. Shook PCRP'SE OF MEETING:

To present. schedule for qualification program of tube repair, and result of preltminary test < perforrea.

Interf ace with technical pe rsonnel on the.,tatus of 9

analytical work.

1.

Inis meeting revealed newer, more up to date plans than previous meetings.

It allowed more direct interactions among technical pe rsonnel present.

e t><

2.

, Pregress was annoetced on the Ordnance Cord-booster ecncept for car.dle Initiation.

This initiation process should result in less explosive reaction cutside the expansion space, and thereby reduce cen tamination

. and potential damage as a result of the m re brisant primacord initiation

' precess that the Ordnance Cord replaced.

Furthe r studies are planned to_

con f i rn that Ordnance Cord tu appropriate.

1 Dr. Pai (Fester Wheeler) expresses belief that the bees te r will have no mo re ' e xp _'o sive e f fe ct than an equivalent length of primacord.

3.

The schedule for delive ry of the six single-tube and two ten-tube test specicens was discussed.

Delivery was to oe in late Septembe r to r '.C.

hoceve r, mo re re cen t information indicates that FRC should have all

- fixtures and hardware by mid-August.

Foster-Wheeler has partially completed their apparatus for cyclic testing of the ten-tube mock-ups.

The test parameters are slightly dif ferent than FRC's but at this time-it appears that results should ba ecmparable.

J4.

The pull-out test s showed that the load required to pull out an expanded tube (after the two step process) is essentially independent of the exter.ded length for lengths above about 5 inches.

Acco re!!n gly, Foste r Wheeler is qualitying a six inch length even.though the minimum expansion at this time will be 17 in.

For tubes which leak in the tube sheet after the 17" expansion, a follow-up 23" expansion can he utilized to seal a new 6" qualified length below the original expansion.

Peripheral tubes m,ay cause problems; the candle cannot be injected easily.

.p 5.

There were some sericus concerns if the expansion is perfore d too'close to the functure of the tube and the tube sheet, placing the transition near the inside surface of the tube sheet.

A double ended rupture was feared that would potentially cause serious leaks.

Particular cencern was expressed in the event of a steam line break during which ti=e a te=perature dif ference wculd exist in which the tube weuld be at a lever temperature than the tube sheet.

The accerpanying tube shrinkage would result in a tensile load on the tube.

This situ tion will be investigated by B&W.

6.

X-ray dif fraction residual stress measurements will be =ade at Penn State University to evaluate the relative magnitude of stresses induced by tube rolling and explosive expansion both in fully expanded areas and in the transition to the caexpanded region.

These tests will take about 2 to 2 me ths to cocplete, and, thus, all results will not be available until the beginning of October.

This should present no problee with regard to the i=plementation of repairs since it is highly likely that the expansion

, _, residual stresses will be less than those fro = rolling.

The tests are to

,, ' demonstrate how much more uniform' and less severe the stresses are resulting f rqm explosive expansion than ' from rolling.

'7.

Candles will be supplied to FRC by Foster Wheeler for any experiments we

, wish to run.

It was cenfirmed that FRC is fully licensed to store explosives

_(Federal, State and City licenses).

8. ; Multiple expansions are planned at Mount Vernon, Indiana on 5 August 1982.

e,.- 'This is a B&W " graveyard" for old steam generators.

FRC obse rvers will attend.

9.

Materials presented during this meeting will be rai'_ed to FRC by GPUN af ter appropriate approvals.

As of this writing these ma:erials have not been received.

11

.i. s t i n g u.e held beu.. :n V:n unt Luk of FRC and 'i-h re of GPU at GIC.

2he purpcse of the meeting was to discuss the Licensee's stress evaluation program.

At the meeting. FRC reviewed a draft CPU Stress Report, T3R No.

3'A, "T"l-1 GTSG.is-Built S t re ss analyses." The report focuses an the review and eva'...atien of the pe rformance of the as-buil t. st a pune rater, and it dres not address the ef fects of the kinetic expansien process en the performance of the Enit.

The report will probably be released to public f arain by 7/23/S2.

FRC cay get a ccpy of the report before the end of July 1982.

At the =ee ting, the scope of stress analyses which FRC would like to be covered by licensee was discussed in detail.

According to Jie Moore, test i tecs covered in FRC's scope will be included in licensee's test qualificatirn progran and the re-maining few itens will be evaluated analytically by either CPU or B&W.

All the results f rem qualification and stress evalu.itien pregrams will be reviewed b;. FRC when they beccce available.

11.

A request was cade to Mary Jane Graham for the f oi ; ~. in g doc e.en t s :

a.

Mechanical drawings of tube bundle and tubesheet-assembly toge ther with tube support system.

~

O,

b.

P rop riet ar,/ Topical Report, B&W-10002, a

"Once Through Stea= Generator Research and Develop ent Report."

c.

B&W-10ll.6, "Dete rmination o f Minitu=

Required Tube Wall Thickness for 177 F/A OTSG's, " Babcock & Wilcox Report."

en m

'W b

es. e O

e O

A:n.

e e

e 9

e

- eae e

e e

1 I

i I

I I

I t,.

(Y\\C h

.r.z...

t a

pn atcoq(o, UNITED STATES

!,}h, p,j, ///,

jfh NUCLEAR REGULATORY COMMISSION o ;g j

WASHmGTOd. D. C. 20555

%;...J s

g

.....f

- c. -

OCT 2 0 1982 I

Docket No. 50-289 FACILITY:

THREE MILE ISLAND, UNIT N0.1 (TMI-1)

LICENSEE:

GPU NUCLEAR CORPORATION (GPUN)

SUBJECT:

SUMMARY

OF MEETING WITH GPUN ON SEPTEMBER 15, 1982 CONCERNING GPUN'S STEAM GENERATOR (SG) REPAIR PROCESS

Background

I As part of their program to recover the SGs from intergranular stress corrosion cracking of the tubes, GPUN has proposed a repair program involving an explosive expansion technique to recover tubes with defects within the upper tubesheet (UTS).

The purpose of the September 15, 1982 meeting was to provide a final briefing to the staff prior to start of the actual repair and to resolve any remaining

~

staff or staff consultant concerns regarding the repair itself.

A copy of GPUN's presentation is enclosed.

A list of attendees is also enclosed.

Discussion GPUN's proposed repair process consists of kinetically or explosively expanding tubes within the UTS.

All 31,000 tubes will be expanded for 17 inches or 22 inches within the 24 inch UTS.

In order to

!~

establish a qualified seal, there must be a six inch area free of defects.

l Hence, a 17 inch expansion will recover tubes with defects only within i

the top 11 inches and a 22 inch expansion will recover tubes with defects only within the top 17 inches of the tube.

The process involves use of low level explosives including prima cord, booster, ordance transfer cord and blasting caps.

The prima cord and booster are inserted into a poly-ethylene " candle" and detonated by a blasting cap outside the OTSG via the ordance transfer cord.

GPUN will be ready to commence the expan-sions in mid October 1982.

Related actions involve secondary side flush, crevice drying, expansions, debris cleanup, plugging tubes' unable to be

. recovered and testing.

GPUN expects to con:plete these operations by December 1982.

The staff issued a Safety Evaluation limited to the steam generator repairs on October 13, 1982.

No staff members or staff con-sultants raised concerns that would postpone or prevent GPUN from commencing the repairs.

A meeting has been scheduled October 18 and 19, 1982 (previously October 13 and 14) to discuss remaining aspects of

TMI-1 GPUN's steam generator recovery program.

GL Ab 1

Richard H. Jacobs', Project Manager Operating Reactors Branch #4 Division of Licensing

Enclosures:

1.

List of Attendees 2.

GPUN's Presentation cc w/ enclosures:

See next page O

0 N C.h me d {p-I Process 3#ect On Cracks / Indications io Test Results (30 gr./ft)

- 100% through wall crack opened slightly

- No ductile growth axially

- No ductile growth circumferencially

Conclusion:

crack did not grow o Leakage Tends to be Self Sealing

- CR-3 operating experience

- ARC model boiler test results

~ 5 tubes defected /1 to 2 gph leakage.

- After 1000 hrs operation - no leakage

- CRUD & corrosion products seal leaks

- Tubesheet corrosion insignificant by inspection

R b A me+ 6-A Third Party Review Purpose To provide a timely, independent, objective, safety evaluation of all activities defined in (the scope of) this charter for conformance to:

1) the NRC rules and regulations governing the operation of TMI-1

2) the adequacy of the steam generator repair program that will allow safe operation of the nuclear unit Scope

Failure analysis program o Eddy current examination program

OTSG performance evaluation

Repair criteria

OTSG repair program 9/15/82

=

l t

I l

l I

~

/ p uuy}*o UNITED STATES NUCLEAR REGULATORY COMMISSION

,9

(

o

.{

q$

WASHINGTON, D. C. 20505 y (N k,, %,,, sp

'8 May 5,1983 GilAIRMAN g

T.he Honorable Edward J. Markey, Chainnan Subcommittee on Oversight and Investigations Connaittee on Interior and Insular Affairs United States House of Representatives Washington, D.C.

20515

Dear Mr. Chairman:

d This is in response to your letter of March 23, 1983 which raised questions resulting from my letter to you of March 21, 1983 and the Subcommittee on Oversight and Investigations' December 13, 1982 hearing.

Responses to the questions in your letter are enclosed.

Additionally, you reminded us of our promise made during the February 22, 1983 Energy and Environment Subcomnittee hearing, to provide information relative to the exact status of completion of items in the TMI Action Plan.

This information was provided to your staff by our Office of Congressional Affairs on March 25, 1983.

We hope this information resolves your outstanding questions with regard to these subjects.

Sincerely, d 6ct.-~-,-

Nunzio J. Palladino l

Enclosures:

Response to Questions cc:

Rep. Ron Marlenee i

t What analysis has the NRC or the ACRS done to eval TMI-1 the probability or consequences of this risk at QUESTION 3; and what were the reasons?

t has been performed RESPONSE:_

Such an assessment is No probabilistic risk assessment of the subject evensons set forth by either the NRC staff or the ACRS for TMI-1.

considered unnecessary by the NRC staff for rearesp start of TMI-1.

l l

.

ML20087N667

Text

SUMMARY

SUBJECT:

Title:

SUBJECT:

SUMMARY

Background

Enclosures:

Conclusion:

Dear Mr. Chairman:

Enclosures:

Navigation menu

Search