ML20112J822
| ML20112J822 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 01/10/1985 |
| From: | Giacobbe F GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| Shared Package | |
| ML20112J780 | List: |
| References | |
| OLA, NUDOCS 8501180383 | |
| Download: ML20112J822 (63) | |
Text
. ~ _ _ _ _
January 10, 1985 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION Before the Atomic Safety and Licensing Appeal Board In the Matter of )
)
METROPOLITAN EDISON COMPANY, ET AL.) Docket No. 50-289-OLA
) (Steam Generator Repair)
(Three Mile Island Nuclear Station,)
Unit No. 1) )
AFFIDAVIT OF F. SCOTT GIACOBBE E. SCOTT GIACOBBE,-being duly sworn according to law, de-poses and states as follows:
1.- I am Manager, Materials Engineering.and' Failure Anal-ysis for GPU Nuclear Corporation. A statement" of my qualifica-tions and experience is attached and incorporated herein by reference.
- 2. The purpose of my affidavit is to address the allega-tions of TMIA regarding the possibility of reinitiation of the intergranular stress assisted cracking (IGSAC) which took place on the inner surfaces of the steam generator tubes in 1981.
-Such reinitiation has not-occurred, and.neither of the bases cited by TMIA -- temporary increases in the concentrations of sulfates and chlorides in the primary system and recent eddy current indications -- indicate that it has recurred.
v
, 9
- 3. Attachments 1, 2, and 4 to TMIA's motion to reopen indicate an increase in sulfates in the primary coolant follow-ing hot steam generator testing in late 1983, and an increase in sulfates and chlorides following refilling of the primary coolant after the system had been drained. (Attachment 5 re-fers to' draining and refilling the steam generators, a second-
'ary side operation which has no bearing on the IGSAC.) Since we began monitoring the primary coolant water for extremely low levels of contaminants after the tube damage was discovered in 1981, we have found that major changes in the pH of the primary coolant, or draining and refilling the primary side of the steam generators, result in increases in certain chemical impu-rity concentrations. The increases are both temporary and of very small magnitude, on the order of 0.1 to 0.5 parts per mil-lion (ppm). Raising the pH through the addition of ammonium hydroxide can result in changes in sulfur solubility, which permits a temporary increase in the concentration of sulfur in the primary coolant. Draining leaves a film of water on the surface of the tubes, which in turn leaves a residue of impurities on the surfaces when the water dries. When the sys-tem is refilled, there is an observed increase in concentra-tions in the new water as the tube surfaces are washed. This quickly abates as the residues are dissolved and cleanup is in-itiated. In each case where sulfur concentrations have in-creased, the reactor coolant is purified immediately, the con-centrations are rapidly reduced, and no further contaminant spikes are observed.
- 4. Thus, levels of sulfur, and for that matter other contaminants, are expected to temporarily increase from time to time for a variety of reasons. As discussed in paragraphs 106 and 108-116 of my February 23, 1984 affidavit, for example, the observed increases in sulfur levels were anticipated. This is why chemistry specification limits are established and why re-actor coolant purification systems are part of normal plant systems. These are the methods by which all nuclear power plants control contaminants. Increases in contaminants can re-sult from a variety of sources such as contaminants in chemi-cals added to the system, impurities from the itmosphere, and dissolution of remaining sulfur from surface oxide films. Ir-recpective of the cause of contaminant buildup, it was recog-nized early by GPUN that this would happen and that it would need to be controlled. In addition, because we knew contaminants would be present, we established the long-term corrosion test with contaminants intentionally added to the test solutions to assure ourselves that our specification lim-its were adequate to prevent IGSAC.
6
- 5. The uncertainty as to the reasons for this increase, as expressed in TMIA's attachments, was that Licensee wanted to be certain that there were no unidentified sources of sulfur or chloride contamination and that there were no analytical errors in sulfur measurement. The investigation into these concerns did not uncover any external sources of sulfur contamination other than normal chemical impurities found in chemical
reagents. Some improvements in analytical techniques were made as a result of_this investigation.
- 6. In any event, the temporary spikes in sulfate concen tration could not cause reinitiation of the IGSAC. The in-creases were far too small, and other environmental factors precluded such attack.
7 Similarly, the temporary spikes of sulfate connentra-tion, whether or not due to the addition of ammonium hydroxide which is used to raise the pH of tne coolant when the reactor is in wet layup, have no bearing on the concern expressed by staff consultant Dillon, as alleged by TMIA on page 10 of its
- brief. Dillon's concern was that the peroxide cleaning pro-cess, which was completed in August, 1983, would put large quantitiet of sulfur (5-10 ppm) in solution at that time. His concern had nothing to do with the subsequent control proce-dures involving the addition of ammonium hydroxide, and he cer-tainly expressed no concern with the magnitude of temporary
?
sulfate' concentrations which we have seen. (As I noted at paragraphs 95-97 of my February 23, 1984 affidavit, the perox--
ide cleaning process generated no more than O.4 ppm of sulfur compounds, which was not corrosive.)
1
- 8. The observations of the small increases in contaminant concentrations have demonstrated that we are able to monitor for minor increases and to control' contamination levels through normal cleanup systems when they do occur. They also confirm that the dissolution of residual sulfur compountls L.
incorporated within the surface oxide films is not a problem, and that, as anticipated, sulfur levels are far below those
+necessary for the reinitiation of the IGSAC.
- 9. The recent eddy current indications reported in TMIA's Attachment 6 to its motion are also not indicative that IGSAC has been reinitiated. As described in detail in GPU Nu-clear Technical Data Report 638, January 11, 1985 (TDR 638, at-tached Eereto), we have performed an in-depth study to deter-
.mine the causes of the new indications, with particular
. emphasis on determining whether they indicate that IGSAC has been reinitiated. The investigation has shown that the degra-
.)
dation is not new, and can best be characterized as intergranular attack (IGA) which occurred in conjunction with the 1981 IGSAC.
- 10. Our investigation included an analysis of the envi-ronmental conditions that the steam generators experienced since the discovery of cracking in 1981. This entailed a re-view of plant
- operation and chemistry records and a comparison of the conditions found to existing data on the behavior of Inconel'-600 under such conditions, sParameters such as pH of the. reactor coolant,scontamin' ant levels, oxygen levels, lithium levels and the water levels in the steam generators were re-
+
viewed. The conclusion of this evaluation was that at no time were the steam generators in a condition which would be consid-ered corrosive to the s' team generator tubing. See TDR 638,
'@ pp. 16-19; see also pp. 33-48.
- 11. All corrosion testing to date has confirmed our posi-tion that, by controlling chemistry, there would be no recur-rence of the IGSAC. Analysis of the eddy current indications, recent bubble tests performed, plus visual observations of the tubes via fiberscopic examination down the tube bore provide sufficient evidence to conclude that corrosion of the type pre-
'viously experienced is not continuing. This evidence in part is made up from the fact that at present there are no leaking tubes, that the current defects have very small circumferential extent, and that visually they appear to be rounded or eliptical, unlike the linear cracks observed before. These de-fects could be classified as intergranular pit-like defects and as such are much like the IGA island or pits which were ob-served during the previous failure analysis of the steam gener-ator tubing. See TDR 638, pp. 20-30; see also pp. 6-15.
.,12 . The current indications had previously gone undetected because of their small circumferential size and be-cause with ICA there is very little volume loss (i.e., loss of metal grains). Because eddy current sensitivity is highly de-pendent on defect volume, detection of IGA by eddy current is more difficult to detect than is IGSAC. If grains of metal in the IGA area should drop out, however, the volume loss from the defect would be significantly increased and the detectability increased. The thermally induced strains and hydraulic forces during the-hot functio'nal testing performed in 1983, subsequent to the record eddy current examinations in 1982,.were more than sufficient to cause grain dropout and grain boundary separation of the previously existing IGA, all of which would increase eddy current detectability. Such grain loss and grain boundary separation from IGA areas have been observed on previously re-moved tube samples, and is expected to continue for a period of time under the action of' thermal or mechanical strains to the tubing which occur during hot functional testing or operation.
- 13. This recent eddy current inspection, as well as fu-ture eddy current examinatione, coupled with leak rate moni-toring, will continue to assure that such defects are found and that the steam generator tube integrity will be maintained.
,.fpp fV
, w s <* r*
F.ASChTTGIACOBBE
/ .
Subscribed and sworn to before me this /d - day of January, 1985.
NOTARY PUBLIC My Commission Expires: 8/ [J'
Attachment to Affidavit of F. Scott Giacobbe STATEMENT OF QUALIFICATIONS AND EXPERIENCE I, F. Scott Giacobbe, am employed by General Public Utilities Nuclear Corporation as Manager, Materials Engineer-ing/ Failure Analysis. I have been in this position since July of 1982.
My education includes a Bachelor's Degree in Mechanical Engineering from Villanova University in 1970 and a Master's Degree in Materials Engineering from Drexel University in 1975.
My work experience has provided me many years of direct involvement in the materials evaluation and failure analysis of power plant components; early in my career it also provided a very intense involvement in heat exchanger tubing evaluations.
In 1970, I began my employment with Westinghouse Electric Corporation in their Heat Transfer Division as a Materials En-l gineer. In this position I worked on the materials selection, corrosion evaluations and failure analysis of heat exchanger components such as feedwater heaters, condensors, radioactive wasta evaporators and other secondary side heat exchangers. In i
particular, I was responsible for assuring that tubing utilized l
in the Westinghouse heat exchangers was properly specified and
- manufactured. This function provided me with in-depth knowl-edge of heat exchanger tubing fabrication practices, corrosion resistant properties and failure mechanisms.
l l
In 1977 I left Westinghouse to join General Public Utilities as a Senior Engineer in their metallurgical laborato-ry. This position afforded me the opportunity to expand my areas of expertise to include materials selection, corrosion evaluation and failure analysis of other components of both nu-clear and fossil power plants, and to gain a broador under- ,
standing of power plant operation.
In 1978 I was promoted to supervisor of the metallurgical laboratory. This was a first line supervising position which gave me the responsibility for the daily operation of the labo-ratory and supervision of the technicians and engineers re-porting to me. This position also carried with it a large technical responsibility which kept me heavily involved in the day-to-day materials engineering problems.
My career took on a slight change in dire'ction in 1980 when the company reorganized and formed the Nuclear Corpora-tion. At that time I became Materials and Welding Manager in the Nuclear Assurance Division. With this position I essen-tially had the same functions as before, with the added respon-sibility for welding at the nuclear power stations. While in this position I was responsible for the technical and metallur-gical aspects of the development of the Nuclear Corporation welding program. During this time I was still supervising all failure analysis activities, including the TMI spent fuel pool pipe cracking incident.
a
't In July 1982, another reorganization took place. At this time my section merged with the materials engineering section in the Technical Functions Division and I took over management-of that newly formed section. In this position I now had func-i tional responsibility for the raterials configuration control of both GPU nuclear power plants as well as welding engineering .
and failure analysis. In addition, my section still provided failure analysis services to the fossil companies.
I have been involved in the steam generator tube failure issue from the beginning. I participated.directly in the ini-tial decision-making regarding the tube sampling and removal operations and was present to perform the initial visual evalu-ations of the removed tubing. I personally planned and oversaw the failure analysis activities performed by the outside la-boratories. I also developed the corrosion testing programs which GPUN implemented to gain insight and understanding into the failure mechanism and responsible corredants. It was also my responsibility to coordinate the input from all our techni-cal consultants as well as plant experience and formulate the current failure scenario.
During the steam generator repair, my section also provid-
! ed materials evaluation and consultation on all aspects of the repair including explosive expansion, flushing, peroxide I
My section also developed and imple-
~
cleaning, and so forth.
mented the long term corrosion testing program and is 4
evaluating the results as the testing progresses.
4 d
Lastly, during the course of the steam generator repairs, I was responsible for making all presentations to the NRC on 4
corrosion testing and failure analysis activities.
Over the years I' have kept fully abreast with the state-of-the-art in corrosion technology through my attendance and participation in technical seminars and conferences, and ,
through attending training sessions. I am a member of the Edison Electric Institute Materials, Piping, Welding and Corro-sion Task Force, a group of industry representatives who meet to share and develop solutions to corrosion problems in the field of materials and welding in the power industry. In addi-tion, I am a member of the American Society for Metals.
Publications
- 1. F. S. Giacobbe, " Examination, Evaluation and Repair of Stress Corrosion Cracking in a PWR Borated Water Piping System", NACE Corrosion 81.
- 2. F. 5. Giacobbe, J.D. Jones, R. L. Long, D. G. Slear, "Re-pairs of TMI-l CTSG Tube Failures" Plant / operations Prog-ress AICHE, July 1983, Vol. 2, No. 3.
1
=__ . _ _ _ _ _ . _ _ . , _ _ _ _ _ _ _ _ . . . . . _ _ - _ _ . . _ . . . _._ _ _ ._..__ _ ____._.._,.___._.
3 GPU Nuclear Corporation UOEMf 100 Interpace Parkway Parsippany, New Jersey 07054-1149 (201)263-6500 TELEX 136-482 Writer's Direct Dial Number:
January '4, 1985 5211-85-2010 R W-0382 Mr. John F. Stolz, Chief Operating Reactors Branch No. 4 Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D. C. 20555
Dear Mr. Stolz:
Three Mile Island Nuclear Station Unit 1 (TMI-1)
Operating License No. OPR-50 Docket No. 50-289 Steam Generator Eddy Current Test Result Evaluation In accordance with the Technical Specifications for TMI-1, an eddy current examination of the steam generator tubes was condric.ted in November and December 1984. An initial report on the resu 1 examination was contained in LER-84-007, suomitted on Decembe m.
We have just completed a Technical Data Report ti. '8) entitled
" Evaluation of Eddy Current Indications Detected Du. ne 1984 Tech. Spec.
Inspection." This TOR supplements the information cons.ined in LER-84-007.
We are continuing our evaluation of the results of the examination and we will provide you any additional information that becomes available.
Sincerely,
, mT
. F. Wi son Director Technical Functions Ir/0537e cc: R. Conte H. Silver Dr. T. Murley C. McCracken GPU Nuclear Corporation is a subsidiary of General Public Utihttes Corporation
l j
TDR NO. 638 REVISION NO. O BUOGET TECHNICAL DATA REPORT ACTIVITY NO. 123125 PAGE 1 Op 49 PROJECT: & Design DEPARTMENT /SECTION h / q Nee Ant 1.
WgaM =1 eEn ineerina )
ntI-l OTSG. REPAIRS -
1 RELEASE DATg 1/11/85 mgVISION DATE OOCUMENT mLE: Evaluation of Eddy Curr:nt Indications Detected '
Durina the 1984 Tech. Spec. Inspection //
ORIGINATOR SIGNATURE DATE Vgtg gg ATE J. A Janiszewski Q O.. *
, )/ti/fg G. E. Ven Niedd W C [/,, h l /////Pc N. C. Kazanas A] sWekW- - a {n {Ar n L cennah- oarklll $LL /*/d-oS~
APPROVAL h EXhANAh DISTRh10N DATE R. F. Wilson h N \q---. nim T \ L Does tesis TOR Inolude roeommendation(s)? CYee E No If yee. TFWR/TR #
o D487AIGUTION ASSTRACT:
- In order to identify the cause of the eddy current R. O. Barley indications detected during the TMI-1 OTSG tube examination G. R. Capodanno beginning in November 1984. Materials Engineer!.g/ Failure Analysis yrfor ad an in-depth review of the eddy current J. J. Colitz results and plant operat.ing/che:istry history since the D. K. Croneberger OTSG's were first filled af ter -hn kinctic expansion repairs.
B. D. Ela:u Two possibic causos for tha eddy current indications were evslust:d: corrosion, aichar continuing or newly M. J. Graham initiatod, an,d enhanced eddy current. detectability of N. C. Kazanas existins intergranular attack (IGA). During unit layup.
CP'J layup specific.ations were followed. Some out of spec-
- 4. J. McGeey aficat. ion periods did occur;.however, they were promptly C frected and were not of sufficient magnitude to have caused i'. A. Richter l
corrosion. Additional corrosion-preventive conditions were G. R. Taylor also maintained during layup.
R. F. Wilson During hot operations, sys:en chemistry conditions were T. G. Broughton maintained within specifications that industry experience and TMI-1 tube testing have shown are non-corrosive.
W. Bloomfield The most likely reason for having eddy current indica -
tions at this time was enhanced detectability of pre-existing areas of IGA. As a result of thermally induced strains and hydraulic forces during hot functional testing, grains could fall out or grain boundaries could separate for a short distance within pre-existing ICA resulting in greater local distu::bance and a correspondingly larger eddy current signal.
i Addit.ional plant data from leak rate observations and i
the fiberscope examination of a sample of tubes also support
, the r.echsnical damage scenario. No leaks have been identi-l fied in :he cubs free span since 1983. In the region of 1984 eddy current indications, patch-like indications sugges-tive of IGA were seen by the fiberscope examination.
~
cCOVER PAGE QNLY aoooooso4.sa
TDR 638 Rev. 00 Page 2 of 48 Table of Contents Page Introduction
Background
Evaluation of Eddy Current Results Post-Baseline Growth Studies In-Process Testing ISI Indications June 1984 Testing 100 Tube Sample November 1984 1984 Technical Specification Required Testing Spatial Distribution Characterization of Indications Degraded Tubes Chemistry Specifications Corrosion Experience with Inconel 600 Corrosion Test Results Long Term Corrosion Test Short Tern Test Results Bulk vs. Surface Effects TMI-1 Chemistry Guidelines Hot Operations .
Layup Chemistry and Operating History Review Data Base Results of Operational / Chemistry Review Chloride and Sulfate Oxygen Other Opirational Considerations
1 TDR 638 Rev. O Page 3 of 48 Page In-Plant Observations Leak Testing .
Fiberscope Inspection of Selected Tubes Discussion General Detectability of Indications by Eddy Current Conclusions 4 References Appendix 9
, - - g -
e mm --- == cv- -
a w - , - . , w - - - - - . -- - - -.e - - . , , .-,-c.-e--w v - - e-e, -
TDR 638 Rev. O Page 4 of 48 Introduction In accordance with the requirements of Technical Specification 4.19, eddy current testing of the OTSG tubing at TMI-1 was begun in November 1984. Initial testing with the 0.540" high gain standard differential probe method revealed previously unreported indications in the unexpanded portions of the OTSG tubes between the tube sheets.
Two possible causes for the eddy current indications were identified and evaluated; first, whether corrosion of the OTSG tubes caused either new defects or growth of existing defects and second, whether straining of existing defects caused them to become more detectable by eddy current. Since the original 100% baseline inspection of the OTSG tubes in 1982, the tubes have been subjected to mechanical loading during the kinetic expansion and thermal and hydraulic loads during the two hot functional tests.
In order to attempt to determine the cause of these indications, the Materials Engineering / Failure Analysis group reviewed 1) the historical eddy current data and 2) plant operational and chemistry data since the OTSG's were filled after the kinetic expansion repair of the tubes.
Based on the results of this review, the cause of the indications is discussed. Data supporting the conclusion are also included.
i
TDR 638 Rev. O Page 5 of 48
. Background
As defined by Technical Specification 4.19, GPUN conducted eddy current examinations of both steam generators at TMI Unit 1. Performance of this examination ultimately resulted in 100% of the tubes in A-0TSG and all tubes in the outer 16 tube periphery of the B-0TSG being examined.
The B-0TSG had only a limited number of indications with an indicated through-wall extent greater than 40%. Due to the limited number of B-0TSG indications, statistically-based analysis is not feasible. All these indications, however, are located near the outer periphery of the B-0TSG.
The following generalizations about the EC indications can be drawn from the A-0TSG results: ,
- 1. They are primarily located in the upper tube sheet and 16th tube span area.
- 2. They are concentrated in the outer periphery, but some indications occur across the entire OTSG.
- 3. Most indications are less than 50% through wall.
- 4. They generally exhibit voltages in the 0.5-2 v. range.
5 By 8 x 1 absolute eddy current, the number of coils tends to be 2 or less, indicating a small circumferential extent.
TDR 638 Rev. O Page 6 of 48 Evaluation of Eddy Current Results Note: This section uses the eddy current data base as of Jan. 3,1985.
GPUN conducted a qualified full-length, eddy current examination program on all tubes from both generators during July to November 1982.
The purpose of this program was to screen out all relevant indications and establish a 6" qualified length in the kinetically expanded zone immediately above the new transition zone which was essentially indication free. It was further established that, although we were using a process that was approximately 175% more sensitive than previously used at TMI in performing eddy current examinations, small defects below the threshold of detection could exist. Reference 1 identifies the maximum size of these small defects which could possibly go undetected.
Prior to the expansion, a 100-tube sample of tubes in each generator was eddy current tested periodically to check for indication changes. These tests were performed on seven occasions over a 7 month period. No growth was observed.
Post-Baseline Growth Studies In-Process Testing During and following the kinetic expansion repair, a total of 437 tubes were inspected in both the A and B generators (Ref 2, 3).
A total of 15 tubes (3.5%) with indications were found that had not been detected by our ECT inspection program prior to the repair.
An evaluation was performed on why these indications were not identified previously (Ref. 3). It was - '"ded that:
- 1) The recent. indications were not i . " the kinetic expansion process nor was there an;, n of ductile propagation of existing indications.
- 2) The defects were small (threshold) type indications that had been either masked by the high background noise levels in the upper tube regions or were sufficiently tight that sufficient metal removal was not present to permit detection. Kinetic i expansion may have altered these areas of IGA to make them l more detectable.
l l
Confirmation on the small size of the indications was established by the visual examination using fiber-optics. Some of the indications appeared to be small pits.
l
~
l 1*
e
TDR 638 Rev. O Page 7 of 48 Additional confirmation was obtained that kinetic expansion would not cause ductfie tearing by using test mock-ups and metallurgical examination (Ref. 2). Small intergranular stress
- assisted (IGSAC) cracks were examined using eddy current techniques before and after kinetic expansions. Expansion caused the cracks to become non-detectable by .540" S.D. techniques. However, the cracks remained visible to the 8 X 1 absolute technique with essentially no change in signal. These specimen tubes were subsequently removed from the test block and metallurgical examination did not reveal ductile tearing or generation of new indications.
ISI Indications During OTSG repairs, a subset of tubes (28 in A-0TSG, 56 in B-0TSG) was identified as having eddy current indications that did not require plugging. That is, the indications were less than 40%
through wall, not in the lane / lane wedge area, and below the 15th tube support plate. This group of tubes (designated as "ISI" tubes by GPUN) was fully characterized and listed for eddy current inspection in the future as a distinct subset.
The "ISI" tubes were re-examined in April /itay 1983. No growth of the existing indications was detected.
As part of the eddy current campaign which started in October 1984, a!1 84 of the "ISI" tubes have been retested. No growth in the IS', subset was detected. (Growth is identified as a substantial increase in the through wall percentage, combined with an increase in voltage and circumferential extent.)
June 1984 Testing During June 1984, 67 tubes in B-0TSG and 3 tubes in A-0TSG were eddy current tested. This set of tubes was retested in November 1984 - no new indications were detected for the two retests performed.
100 Tube Sample November 1984 Since discovery of the additional indications in tlovember 1984, a second 100 tube sample with indications has been re-examined at approximate two week intervals. As of December 18,
. 1984, no growth and no new indications have been detected for the two retests performed.
TDR 638 Rev. O Page 8 of 48 1984 Technical Specification Required Testing In flovember 1984, eddy current testing required by TMI-1 Technical Specification 4.19 was conducted as specified. 3% of the tubes in each generator were initially examined. This examination included tubes randomly selected across the entire generator plus a concentrated examination in the periphery of each generator. The more extensive examination in the periphery was performed becruse this was the region of highest previous (1981) damage .
As a result of this initial examination, OTSG A was classified as category "C-3" per technical specification and OTSG B was classified as category "C-2". Subsequently the entire A-0TSG was inspected while the B-0TSG inspection was complete after the entire 16-tube periphery, approximately 6500 tubes, had been examined.
The number of indications is much higher in A-0TSG than B-0TSG. In A-0TSG, 2.0% of the tubes (299 out of approximately 14589) have indications greater than 40% through wall while in B-0TSG, 0.5% (33 out of approximately 6576) have such indicatio,ns.
Spatial Distribution The indications with greater than 40% through wall extent are concentrated toward the outer periphery and top of A-0TSG. In the outer periphery, the percentage of tubes with greater than 40%
through wall indications is higher than the 2.0% average, wh11r.
inside the outer support rods the percentage of indications is below 1%. 71% of the indications are located above the 15th tube support plate (TSP).
Characterization of Indications To understand the nature of the defects . .. .-
characterized the indications reported back in t!.c ~~;-1982 time frame and compared them to the indications discovec today.
The axial and radial locations of indications in A-0TSG are essentially the same in 1984 as in 1982, if one does not consider the 1982 indications in the kinetically expanded region in the 1984 evaluation.
TDR 638 Rev. O Page 9 of 48 1
Table 1 characterizes the 1982 and 1984 eddy current signals.
The 1984 eddy current indications exhibit a similar type of signal response as the previous test program. Detafis of the differences in responses are noted below:
- 1) Reported voltages are essentially the same. This indicates that the 1984 indications present a similar volume for the eddy current probe to detect as the 1982 IGSAC.
- 2) Both through wa11' penetration and number of cofis is significantly lower in 1984. Thus, the 1984 indications extend a shorter distance both into and around the OTSG tube.
Statistical analysis of the eddy current data reveals that 90% of the observed indications fall between 10% and 50% through wall penetration, and between .020" and .190" long.
Degraded Tubes Per GPUN procedure, tubes with indications reported between 20 and 40% through wall were not required to be plugged if the tubes were not in the lane or lane wedge and the indication was below the 15th tube support plate. At the completion of the 1982 kinetic expansion repairs, a total of 15 A-0TSG tubes and 51 B-0TSG tubes were classified as " degraded" and were included in the ISI group.
As of January 4, 1985, 347 additional A-0TSG tubes and 98 additional B-0TSG tubes are classed as degraded.
s
-_,_~,,,------vr--w. -,--*.-+.%- ---- -- -
-y * +-- -m-_y -
i TDR 638 Rev. O Page 10 of 48 Table 1 Comparison of 1982 and 1984 Eddy Current Data a) Reported Voltage - % of indications reported A-0TSG B-0TSG Voltage 1982 1984 1982 1984 t1 34 40 24 27 1 44 35 30 21 2 16 20 25 29 3 4 4 10 12
>3 2 1 11 11 b) Reported through wall penetration - % of indications A-0TSG B-0TSG
% T.ll. 1982 1984 1982 1984 4 20 41 41 12 20-40 3 61 28 75 40-60 21 25 24 18 60-80 17 10 15 5
> 80 59 4 21 2 c) Number of coils on 8 x 1 examination - %
A-0TSC " 0TSG Co f f s 1932 1984 2 1984 1 20 90 18 80 2 2G 10 24 20 3 16 41 15 41
>3 38 41 43 4.1 NOTE: 1982 data includes inspection of original tube roll transition area.
The 1984 data does not include inspection from the top of tube sheet to the bottom cf the kinetically expanded region.
~
e 9
TDR 638 i
Rev. O Chemistry Specifications Corrosion Experience with Inconel 600 Three types of primary-side initiated attack have been identified in Inconel 600. In recirculating steam generators using mill-annealed tubes that have not been stress-reifeved after U-bending, stress corrosion cracking (SCC) has initiated from the primary side in the highly stressed bend areas. Also in mill-annealed tubes in recirculating steam generators, SCC has been found to initiate from the primary side at highly stressed transition areas in the lower tubesheet. Laboratory studies have shown that the stress relieved Inconel tubing used in OTSG's is significantly more resistant to SCC than the mill annealed type.
The other primary side attack of Inconel 600 that has occurred in steam generators is the intergranular stress assisted cracking (IGSAC) caused by reduced sulfur species on sensitized 0TSG tubing. This is the mechanism which caused the TMI-1 OTSG 1eakage in 1981. This mechanism requires sensitized tubing, low temperatures, oxygen, and significant levels of reduced sulfur species. ,
1 Corrosion Test Results As part of the overall program to evaluate the most recent eddy current testing results, we have reviewed the results of corrosion tests performed as part of the original failure analysis and OTSG requalification programs. These data provided a partial basis upon which we could evaluate the layup and test conditions to which the steam generatort, had been subjected.
Lor.g Term Corrosion Test (LTCT)
The primary purpose of the long term corrosion tests was to verify that the proposed operating chennstry specifications are satisfactory to prevent corrosive attack of the OTSG tubes. To this end, chemistry conditions for the testing were established at the maximum allowable values consistent with the upgraded THI-1 operating specification (Ref. 4). The LTCT was conducted using actual THI-1 tubing. Temperatures, tube loads and heatup and cooldownrateswererepresentativeofactualplantoperating conditions.
In addition, as the LTCT was actually performed, specific factors which prallel actual plant layup conditions were .
experienced. The tubes were held in a cold, aerated condition for several days after the completion of each operating cycle.
Aeration was done after cooldown. Before heatups, or while waiting for other autoclaves in the test program to be ready for operation, the test loops were operated in a cold, deaerated, circulating mode. Because eddy current examinations were done after each test cycle, the tubes had to be removed from the autoclaves and drained. Thus, drained aerated layup conditions were also included.
TDR 638 Rey, O Page 12 of 48 Table 2 sunearizes LTCT operational times in each mode. AIT loops spent significant time under drained, cold deaerated, and aerated conditions.
Review of the chemistry history of the LTCT's revealed that the conditions were comparable to the plant's experience. The LTCT specification (Ref 5) for sulfate and chlorides was 0.100 ppm +
.050 ppm. Actual analysis results (Ref. 6, 7, 8) revealed that the concentrations of these species were maintained at or slightly above the .150 ppm upper'lfmit. The actual values measured in these tests bound any of the contaminant " spikes" reported in the Chemistry and Operational History Review.
C-ring tube samples from archive tubing (tubing never installed in the Tt1I-1 OTSG's, wht:h uas included as a control sample) showed no evidence of cracting, pitting or general corrosion.
Some intergranular attack (IGA) was noted on 4 C-rings made from a single TilI-1 OTSG tube; this ICA was evaluated to be pre-existing damage associated with the 19C1 IGSAC incident. Of a total of 38 C-rings evaluated, 31 had no visible defects, 3 showed very shallow cracks when strained f.everely, and 4 had IGA as described above.
Five full tube samples were meta 11ographically examined after the LTCT. In addition to previously reported defects, four samples exhibited scattered, shallow cracking or IGA which was not detectable by eddy current testing. This IGA was consistent in size and shape with IGA that had been seen during the failure analysis (Ref. 9). Therefore, the observed '" on these four tubes was judged to have been present at the st- ? LTCT.
One tube sample had severe IGSAC and M ... ~ad progressed during the term of the LTCT and had been deteu.n .., eddy current.
The tube sample which showed flaw growth during ~ 'TCT was exposed in the test loop in which the sulfur species was thiosulfate, at a concentration of 0.100 ppm + 0.050 ppu (as sulfate). Therefore, the only tube sample exhibitin during the LTCT was exposed to intentionally added, g flaw growth reduced corrosive sulfur species.
The four C-ring samples showing IGA and the full tube sample showing flaw growth were removed from the same 0TSG tube. This tube was recorded as having multiple eddy current indications when inspected in the OTSG. The IGA seen in the post-test examination is therefore consistent with an original tube sample which had multiple defects and, presumably, associated IGA.
Results of metallographic examination of the LTCT samples (Ref. 8) confirmed that in the absence of intentionally added aggressive sulfur species, normal operations would not cause corrosion of Till-1 OTSG tubing.
1 TOR 638 Rev. O Short Term Test Results l Several sets of tests were previously run on Inconel 600 tubing to establish corrosion resistance under various conditions representative of TMI-1 service. Those results which apply to the period of this review are summarized below:
- 1) Screening work on actual TIC-1 removed tubes and archive tubes (Ref. 10) identified that at oxidizing potentials, 1 ppm of thiosulfate was required to cause IGSAC.
Sulfate levels'as high as 10 ppm did not cause IGSAC.
- 2) Simulation of hot functional testing and cooldown (Ref.
- 11) utilizing thiosulfate contamination and actual operating temperatures and times revealed that 1 ppm of thiosulfate caused IGSAC.
These short term tests thus confirmed that in the absence of thiosulfate contamination, no short term attack of OTSG tubes is expected.
Bulk vs. Surface Effects The above corrosion tests were performed using actual THI-1 OTSG tubing. The surface film condition was therefore representative of that in the plant. Chemistry control in both corrosion testing and actual operation is done by the measurement and control of species of interest in the bulk fluid.
Since both surface conditions and chemistry control were identical between the laboratory tests and plant operations, the r'esults of the corrosion tests can be directly applied to the plant environment, and, conversely, plant bulk chemistry data can be used to evaluate the propensity for corrosion.
.MI-1 Chemistry Guidelines Hot Operations After sulfur was identified as the causative agent of the 1981 IGSAC, hot operational guidelines (Ref. 4) were reviewed to ensure that adequate corrosion protection was maintained. As a result of this review, two changes were made to provide increased margins against corrosive attack.
First, a requirement was added that primary system sulfate be maintained below 0.100 ppm. Sulfate at this level does not cause corrosive attack of Inconel 600 in primary coolant, and maintaining sulfate below this level provided assurance that intermediate sulfur species could not exist at harmful concentrations.
Second, the lower ifmit on lithium concentration was increased to 1.0 ppm, to take advantages of lithium's inhibiting effect on sulfur-induced IGSAC in Inconel 600 (Ref. 12).
l TDR 638 Rev. O Page 14 of 48 The net result of these changes is to ensure that total sulfur species concentrations are a factor of 10 below the levc1 at which corrosive attack might occur. At the same time, the minimum Lf/S ratio will be 30 (or Li/SO4 of 10), which is a factor of 3 over the recommended (Ref. 12) ratio of 10 for inhibition of IGSAC initiation.
Layup For cold layup conditions, guidelines have been estabitshed to maintain as many protective conditions as feasible. The individual protective conditions that are feasible for the THI-1 RCS are:
- 1) Elevated pH - during layup, pH has been elevated, using amonia, to at least 7.2. The normal pH without amonia is 5.6 - 6.5.
- 2) Control of contaminants - The primary water contaminants or concern are cniorices and sulfates. Chlorides have traditionally been limited to less than 0.100 ppm during operation; we have maintained this level as a general guideline during layup. The sulfate level of less than 0.100 ppm used during hot operation also applies to layup.
- 3) Control of oxygen level - When the system is ft11ed and able to be pressurized,'the cxygen level is to be maintained below 0.1 ppm. For cases where the prieary system is open and oxygen cannot be excluded, air saturated conditions are specified as this is more protective than some intermediate oxygen level.
- 4) Control of OTSG 1evel - One of '
.ibuting factors to tne in: 16dAL, incident was -ance of a water line on the primary side of the 7. . .. For layup of the OTSG's, wherever possible, no stan: ::sterifne shall be allowed to exist in the OTSG tubes. ither the water level should be above the upper tubesheet or the OTSG primary side should be fully drained. *
- 5) Inventory Turnover - Periodic replenishing of the OTSC contents will assure that local buildup of contaminants will not occur. Layup guidelines have included provisions for periodically turning over the water inventory on the OTSG primary side to meet this objective.
- - - - - - g --- - .-
TDR 638 Rev. O Page 15 of 48 TABLE 2 Summary of Operations for Long Term Corrosion Tests Operating Days Cold Circulating ' Drained Loop Hot Deaerated Aerated Layup (Note 1) Comments 1 348 52 28 132 2 308 69 27 157 Thiosulfate loop 3 241 42 23 58 4 242 40 22 61 i
Notes
- 1. Dces not include drained layup between completion of operational cycles and start of metallographic examination.
i i
_ _ . _ ,, . . _ . . , . . _ . - . . , . , ,....,_.,...-,,,,_,_.._,,,..__...,_-_.,_,,___,_,__m,..;_ ___,._,_.._,__,r.~., , . _ _ __ , . , , , _ _ .,
l TDR 638 Rev. O Page 16 of 48 Chemistry and Operating History Review Data Base The chemistry and operating history data were obtained from two sources. First, the on-site Plant Analysis group reviewed operational records to identify plant conditions during this time period (Ref.13).
Then, we retrieved the primary plant chemistry parameters of interest from the GPUN ccmputerized chemistry data base.
The major plant activities that occurred between May 1983 and October 1984 are listed in Table 3. Within each of these periods, we identified different plant conditions of RCS level, tem?erature, pressure, circulation, and pH. Then, we reviewed the c1emistry data for each time period.
Chemistry data selected to be of interest with respect to corrosion were pH, oxygen, 11thium, sulfate and chloride. As an additional check on the effectiveness of chemistry controls, we calculated the lithium to sulfur ratio for each operating period. In cases where simultaneous analyses for 11thium and sulfate exist, we calculated the Li/S ratio for each data point.
The data from the operational and chemistry investigations are plotted as a function of time in Appendix A.
Results of Operational / Chemistry Review During both hot shutdown and cold layur -ditions, TitI-1 has maintained conditions within chemistry guid . ::s about 95% of the time. For short time periods, some deviati /- ccurred which are discussed in the balance of this section.
Chloride and Sulfate There have been short time periods where chlorides and/or sulfates have exceeded specified limits. In all instances chemistry data reflect that corrective actions were appropriately and promptly taken to return the concentrations of these species to specified levels. Collectively, these out-of-specification periods can best be described as normal chemistry " spikes".
TDR 638 Rev. O Page 17 of 48 0xygen In preparation for both the September 1983 and May 1984 hot functional tests, it was necessary for the RCS to be taken from a layup to an operating mode. During this transition, oxygen levels were higher than desired for optimum protection, but other factors made it very unlikely that corrosion occurred. First, chloride and sulfate concentrations were controlled to acceptably low levels.
Second, the lithium level was maintained such that the minimum lithium to sulfur ratio was 66; the recommended minimum value for protection against IGSAC is 10 (Ref. 12). Chemistry control during these periods is summarized in Table 4.
Other Operational Considerations During the Integrated Leak Rate Test (ILRT) in April 1984, the primary side water level was maintained at about the 12th tube support plate for 8 days. This condition was both preceded and followed by drained layup with elevated pH, aerated water. Both sulfate and chloride levels remained within specification. Therefore, no OTSG tube corrosion was expected.
In August 1983 and May 1984 oxygenated water was injected into deoxygenated RCS during HPI testing. Most of these tests were conducted prior to the high temperature portion of the hot functional tests, and the oxygen introduced would have been consumed by hydrazine and/or hydrogen added for that purpose. One test was conducted on May 26, 1984, at the end of HFT and may be postulated to have injected 5000-6000 ga113ns of oxygen-saturated water. During this time period, however, the lithium to sulfur ratio was greater than 30 which was mdre than adequate to inhibit corrosion during this test.
l TDR 638 Rev. 0 Page 18 of 48 l-l TABLE 3 t
Major Plant Evolutions, 5/83 to 10/24 Event Duration Fill & Bubble Test June 1983 Peroxide Clean July 1983 Hot Functional Test Aug - Oct 1983 Circulating Wet Layup Oct - Nov 1983 DH-V1 Repair Nov 1983 Circulating Wet Layup Nov 1983 - Jan 1984 RC-P1B Repair Feb - April 1984 Integrated Leak Rate Test April 1984 Hot Functional Test itay 1984 Non-Circulating Wet Layup . May - June 1984 Tube Plug Rerolling and Oct 1984 Bubble Testing
^%
TDR 638 Rev. 0
- Page 19 of 48 TABLE 4 Chemistry Sumary Before Hot Functional Testing Oxygen, Li, SO4 , C1 Lf/S Period Days ppm ppm ppm ppm Ratio 8/83 29 0.3 .82-1.96 .047 .079 .05 .156 66-123 5/84 19 .075-2.2 1.06-2.17 .02 .047 .05 .110 127-240 1
4 I
s V
. - - . - - - + - -
. ,.. - - - - - . - - , - . --. . . _ . . . . _ , - . - - - s--.._ -. , - . , .. , - , .
TDR 638 Rev. O Page 20 of 48 In-Plant Observations Leak Testing
. Since completion of the kinetic expansion repairs, several leak tests have been performed to measure primary-to-secondary leakrates and identify individual leaking tubes. These tests are summarized in Table 5.
No pattern of tube leakag'e can be seen. After the cooldown tests included in hot functional testing some increase in leakage was seen.
Further investigation showed that this leakage was the result of leaks through a small number of tubes. These leaks were located in the expanded region within the upper tube sheet and were repaired by 2
mechanically rolling a portion of the expanded area.
Of greatest significance is that since 1983 no tube which is in service has had a leak in an unexpanded portion of the tube. All leaks have either been due to bypass leaks in the expanded area or leaking plugs.
Fiberscope Inspection of Selected Tubes A fiberscope inspection was performed (Ref. 14) of six A-0TSG tubes which exhibited typical eddy current indications. During the inspection features were observed on 4 out of 6 tubes at the same elevation as the eddy current indications.
The visual features were "patchlike" rounded areas having an outer ring which was darker than the general tube surface and slightly ,
reflective components in the interior. Th- '
-hes were between 0.020 and 0.060" in diameter.
The patches appeared similar to surto. et"s seen during the initial tu)e failure analysis. These earlier v : ::s were found to be associated with partial through wall intergranui. .ttack.
1 i
P
._,,,..__s ,-.,,m ,,.__m- _.,_,y ,_,.._, ,,,,,, .g,, m,, _,m__,,--.-_
~#-. . --- - , -. ..-__.,,._.m-mm..__,,_, , . , _ ,
TDR 638
-Rsv. O Page 21 of 48 Leak Tests in OTSG's Since 5/01/83 Month / Year Test Type Reason For Test Results Repairs May 1983 Drip- Test of Kinetic 2 Leaking Tubes, .
Plugs In. stalled / Rerolled Expansion 8 Leaking Rolled Plugs 10 Leaking Explosive Plugs June 1983 Bubble / Drip Final Test of Small Number of Slightly Repaired welded plug Kinetic Expansion Leaking Tubes and Plugs in A OTSG - 1 Leaking welded plug Sept 1983 Kr-85 Tracer Establish Baseline Baseline Leak Rate None Required Leak Rate 1 gph ,
May 1984 Kr-85 Tracer Measure Baseline Slight Increase in None Required Leak Rate Leak Rate June 1984 Bubble / Drip Identify Leaking 4-5 Leaking Tubes in Plug 3 tubes l Tube (s)
B-0TSG w/ welded plugs 6 Rolled Plugs Missing Reroll all W plugs Replugged tubes.
Oct. 1984 Bubble / Drip Test Rolled Small Number of Leaking Roll 8 Tubes Repairs Tubes, one welded plug Reweld Plug Note: No leaks seen in final October 1984 Bubble Test, after tube rolling.
_.__m.-...,_,_- ....e..
TDR 638 R2v. 0
[
Page 22 of 48 Discussion t
General Removal of sodium thiosulfate from the TMI-1 site and tighter operational chemistry controis implemented since 1981 have made it highly unlikely that the conditions to cause sulfur-induced IGSAC could be f recreated. The steam generator layup guidelines are specifically designed to protect the steam generators from additional corrosion and are more stringent than BW's generic recommendations, particularly in the areas of contaminant control and the use of elevated pH during cold layup.
Industry experience on B&W PWR's also does not reveal any other primary-side initiated attack mechanisms on Inconel OTSG tubing. l TMI-1 compliance with operating and layup specifications has been excellent. Transient out-of-specification conditions, which were identified during plant operation, have been infrequent and corrected promptly by the plant operators. Plant conditions have always been bounded by those which were evaluated during corrosion testing and found to be satisfactory.
The only period of possible vulnerability to corrosion would have existed during the time when the OTSG's were drained for the kinetic expansion repair. During this period sulfur would have remained in the oxide film on the tube surfaces as peroxide cleaning had not yet been performed. During this time, however, eddy current testing done on the 100 tube surveillance sample did not reveal any growth of existing indications or any new indications. Thus, while the oxidre film may have contained sulfur during this time, there is no evidence t1at corrosion continued.
Previously detected IGA, both ir '!ure analysis (Ref. 9) and long term corrosion test (Ref. 8), ha: -)een in the form of hemispherical pits penetrating approxf .- through waii. A pit of this shape and penetration would appear au ' t e on the surface of diemeter of approximately 0.035. Areas of th '; ::frcumferential extent would not be predicted to be detectable by the .540 S.D. eddy current technique (Ref. 2). .
Under mechanical loadings induced by kinetic expansion or cooldown, these areas could become more detectable by eddy current through several mechanisms:
- 1) creation of a linear grain boundary separation within the IGA islands as was seen in the LTCT (Ref. 8), or
- 2) .isconnected d grains dropping out and leaving pits.
I i
TDR 638 Rev. O Page 23 of 48 Two additional pieces of data from Ref.161end support to the mechanical scenario. First, peripheral tubes consistently see higher loads than core tubes. Therefore, in the periphery, the highest stresses would also act on this IGA. Second, the A-0TSG cooled down more quickly than the B unit. The peak load during the most rapid cooldown (Ref. 16) was 200 lb, or higher (12%), in the A-0TSG than in B-0TSG. Figure 1 is a representation of how the A-0TSG would have had significantly more tubes carrying loads high enough to cause IGA to become more detectable.
A previous study (Ref.15) on crack opening displacement of archive ,
tubes with approximately .5" long through-wall cracks found that loads between 1500 and 2000 lbs. would induce permanent displacements in the vicinity of the cracks. Loads less than this would induce only elastic l displacements with a load of 1000 lbs. producing an elastic displacement i of approximately .002". Although tubes with cracks of this size are no longer in-service with the steam generators, this study does point out that one can expect local straining in the vicinity of smaller defects, l
l but that it would be of proportionately lesser magnitude.
During the 1983 HFT, the most rapid cooldown was calculated to have induced loads in the tubing of between 1600 and 1700 lbs. (Ref. 16). It is such loads acting on the regions of IGA which we believe leads to grain dropping or grain boundary separation.
Visual observations made during the fiberscope examination of selected OTSG tubes support the cause of the present eddy current indications being mechanical dama At locations where i
eddy current indications existed,ge to existing IGA.we frequently saw rounded, darkened areas of a size consistent with IGA detected in the original failure analysis.
Detectability of Indications by Eddy Current l It should be noted that the l tube integrity (i.e. tube rupture) primary defects of concern are circumferential for OTSG cracks. The production of 0.540" standard differential eddy current technique is optimized and qualified for this type of defect. However, it can also be used for detecting different defect geonetries as discussed below.
The 1984 tube ID indications as detected by eddy current and as seen during the fiberscope inspection had significantly different l characteristics than the IGSAC responsible for the 1981 tube leakage.
I The 1981 IGSAC consisted of tight, circumferential cracks that penetrated completely through the wall. The 1984 observed IGA is more rounded and does not completely penetrate the tube wall.
The different geometry will have a direct effect on detectability.
The current .540" S.D. eddy current technique was optimized for the IGSAC geometry; therefore, a different geccetry will have a different detectability. The balance of this section of this report will discuss changes in sensitivity due to changes in indication gacmetry.
TDR 638 Rev. 0
- Page 24 of 48 Figure 2 (Figure 2 from Reference 2) shows the measured sensitivity I of the .540" S.D. technique in the range of short circumferentially oriented defects. The shaded region in Fig. 2 identifies the area in which S0% of the 1984 indications fall. It can be seen that the eddy current calls span the 0.3 volt detectability limit. Thus only slight changes in indication geometry could cause a particular indication to become detectable.
In Figure 3a and 3b, we have taken the eddy current data and visual observations from the fiberscope inspection (shown in Table 6) and indicated where the indications would be in relationship to the calibration curves. The tubes for fibrescope inspection were chosen to be representative of the types of indications being found in 1984.
All of the below-UTS indications (Figure 3b) are close to the 0.3 y detectability limits; the within-UTS indications (Figure 3a) do not fall into the detectable range. Therefore, it is reasonable that before mechanical loading these indications may not have been detectable.
tiechanical loading, as discussed in the previous section, can alter IGA geometry.
Because the calibration was done on a length vs. through-wall basis using Edit notches of a constant axial width of about 0.004", IGA geometry could produce a different signal. Patch-type indications of the same length would have a larger axial extent, and therefore a larger volume, and could be expected to give a higher voltage signal. The S.D. response would also be enhanced by increased axial extent, even at constant defect volume, since the differential coils are wound in the circumferential direction and are more sensitive to the axial extent of nefects.
The large increase in the number of degraded tubes in A-0TSG and B-0TSG is also consistent with the scer-
- of pre-existing IGA becoming more detectable. IGA islands of 20-4' 5 wall extent would be expected to have a length of about .0. thes; this is below the 300 mV sensitivity for free-span detet.. e 2). The additional l disturbances of mechanical, thermal, and n; e loading could easily disturb these islands enough to now make th w ~e detectable.
l l
'lVR 611D Rev. 0 Page 25 of 48 Table 6 - Comparison of Preliminary Eddy Current Data arxl Fiberscope Results EC Results
.540 S.D. 8K1 Row Tube Elevation % T.W. Volts Volts Coils Visual Observations 4 89 124 US+5.4 98 1.6 1.6 2 US+4 Rounded indications possible ICA US+5.8 Axial alignment of 3 rounded indications 76 119 US+2.4 97 2.1 0.8 2 US+5.5 Small dark spot when scanning w/90* head 66 129 15+27. 6 70 2.8 1.3 2 15424.5 Rounded indications possible ICA 61 123 15+21.8 67 2.3 1.1 2 15+26 Small dark spot - no detail visible 15+24.7 45 1.7 0.5 1 57 128 US-2.6 92 1.3 0.3 1-2 US-1.5 Axially oriented rounded indications'
- e.
63 126 15-14.2/15-6.5 37/42 1/1.0 NDD Small single rounded indication
TDR 638
" Rev. O Peas 26 of 48
. "S
---a - -
l Peripheral Tube Location & Lead 1
E I B i e
U l "@
, a il l
T U l "6 I 8 j O O I
,, a lh
~
g ~T~~~~~~~ Average fEb~e I3catioT1"A lo~ai ~~
~
2 ta l 5 s 8 i l . s i s } i a $1 I 3
$O I .b l y 3 2 i -
=
$ *2 l 1=
1 I 5
ll- ]J *2
' a l
,e" a e
h 4%
, d el ,
i -
s e
- a 2 gli
- I 5
I
- u. I L I 'S 1
l 1 1
8 I
I k' h_
$ h 5 f_ ~ - ~
e OTSG Tubes Axial Lead (Lbs.)
TDR Rev. 0 Figure 2 - 1984 Eddy Current Date Page 27 of 48 Compared to Detectaability Limits ._ ._
1 I
187" ---- - - - - - - - - - - - - - - - - - - -
g
- 1 v. Detectability s Within UTS
\ '
l 100" ----
bs T
N x
.060" ----
's .
'N N'x s l .020" ----
N\
\
'N 0.3 v. Detectability below UTS
- ~ ~ - - - - - - - - - - - -
ANNNN'NN'N.
l I
O 20 40 60 80 100
% Through Wall 90% of 1984 Indications N' - Lie Within This Area
TD2 538 R;v. O Figure 3a - Within - Tubesheet Page 28 of 48 Fiberscope Indications Compared to Detectability Limit. .__ ... . ..
Notch Length
.187 __ __ ____ _ _ _ _ _ . . . _ _ _ .,
Non-Detectable Detectable 1 Volt Detectability Limit Within UTS 1
(
.100 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
.060 ----------------------------------------m
& A
.020 - - - - ~ ~ ~ ~ ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - *' -
0 20 40 60 80
% Through Wall i
Tube Identification o"
A A-89-124 s B A-76-119
- - TDR 638 Ev. 0 Page 29 of 48 Figure 3b - Below - Tubesheet Fiberscope Indications Compared to _ . . ,
Detectability Limit Natch Length
- ----------~~'
- 187
.Ton -6etectable
ble Detecta l
0.3 Volt Detectability Li: nit Below UTS
.100
.060 -------------,r < r - - - - - - -- m r - - - - - - v - - "
/ .
5 sre O C l .020 -------------d-'--------#------ 4 ~ --
l I
20 'O 60 80 O
% Through Wall Tube Identification ,
C A-66-129
. D A-61-123 E A-57-128 F A-60-126
TDR 638 Rev. O Page 30 of 48 Conclusions i 1. The TMI-l layup guidelines are adequate to prevent any l identified mechanisms for primary side initiated corrc<.fon of Inconel 600 OTSG tubes.
l
- 2. The THI-1 1ayup guidelines have been adhered to since l completion of the kinetic expansion repair. Minor deviations have been corrected promptly.
l
- 3. Vulnerability to corrosion may have existed during the period f when the OTSG's were drained for repair prior to peroxide cleaning. However, eddy current data and the absence of OTSG 1eakage during this time period do not show evidence of corrosion of OTSG tubes.
- 4. Results of both GPUN-sponsored and industry corrosion test programs confirm that corrosion would not be expected during TMI-1 operations since May 1983.
- 5. Results of eddy current tests since 1982 do not indicate any trends of indication growth of pre-existing indications.
- 6. Leak rate testing and 0TSG bubble testing do not indicate any increases in leakage or new leaks in the tube free span.
- 7. The eddy current data and visual observations are consistent with a mechanism where previously existing areas of intergranular attack are raade more detectable by mechanical loading during kinetic e).pansion and thermal and hydraulic loading ' 'm cooldown from HFT.
( -
TDR 638 t
Rev. 0 Page 31 of 48 References f
- 1. R. Barley, J. A. Janiszewski, G. E. Rhedrick and M. Torborg, "Three Mile Island - Unit 1 OTSG Tubing Eddy Current Pro Qualification," GPUN Technical Data Report 423, Rev. 1, gram January 1984.
- 2. G. E. Rhedrick, " Task IV Report on Eddy Current Indications l Found Subsequent to Kinetic Expansion of THI-1 Steam Generator Tubes," GPUN Technical Data Report 401, Rev. O. April 1983. l
- 3. T. M. Moran, " Assessment of TMI-1 Plant Safety for Return to Service After Steam Generator Repair," GPUN Topical Report 008, Rev. 3, August 19, 1983.
- 4. " Primary Water Chemistry," GFUN Specification SP-1101-28-001, Rev. 3, July 11, 1984.
- 5. "Long Term Corrosion Testing," GPUN Specification SP-1101-22-008, Rev. 2, Oct. 29,1983.
- 6. "Long Term Corrosion Test Program of Nuclear Steam Generator Tubing Samples from Three Mile Island Unit 1 - First Interim Report," Westinghouse Electric Corp. Report No.
0914c/0127c/010684:5, October 1983.
- 7. "Long Term Corrosion Test Program of Nuclear Steam Generator Tubing Samples from Three Mile Island Unit 1 - Second Interim Report," Uestinghouse Electric Corp., May 1984.
S. "Long Term Corrosion Test Program of Nuclear Steam Generator Tubing Samples from Three Mile Island Unit 1 - Final Report,"
f Westinghouse Electric Corp., unpublished draft.
f 9. A. K. Agrawal, W. N. Stiegelmeyer, and W. E. Berry, " Final Report on Failure Analysis of Inconel 600 Tubes From OTSG A and B of Three Mile Island Unit 1," Battelle Columbus Laboratories, June 30, 1982.
- 10. J. V. Monter and G. J. Theus, "TMI-l OTSG Corrosion Test Program - Final Report," Babcock & Uficox Report No.
RDD:83:5433-01-01:01,11ay 9,1983.
- 11. J. C. Griess and J. H. DeVan, " Behavior of Inconel 600 in Sulfur-Contaminated Boric Acid Solutions," Oak Ridge National Laboratory Report ORNL/TM-8544, March 1983.
Task 638 Rev. O Page 32 of 48
- 12. R. Bandy and K. Kelly, " Investigation of the Sulfur and Lithium to Sulfur Ratio Threshold in Stress Corrosion Cracking of Sensitized Ailoy 600 in Borated Thiosulfate Solution,"
USNRC NUREG/CR-3834, July 1984.
- 13. J. R. Kasper, "TMI-l Primary Plant Status from 5/01/33 to ~
11/27/84," GPUN IOM PA-TMI-84-47, Nov. 27, 1984.
- 14. J. A. Janiszewski, " Observations During Fiberscope Inspection of A-0TSG Tubes, Dec'. 19, 1984," GPUN IOM MT1-1550, December 28, 1984.
- 15. J. A. Janiszewski, " Leakage and Crack Opening Displacement of OTSG Tubes," GPUN Technical Data Report 480, Rev. 1.
- 16. G. L. Lehmann, T. M. Moran, J. R. Sipp and D. G. Slear, "TMI-1 OTSG Hot Testing Results and Evaluation," GPUN Technical Data Report 488, Rev. O, October 25, 1983.
mhmmai mmai s isici
~ -
TDR 638 Rev. O Page 33 of 48 APPENDIX A TMI-1 CHEMISTRY DATA MAY 1, 1983 to OCTOBER 26, 1984 Contents Table A1 - Chemistry Guidelines Applied to TMI-1 5/1/83 to 10/26/84 I Figure Al Al Chemistry Data for TMI-1
- 5/1/83 to 10/26/84 1
i f
l-s I
s
TDR 638 Rev. O i Page 34 of 48 l
l- Table Al CHEttISTRY GUIDELIllES APPLIED TO THI-1 ,
5/1/83 to 10/26/84 Operating Wet Drained Hot Shutdown Peroxide Mode Layup Layup (Hot Functional Testing) Cleaning ,
1 OTSG Primary Level Full Drained Ful's Full
!!aximum Chloride, ppm 0.1 0.1 0.1 0.2 Maximum Sulfate, ppm 0.1 0.1 0.1 Note 2 Maximum Oxygen, ppm 0.1 N/A 0.1 Note 2 pH greater than 7.2 4.6-F,.5 4.6-8.5 8.0-8.5 Li, ppm ' 0-?.0 1.0-2.0 1.0-2.0 1.8-2.5 Minimum Lf/S ratio 10 10 fl/A Notes:
- 2. Sulfate and oxygen were monitored but no limit was applied.
-~
TDR 638 Rsv. O Pags 35 of 4G 1
~
O O
,e O et. _
-=
C T1 M_ -'I
.- =
1 ~ ^
-R
-I I .
- a 2
-=
g f g i
m g
lll f :
2 _-
M :"
f et - -=
i g'j l
- o w
o n
o 4
o A
d .a I:n 1
0 TDR 638 R3v. 0
$ Pass 36 of 48 2
1 3 h o a e
E en a.
~
M o"
m, .
l
-, 5
~
iE f [ -s
. g
-5
-s
-R g m 4 3 .
i e lIg* -sk=}
g
-t
-E
-s T u.
1#
=
A
......m
-s a-i 9 ..
5 b, g=
, a 4 : .
4 o n-__--___ _ o .n a n. -____ - _ __ -_ __ _ ___ _ _ n n _ n n s_ ~ a n _ _ oncoo on 99 o.m.
(
i s= $ '
~y 1 I m ~jta! :
p
- m p p
- p -
- 4 l O C
T .
o S
P O H
R O
F g '. s n
S i
t s
N T e
E O
l a
I n T i o
A t c
n
_ E R F u _
A3 t o
P89 E1 Rt H
f r
o e
e A 2
s Pse r p
a -
5 1
A s e r e Da P -
w g
NA 3 H
Ay l Hl Na
&o v -
GJu m
4e OR N
I S
I l
k N g A i n
E n a
e I
L l C
C d e
i x ,"
E o r
D I
P e
1 r
X -
, 1 1 a -
O 6 3 2 1 0
n o
R 0 0 0 E
P (l l{I lIl f 1 1, f i1 ! >) Ij I ;
1 lll
, g o.
i wo<, o ,
.o,g r ,= 2 D00 m
p p
En S i h /i.
L PT T o e P
H o w
0 2
1 o
o l
o s
o s
0 4
a 2
O
- - - s a
R O
F .s e
t s
- e S T ,
N l a .s e O
I i n
o T
t c
n
_ A F u =
= .w AR t o
TA3 AP8 H
r o b-d' .n 9
DE1 f
e '
r Rt ap h YPs u P e
r
.e 2 1
A R s T e TDsA .i S
U G
u r
g NN-y Ela 3
, ( U A
R .
MA ,
l E
_ R GJu O" 4
.r e S *
.- HN CI ,1' 5=
N g s
e
.a s
A i n
n
=
=
s a
E a =
_ e "
- .w L l C = ,
C =
_ d e o,
_ i x s E s .a W'
o r
_ D P e ,e I
Y3 X
,rA8
- - I O 0 3
1
'n a 4 2 .
0 K. 19 R -
E P r:o.1Y I g' .) llI 1B I ; (*: f, h1- ll F lp' !
? y E
l 1ll l
l gw; ,'s y< o m
p y.* g P. e m p m
p p p p -
z
. 1 C = o O o e P
U ,
Y 1 Vr ;e n
A Map i
- e m
L I e v i ;o m
- T e 4
E p W
u y
a y g.
L y. .
i.
D38 t
. s .
A W e
a e.w T N9 l 9- !
AA1 D r l
u F
- r1 i'
Gbe l" , ;
a s
YNsn -
Q. ,_;n v RIe Tv '
y-
- a e A e
T Sh r
c p REE h.qFma :., " ,
o i F
g
- MT-s t g
s i I .rpT R u t
% y;i H Lg s e
C NA Au T
l . .
a O
I n
o
- j l.
T i C
t c
n m
T; u
N F
- a U
F t
o H
4l E
$t. , ;ssu T )d. Am; eS eS s
O 5 4 3
2 1
8 W
- H 0 o 0 0 0 "
l li 1\llllll1ll
TDR 638 Rcv. O.
~
Page 40 of 48 g
LL/S k'
f a 41 3 ,,l 5 :
7- a w.g . 9
~8
. ./ -.
.. 2 og -- - aw. .. <
- 3 c .. -
0 l .#' h -,
l <
e '
-a
@# /
h *? t
?
I l
- c. k
~
.. . .I, l
- ,8 =
4 4 ,_
a Oc
- . " a Li (ppm), pH
.& h DR533 Rey, o
$ Page 41 of 4g l
I 9 c8g O O e l
-3 t -5 n'
m
- g -.
1 .
t
.s
-s
- 5
- -- i 12 5
- -
-R y i
n .
n =.
l .
=
r r ... .
O
-2 i n 8
h a 9 : i g
d
TDR 638 Rey, o Pa83 42 of 48
\
k 3 % '$
Li/5 o e e g $ I g I' E'
- s s -
-n g; *. - J ,.
. p, 1
g$
i j 'a .
. sf. 7 1
9 i s.1 h; .k .
4 .....
-sl k 9 . .
g 8
-l
$ . e 4**
, -5
=
- i a a O "
, s. " - , .
=
tt < ppnb pH I
l
\ . _-____________.nn.~~ nn ~n.en ne nn .on 1
ll d, u=
.;o o m
p o. g :w P a s
m p m p
p -
m
- 4 -
1 O
. C S m o D .
G
_ N I
,=
T ,=
S E ,.
T d e
n
,R R
i a P B
I D r A P a
,n A4 T C8 G
S T
O
,w 9
A R1 D -
- ,n l r n Y Dir i a
p ,.H E
_ R Np A R e C R
A l
A e
T A-B ,wM r u
1 g EPra y P C
,=
i F
MUu RYn R
~=
HAaJ -
- Y CL 1 $
.R A
0 .1 l
4- ,B R T p =UE E uy a ,=
F W L t
e ,=
w L l L l u
- U F ip Y
R A
F - - - -
~ -
a t
mu 5 4 s. 2 o. A 0 o 0 0 oJ ll( 1I 1llll1
1l 1' d" S
- m 7? o p 2T w* ,% 5 p
- 9 i hp44
. L e o a .
o o 0 0 0 0 s
= 4 1 n O
9 5
4 a
m R
I E A
P E E
R ,
e L
B I
! W P
A P E A -
TC8 AR9 4
r id ae E
1 pn D ei Ra E Dli BD r b i
YNrp I I H C
M RAA T -
EM R
A A
m u
O g mP, yU. E i
F HAa Y.n , $
H LJ C
An y
,g A T 1 g Ea u
n B p u
y m W wL a E
t e
L W 4 L l U
l j u t .
F c _
F u _
iwM
. 7 5 4 3 2
i aW Eg m
. 3 1 .ll
i' g=
a r
m n m gSO p p p p p .
- 4 -
1 o z C s o o a e P
U Y .- n A
L p
& , W1,G.
u T y a m E t L
(
W W e
A D l
l F
u e '
c;-
TN 9 O .' ,
nAW$
AA498 S .E9 9 D )
4 ..
G1 .-
YNe Q ,
.g or s
RIn TJu t
A
%,b T S-eel T F
e r
u g
4 MTi - i H 1 F r -
E p o\
1 R '
HLA e CAN ,.
O I ,
T 4 u , ,,
C -
lr N lo a U
F uf F
mr eat e /, qM .g, t pF seH yr 9 T SP
,1
/e O - - '
b],g..
H 6 O
A D
5 0
- a
- v. 0 0
llfI! ll '
,Il ll1lll)1ll1 l il .'
gw .M c - =I*
m p =E* & 2 &
p L
i hpl [i gb o a e P
o e
s n
o s.
, - ~ o s o U . _ - - - e s
Y A
L ms e'
, E s
- T E p uy
.l le LP I
4l1 E e
W a L
e E
D t
W e
\e n AN l E
T A A4 8 l
u F
. e D G19 . n- I E
e e E N b Y In Ne 0UJ 6-RTuJ 1
- , A e
TS- e r
n , J e u E E' T .
e i
F g
- MT .
E E L/
e HA -
A CN -
e O
I lF T e N
A T lH u -
C Fr o e e N s mf e
- e U te sr F ya
~' - Y y
Sp iA e
r
- M T P
% L4 O - ' -
o -
eG 5 sP 9 H o i
g q1 8
2 o A1 I ll l l1
TDR 638 Rev, o g,
- N Page 47 of Ag I
I I *
.' c88 o a .
-g
\*u E
p AA7 -3 gp. -2
-3
-e
- t
-R I
- -t b -.
=
-s ? II 8 -n f
-a I4 llg
-s e=
-e
-5
-3
-s
~E
=e
- 1 4
4 4 o R O o o a
(
IDR 638 Rey, c' g Page 43 of 4g A
A am&Ql I
.- A Li/s o e e h O
'$~! ' f . p
. -5 I
-5
-e
-t
-8
-I g
I2 I
g -R 5 -s
- 6. $ -*
. -- j
-3 4 h
-2 $
' -B L.
i .,
Il($
~
. i
~
-3
-3
-5
-R
-c
-e
- a=
g = = ,. .
n .a *3 Li (ppm), pH
I
\'
2 UNITED STATES OF AMERICA NUCLEAR REGULATORY, COMMISSION P':27 Before the Atomic Safety and Licensing Appeal Board 2 j[
In the Matter of )
)
METROPOLITAN EDISON COMPANY, ET AL. ) Docket No. 50-289-OLA
) (Steam Generator Repair)
(Three Mile' Island Nuclear )
Station, Unit No. 1) )
CERTIFICATE OF SERVICE This is to certify that copies of the foregoing Licensee's Brief in Opposition to Appeal of TMIA From Initial Decision and Licensee's Answer to TMIA's Motion to Reopen the Record were served by deposit in the United States Mail, First Class, post-age prepaid, this 14th day of January, 1985, to all those on the attached Service List.
/
EVANS HUBER DATED: January 14, 1985
q l
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION Before the Atomic Safety and Licensing Appeal Board In the Matter of )
)
METROPOLITAN EDISON COMPANY, ET AL. ) Docket No. 50-289-OLA
) (Steam Generator Repair)
(Three Mile Island Nuclear Station, )
Unit No. 1) )
SERVICE LIST Gary J. Edles Dr. James C. Lamb, III Administrative Judge Administrative Judge Chairman, Atomic Safety and Atomic Safety and Licensing Licensing Appeal Board Board U.S. Nuclear Regulatory 313 Woodhaven Road Commission Chapel Hill, N.C. 27514 Washington, D.C. 20555 Dr. W. Reed Johnson Mary E. Wagner, Esq.
Office of Executive Legal Director.
Administrative Judge U.S. Nuclear Regulatory Atomic Safety and Licensing Commission Appeal Board Washington, D.C. 20555 U.S. Nuclear Regulatory Commission Atomic Safety and Licensing Washington, D.C. 20555 Appeal Board Panel U.S. Nuclear Regulatory Dr. Reginald L. Gotchy Commission Admini trative Judge Washington, D.C. 20555 Atomic Safety and Licensing Appeal Board Atomic Safety and Licensing U.S. Nuclear Regulatory Board Panel Commission U.S. Nuclear Regulatory Washington, D.C. 20555 Commission Washington, D.C. 20555 Aaministrative Judge Docketing and Service Secticn (3)
Chairman, Atomic Safety and Office of the Secretary Licensing Board U.S. Nuclear Regulatory U.S. Nuclear Regulatory Commission Commission Washington, D.C. 20555 Washington, D.C. 20555 Dr. David L. Hetrick Joanne Doroshow, Esq.
Administrative Judge Louise Bradford Three Mile Island Alert, Ir'.
Atomic Safety and Licensing Board College of Engineering 315 Peffer Street Dept. of Nuclear and Energy Engr.
Harrisburg, PA 17102 The University of Arizona Tucson, Arizona 85711
~
k ASLAB Servics List
'f PEga Two TMI-l Thomas Y. Au Assistant Counsel Commonwealth of Pennsylvania Department of Environmental Resources Bureau of Regulatory. Counsel Room 505 Executive House P. O. Box 2357
'Harrisburg, PA 17120 o
i e
i
. - _ . _ _ . _ . _ _ _ . _ _ . . _ _ . . . _ _ _ _ . _ _ _ _ _ . . _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _