ML20126J606

Forwards Proposed Response to M Lewis Re Safe Operation of Repaired Steam Generators

ML20126J606
Person / Time
Site:	Crane
Issue date:	05/11/1985
From:	Johnston W Office of Nuclear Reactor Regulation
To:	Lainas G Office of Nuclear Reactor Regulation
Shared Package
ML20126J603	List: ML20126J606
References
NUDOCS 8506100646
Download: ML20126J606 (6)

Text

.

,/'p *,c,o, o

UNITED STATES

['

~ 'g NUCLEAR REGULATORY COMMISSION

p WASHINGTON, D. C. 20555 a

/

NAy 111985

....+

MEMORANDUM FOR:

Gus C. Lainas, Assistant Director for Operating Reactors Division of Licensing FROM:

William V. Johnston, Assistant Director Materials, Chemical & Environmental Engineering Division of Engineering

SUBJECT:

RESPONSE TO LETTER FROM MARVIN LEWIS Enclosed is our proposed response to Mr. Lewis' letter of Aprik16,1985.

T If you have questions, please contact Conrad McCracken at ext. 28595.

ylxI i L =.v-V

• N 'Y William V. Johnston, Assistant Director Materials, Chemical & Environmental Engineering Division of Engineering

Enclosure:

As stated cc:

J. Knight H. Thompson J. Stolz V. Benaroya C. McCracken

0. Thompson J. Thoma P. Wu 8506100646 850530 PDR ADOCK 05000289 P

PDR

UNITED STATES e'S 3 p,

NUCLEAR REGULATORY COMMISSION 1

r,,

aj WASHINGTON, D. C. 20555

-4 4

a., -

,c b

Q e...+

Marvin I. Lewis 6504 Bradford Terr.

Philadelphia, PA 19149 I

Dear Mr. Lewis:

I I have been asked to respond to your April 16, 1985, letter to the Comnissioners regarding the safe operation of the Three Mile Island, Unit 1 i

(TMI-1) repaired Steam Generators (SG).

I In your letter, you request that TMI-1 not be restarted because the fiRC and the licensee have made errors and omissions in their analysee of the SG 7

corrosion problems and that these will directly endanger the public health b

and safety.. You state that the patchlike indications which have been identified as Intergranular Attack (IGA) caused by sodium thiosulfate, may 7

instead be concentration cell corrosion. You also note that the observations at TMI-1 better fit concentration cell corrosion than IGA or Intergranular Stress Corrosion Cracking (IGSCC).

Your concern that the TMI-1 sulfur induced IGA of the SG tubes may be related to concentration cell corrosion by some unidentified contaminant is based on some incorrect assumptions.

Regarding ycur statement that the corrosion is occurring in crevice areas, as we stated in our safety evaluation (14UREG-1019) page I, parcgraph 1, the TMI-1 SG corrosion is on the primary side (ID) of the tubes. The primary side of the tubes does not have crevices. The SG design is discussed in more detail on page 2, section 2 of fiUREG-1019. All tube supports, including the upper and lower tube sheets contact the tubes on the outside (OD) of the tubes.

Recardirg your statement, that there is no physical evidence to suggest the patchlike areas were produced by the original thiosulfate contemiration, the record clearly documents the presence of IGA prior to fiovember 1984, when eddy current testing (FCT) showed enhanced detectability of IGA patches.

In the licensee's technic 1 data report (TDR-341), dated August 27, 1982, IGA on tubes removed from the So is discussed. The staff motion for summary disposition of joint intervenor r.ontentions, dated February 24, 1984 discusses IGA in TMI-1 SGs on pg. 5.

Additionally, the licensee's summary disposition motion, dated February 24. 1964 discusses IGA on pages 58 and 59.

Further, a review of ECT data prior tn November 1984, showed the presence of IGA.

b l

i p

Marvin'I. Lewis,

n

[

With regard to the potential safety consequences of the recently identified t.

grain dropout from.previously existing patches of IGA, the issue was addressed in our response to the April 5, 1985 letter from the Union of Concerned Scientist dated May 10,.1985.

In the response, we stated that tubes with IGA defects exceeding the plugging limit defined in the plant Technical Specifications-have. been removed from service by plugging. Any tube degradation smaller than the plugging limit is not expected to rupture or.

c to result in a leak because the defects themselves are not substantially influenced by axial. tube stress caused by a cooldown.

The discussion on page 2, paragraph 1 of your letter provides a good text book sumary of some factors involved in concentration cell corrosion.

However, the factors which need to be considered when making a determination

.as to-the actual type of corrosion that exists in a specific case were not listed. To make a final determination of a specific corrosion mechanism,

+

many factors such as material, environment, physical configuration (such g

as crevices), and microstructural evidence must be considered.

On a microscopic basis, the patch like degradation at TMI-1 was present and clearly identified as IGA on some of the twenty-one tube sections which were removed from the TMI-1 SG and examined. On the tubes which were examined grain boundary attack and concurrent grain dropout were visible.

Concentration cell corrosion produces localized dissolution of the base.

metal.end is clearly differentiable from the IGA which occurred at TMI-1.

Additional factors considered in determining that the IGA is r.ct progressing include:

1)

The licensee has maintained an adequate reactor coolant chemistry control and an effective lay-up procedure to mitigate initiation of new corrosion attack or propagation of existing IGSCC.

Further, no conteminants have been identified that would be associated with concentration cell corrosion.

2)

Long-term corrosion tests demonstrate that new corrosion would not be expected and propagation of existing IGSCC should not have occurred in the TMI-1 water chemistry ccrditions in existence since 1983.

3)

Periodic, preplanned, ECT of a small number of tubes (approximately

20) containing known IGSCC, at less than the plugging limit, has shown no. measurable growth of IGSCC since 1982 when it was originally tested. These. tubes are scheduled to be retested, to monitor for growth of IGSCC, every time an ECT is performed.

Regarding your concern that the steps used to control thiosulfate contemi-notion can also cause or promote other corrosion processes. The purpose of the pcroxide cleaning of the TMI-1 OTSG tubes and the RCS was to convert intermediate metastable sulfur species, such as the sodium thiosulfate, volatile polysulfur and sulfide into stable non-corrosive sulfate (NUREG-1019, page 27). No corrosion was found in tests performed before the cleaning in

Marvin I. Lewis,

which actual TMI-1 tubing and simulate solutions spiked with concentrations of up to 10 ppm sulfate were used (Licensee's Motion for Summary Disposition of each of TMIA's and Joint Intervenors, Contentions, February 24, 1984, page 100). This is a factor of 100 above the current 0.1 ppm sulfate limit.

Your concern regarding the addition of lithium hydroxide as a base additive at TMI-1 may enhance the potential for galvanic corrosion is not supported by experience. All Pressurized Water Reactors (PWRS) operate with lithium hydroxide as a base additive to the reactor coolant. The concentrations of lithium hydroxide at TMI-1 are not different from other PWRS. No instance of galvanic corrosion due to lithium hydroxide addition has been reported from any PWR.

Furthermore, the addition of lithium as a base in the reactor coolant would decrease the propensity of sulfur-induced IGSCC as demonstrated by the staff consultant (NUREG-1019, Attachment 2).

Even if contrary to the facts, were concentration cell corrosien takirg

~;

place by some unknown mechanism, it would not constitute a significant public health and safety issue. Concentration cell corrosion is a slow process

.1

'?

when compared to IGSCC. Therefore, the measures taken to ensure public health and safety for the IGSCC which is present would be even more conservative if concentration cell corrosion were active.

Further, regardless of the corrosion mechanism, the steam generators have been repaired to their original licensing basis as discussed in more detail in my letter of May 10, 1985, to representatives of the Union of Concerned Scientists.

I hope this letter answers you concerns regarding the cerrosion aspects of the TMI-1 steam generators.

Please feel free to contact Mr. Ccr. rcd McCracken (301 l,S2-0595) of my staff if you want to discuss this subject further.

Sir.cerely, harold R. Denton, Director Office of Nuclear Reactor Peeviction l

l

L/

Kamer y

',4,,

y UNITED STATES g

NUCLEAR REGULATORY COMMISSION y,

E WASHINGTON, D. C. 20555 EDO PRINCIPAL CORRESPONDENCE CONTROL FROM

DUE: 5/-24g EDO CONTROL: 000575 DOC DT: 04/16/85 U S J-3[

M.

I. LEWIS FINAL REPLY:

TO:

CHAIRMAN PALLADINO FOR SIGNATURE OF:

GREEN SECY NO: 85-348 Denton DESC:

ROUTING:

ERRORS & OMISSIONS IN BN 85-028, "RECENT EVENTS MURLEY.

RELATED TO TMIr[. STEAM GENERATORS" GGUNNINGHAM TAYLOR DATE: 04/26/85 ASSIGNED TO: NRR CONTACT: DENTON CPECIAL INSTRUCTIONS OR REMARKS:

FOR APPROPRIATE ACTION received NRR: 4/26/85 contact:

Thompson cc: Eisenhut/Denton bhL PPAS M

,,a4

~ CORRESPONDENCE CONTROL TICKET SECY N!JMBER: 85-348 LOGGING DATE:

4/24/85 0FFICE OF THE SECRETARY i

ACTION OFFICE:

ED0 AUTHOR:

Marvin I. Lewis i

AFFILIATION:

Phila,PA LETTER DATE:

undated FILE CODE:

ID&R-5 TMI-l A00RESSEE:

Chm, Cmrs SUSJECT:

Errors and omissions in BN 85-028 d

ACTION:

Appropriate

.....%. 5Dh[.f,......

DISTRIBUTION:

"," * * * *

• g..g* * * * * * * * * * '

SPECIAL HANDLING:

None

$1GNATURE DATE:

FOR THE C0petISSION: Champ i

e EDO --- 000575

M. l. LEWIS tv

- 'p Y

7-M BRADFORD TERE

~

i ',

PHlUC, PA.19149 Chairman *allidino Commissioners Roberts, Asselstine, Bernthal, and Zech Eirs:

.This letter deals specifically with errors and omissions in Board Notification 85-028,"Recent Events Related to TMI-1 Steam Generators. " These omissions and errors are 'of a degree and of an importance that they will directly endanger the health and safety of thrt public if TMI#1 is restarted.

Background:

The SRC has agreed with the Licensee, GPU/ Met Bi, that the corrosion of Steam Generators at TMI#1 relates to the addition of thiosulphate contamination of reactor coolant.

Repairs have progressed designed to obviate thiosulphate induced intergranular stress' corrosion. cracking. Some.of these repairs have not succeeded in producing

'the hoped-flor results, and the Board Notification d5 a3 spells out some of the

• - remainirig problems.

One' of the problems is trat the tubing has shown'some indications that were not

'previously observed with eddy current testing. The ~icensee has characterized the

, new" indications as " enhanced detectability.of pre-existing areas of ICA."

Furthermore,

,.the Licensee defends his supposition of intergranular attack with the statement

'"pttchlike -indications suggestive of IGA were seen by the fiberscope examination."

~0 missions:'

" Patch-like ind6 cations" are not only suggestive of IGA induced by thiosulphate contaminations.

They are also suggestive of other processes which are ongoing and may actually be accelerated by the very steps used to control thiosulphate contamination. Ato whether the original attack came from thiosulphate of not, the steps used to contro1L the 'thiosulphate contamination leads to a condition ripe for other corrosion processes.

These are two situations completely omitted from the Br 8$-028 and previous.

corresponden6c efconcerning the steam generator tube problem:

1.Patchlike indication. used to justify the conclusion of it* *preex1yt&ng 9,, areas of IGA" can also indicate other corrosion processes, and

2. Steps used. to control thiosulphate contamination can also cause or

• j promote other.' corrosion processes.

An example:

k 8

5 L..

X 2.

'c o

An example :

g 7

"The type of corrosioniassociated with surfaco crevices, scale and other rfaces deposits, is known as" concentration cell corrosion." The types of cells most frequently encountered in the stainlesssteels are the oxygen and

~

active:pasalvo calln. Oxygen cells will develp into active: passive cells when the re-establishment of passivity in the anodic area becomes impossible. Because of the Lsrge areas of pasive metal that normally surround the anodic areas

[

[the active"pascive cells are extremely energetic, cince the currounding Exygen serves as a dopolarizer by reacting wi th the hydrogen from the reaction.

7*

Solutions that contain chlorides are nost active in promoting these cells, ani

(

severe pitting may deve.lcp from their action. "

r Page 559 AeM Metals Handbook 03pyrighted 1 @ 8 :1958 Edition.

t The above write up of " concentration cell corrosion fits exactly the observations in the nteam generator 'orrosion problen.

In fact, the obsersations f

fit the " concentration' cell corrosior." much better tha n they fit the thiosulphate 100C0 theavy, Please note how the obaervations in the BN 85-028 fit the Oconcentration cell' t.

phenomenon:

1. The corrosion and cracking occurred near or in crevices.

Concentration c..

cell corrosion is also called cryise corrosion because it occurs in or near 1

crevices..

[

2.

Once ain area is depassivated or activated by oxygen, concentration cell corrosion.is likely due to large area of cathodic or passivated area.

Only

(

small ares were attacked leaving a large passive area. This also fita concentration f

cell corp ion. The attack appears to be continuing as would be foretold with

?

concentration cell corrosion since the large uncorroded, passivated areas h uld make repassivation of the anodic area difficult. Also the recent FEk indications would nuggest that the corrbnosion is continuing rather thaa g

q'pe,chanical forces are causing previous intergranular attack to allow grains to

' fall out or separate.

t t

3. Active sulfur compounds would have the ability to depassivato small crevices, The lithium and other additions used to control residual sulfur compounds would

^

t I

also be electrolytes that could reduce the coolants electrical resistance allowing

.[

easier galvanic current transport in the electrolyte to promote concentration 5

cell corrhion.

Inotherwords, the very steps used to control the thiosulphate contaminat.lon may lead to worsening )f the contentration cell corr #sion.

A s

b r

F F

t cg 3

4..: The pastch like~ areas described in BN 85 028 on page 20 suggest. galvanic or concentration cell corrosion..

There is no physical evidence to suggest that W6-the ar y iginally produced.by thiosulphate contamination and even if the thio-sulphate contamination were the original source of the corrosion,-concentration cell corrosion could now be the controlling corrosion process.

Raeuest:

~Under.certain rules of the NRC,.I respectfully petition that the TMI//l reactor

. not be restarted. The above scenario and exposition of corrosion problems in the steam generator tubes show that there 31s a -'

great potential for harm

. to the public if the reactor is allowed to. start.

Very truly yours,

",A.

'J euwt wta) 2 /5 289 5964 M. -1.. LEWIS M 88f 6504 BRADPDRO TERR.

PHRA, PA.19149 l

H O

F s

4

\\

-~ as c-

-x4 v/

m..

• ~ ^ D..

8C 9

.(3 i m igi 1

~ -+:

1.,.

IG hP3

" ~ ~ * ",

'c r

~__

l Citizen Action Nk ArtysA22 -

~~

in the Northeast l

M. I. LEWIS 6504 BRADFORD TERR.

~

Chairman Pallidino PHILA., PA.19149 e

trNRC washington, D,C. 20555 O

I e

9 Il$$ I $$ I I$ $ I$ $

I w--..-.

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

.-.i

- -