ML20196D122

Comment Opposing Proposed Rule 10CFR50 Re Leak Before Break (LBB) Technology.Requests Not to Extend LBB Technology & to Retract Present LBB Use Because LBB Failed to Predict Pipe Ruptures

ML20196D122
Person / Time
Site:	Three Mile Island
Issue date:	07/01/1988
From:	Lewis M LEWIS, M.
To:	NRC OFFICE OF THE SECRETARY (SECY)
References
FRN-53FR11311, RULE-PR-50 53FR11311-00003, 53FR11311-3, NUDOCS 8807010312
Download: ML20196D122 (4)

Text

DOCKET NUidBER no I PROPOSED RULE EB 53 FR H3n ocacito 3 SNRC Marvin 1. Lewis V801 Roosevelt 81vd.#6a 2 3 27 P3 30 Phila., PA 19158 0FrlC: /Et Y%

becretary of the Commis*lon 00CKCi g y UbNRC Washinaton, D. C. BObbb

Dear Mr. Secretaiy:

Please accept the following letter as my comments on 10LFR Part SO Leak Betore Break technologyi Solicitation of Public Comment on Add 1hlonal Applications. (Federal Register Vol. 53 #66 4-6-88 Pages 11311 & 1131d. )

Introduction:

Leak before break technology has allowed a reduction in the restraints used on piping in nuclear power plants. ihese restraints or snubbers had tailed and continued to tall many design and quality control criteria. The snubbers were otten in inappropriate places, lhey otten failed testing to nimulate earthquake conditions. Getting rid of the snubbers may have actually improved the design and operation of some nuclear power plants, but not t'o r the reason stated in the final rulc daoed October 8V,1Y87.

betting rid of the snubbers allows designers and operators to design powerplants more realistically. Operators and designers cannot depend on snubbers to provide earthqueke protection, lhe insult is that the design may actually wind up sater because the snubber s ar e no t depended upon to do a job to which they are incapable. Since snubbers could not perform their safety '/unct ion adequately , the NRC used the ruse to elim3nate the need for nnubbers. lhe r use which was used to eliminate snubbers was o non-technology called Icak betore break. I have been commenting upon this non-technology for many years, but i shall review some of my comments to rami 11 art:c those who have not had prior experience with this non-technology.

Ihc idea behind leak betore break seems logical. Leak before break technology assumes that a detectable, and detected, leakage will precede a double ended , guillotine pipe break during reactor oper ation. Since the assumption is that leakage will occur and be detected betore a double ended pipe break, the design need only require detection ot leakage. The design does not need to withstand a double ended pipe break while the reactor is operating. Leak bet' ore break sounds logical, but it ta11s every test that LBB should logically meet.

l 8807010312 080701 PDR PR PDR SO 53FRll311 i

'W 3S-/D

(-

i Uhat-tests should iBU meet to be usef ul?

LHB assumes that a detectable and detected leak will precede-

~a'doublo endeds quillotino pipe break.

.l. LBB is an assumption. Ihcre have been no tull scale, in reactor, operating test to prove this assumption. Even atter L88 has been adopted as part of the rules in bDC 4, full scales tests are not tunded or contemplated. This sort of tull. scale testing should have proceeded the adoption of Lub into the rules.

a. L88 technology assumos that a detectable leak will precede a doubic ended, guillotine pipe break. lhroughout reactor history, leakage measurement of coolant has been a subject ot disagreement in method and results. There are a thousand ways to tudge or cheat on those numbers.

Iwo excellent examples of incredibly poor measurement of coolant level inventory have just como out this month.

Measurernent of coolant level inventory over time provides the leakage rate. Poor measurement of coolant level inventory results in meaningless and. wrong leakage ratec.

l'he first example is reported in NRC Information Notico No.

80-36: Sudden Loss of RUS Inventory During Low Coolant Level Operation. Entrapped air and other problems lead to incredibly inaccurate RCS inventory errors and subsequent pressurinations, this is exactly the scenario needed for a double ended, guillotino pipe break. LBB would not be predictive in this situation, and a double onded pipe break would ensuo. At the same time, the rules do not require protection in the design from the consequences of a double onded, quillotine pipe break. NRC Information Notice 88-36 refers to Uencric Letter 8'/-18 which lists 37 ovents attributed to low coolant levels many related to poor measurement. Ihlr first exampic shows that measurement is historically poor and unreliable. Depending upon measurement to detect Acakages contradicts historical experience and logic.

The-second example of why detecting a leak is unlikely relates to the recent decision on IMIME. the Memorandum and

Ordor, CLI-OH-Ud , Inquiry into the fMId Leak Rate Falsification, states , "The Board found that 50% or more of the tests were

discarded." Page 3.(Docket LRP dated 28 April 1988.) No matter what the technology, it halt the test are discarded, leaks will not be detected in a timely manner to stop a double ended pipe break. . Leaks have been. ignored in the past. Leaks are still being ignored. Leaks which are ignored cannot warn of an impending

! double ended guillotine pipe break. Leak before break technology

' cannot war n ot double ended pipe breaks.

l

)b

3. Leak before break technology assumes that a pipe remains as it was designed. the pipe does not romain as it was designed.

Some of these changes.are minimi:cd during operation. The loss of pipo thickness ta ec,rosion is minimized by addition of corrosion inhibitors.such as hydrogen and all volatile acidity control.

Pipe walls do thin out for other reasons than corrosion. Some of these other reasons have just becomo evident. (Summary of Responses to NRO Bulletin 87-01, "Thinning of Pipo Walls in Nucicar Power Plants.") The thinning is a newly observed phenomenon, and was not included in the design basis. The result of the thinning was a pipe break at the-Vepco Surrey plant which killed tour workers. the gig _e did not leak before it broke.

4. the Advisory-committee for Reactor Sat oguards has branded this leak bef or e break technology as a "dclusion." AI.RS has stated. "We know of no way to demonstrate its validity."(

Uroundswell, Nuclear Information and Resource Services Winter Su Page 6.)

S. Other situations have caused piping to rupture without leaking. (NHC Information Notice No. 88-13: Water Hammer and Possible Piping Damage caused by Misapp11 cation of Kerotest Packless metal Diaphragm Globo Valves.) Great amounts of inf ormation have recently surtaced concer ning improper materials, impropor markings, improper valves, and improper testing tor severe accidents. For brevity, I shall not put down all the retorences. NRC.Intormation Noticos and Generic Lotters contain most or these references going back to 1984 to the present.

Deviation trom the design basis is a long and hallowed tradition in the nuclear industry. Leak betore break technology depends upon the design and the actuality being fairly closo. the design and the as~ built plant have many ditterences . These

~

dit to coces ar c signat icant and f ar reaching and invalidato any assuinp tion r equired by leak before break technology. Leak before break technology has lost any validity to which it ever aspired

due to these many datterences.

Reason to retract , and not to extend, the uso of leak before break technology.

the procccdsog discussion provides substantive reasons to retrac t the use of leak before break technology. Loak before break technology is a delusion. Leak before break technology has tailed to predict pipe ruptures. The failure has cost lives. Leak before break technology depends upon materials and tests which have talled to moet- any criteria of honesty. Inferior materials have bcon subststuted and tests have been discarded. Leak before break technology is a dangerous delusion.

?AA

i 1 H 4 ...; .

q

' Spec i t. i c s -ai:, to.why ieak before. Break shottid not be' extended.

1.. Containment: the containment operates at pressures lower than those it.util sec during an accident'. the containment is usually at ver'f low piossure. During the TMI#2 accident, the containment saw over 27 psi-due to a detonation. Calculations in support of NUHEG-1150 have shown that even higher pressures.are possible from hydrogen detonations'in an accident. Leakages, which would

.be unobservable at normal.. test pressures, would lead to

. containment tallure at these higher accident pressures.

a. Emergency Coio Cooling Systec .Ouring an accident, this system would be challenged repeatedly. .it a leak developed , the ECUS could no t be turned ot't for repairs during an accident. . fhis would be a situation wherein even it leak before break technology worked p oi .t oc t } y , the result'would be academic. The opportuni ty

- would have past wherein the pipe which war. Joaking could'be

.repaised betore catastrophe struck. The ECCS could not be tur ned ott'during an accident'because the-EL'US is needed to mitigate the accident. Catastrophe-cold strike botore the ECCS could be.turnad

-ott to allow access to fix the leaking pipe.

3. ^ t'nvi ronmental Qualification: The discussion ot' what changes in EU due to LBb does not appear in the RFO. Please put in enough discussion in plain'English'to lead to coniments.

Please do not extend the use of Icak before break technology.

' Flease retract the present use of LOD.

Respectru1]y submitted, A .

(

,f

. 'WhhW

/ aivin 1. ' t .c al s

^ .

l 7801 Roosevelt Blvd.#68 Phila.. PA 1915d 4 (Bib) 684 16Y4 a