ML20093P018

Forwards Comments on ACRS 840523 Tour of Plant Site,In Response to 840515 Ltr

ML20093P018
Person / Time
Site:	Diablo Canyon
Issue date:	08/01/1984
From:	Mysinger E TENNESSEE VALLEY AUTHORITY
To:	Mckinley J Advisory Committee on Reactor Safeguards
References
ACRS-CT-1761, NUDOCS 8408020242
Download: ML20093P018 (4)

Text

.- t 27- /96/

TENNESSEE VALLEY AUTHORITY KNOXVILLE. TENNESSEE 37902 -

400 West Summit Hill Drive, W9C165 gj gy /[

R .C.* p r.- j p. _,

f: . ..,.;,...

Mr. John C. McKinley, Chief , J j \ ,.. p s.;. .. .

Projec t Review Branch #1 7.I p.,

tN%C, ACRS 3 5 a,!C 1'.

1 ..t-

. "< #* ' 8;*:#8. N G Ua :.h ' og ton , D. C. 20555

Dear Mr. McKinley:

In rc aponse to your letter of May 15, 1984, I accompanied members of the ACRS cn a tour of the Diablo Canyon Nuclear Plant on May 23, 1984. As requested by your referenced letter, I observed the installation and sizing of the pipe support / restraint systems. Comments on what I . observed are enclosed.

I am very appreciative of the opportunity to make the tour with you.

Very truly yours, TENNESSEE VALLEY AUTHORITY

(,(0.fhj4 0 k

E. D. Mysinger, Principal Mechanical Engineer Civil Engineering Support Branch

-Eiclosure DESIGNATED ORIGINAL Certified By _/# _,, ATTACINENT

~

8408020242 840801 PDR ACRS t CT-1761 PDR An Ecual Opportunity Employer ,

'J.

Disbio Canyon Nuclear Power Plant Pipe %ppart/ Restraint Observations e i> j E. Douglas Mysinger s

~

On May 23,1984, I accogenied members of the ACRS on a tour of the Diablo Canycu Nacicar Power Plant. As requested by Mr. John C. McKinicy, Chief, Projec t Review Branch No.1, I obser'v ed the installation and sizing of pipe e

supp or t / re st r aint sy stems . The purpose of this paper is to connent on what I observed.

On the tour Mr. Isa Yea, 1 NRC inspec tor previously assigned to the site, pointe

• ou t typical exampics of concerns he had documented.

Concern No. 1 - A snubber had been placed on a small branch line relatively close to the run line. Thernal movements were not su f ficient to justify a snobber and seismic movements were not as large as the snubber free travel.

Thus, the snubber would not provide the support as modeled in the piping ana ly sis .

Response: The small branch line had initially been qualified by

• conservative span tables (alternate analysis ru le s) . The designer had prudently specified a snubber adjacent to a valve to accomodate relative thermal movement of the run and branch line and to provide seismic support of the valve. PG & E has subsequently performed a computer analysis of the line and it is qualified with or without the snubber. There is no safety concern for leaving the snobber in.

Concern No. 2 - A spring and snubber and two rigid supports were very close together near a valve. The spring and snubber could not be ef fective due to the close proximity to the rigid supports.

Response: PG & E had run the problem with and without the spring and snubber and the pipe was qualified. At one time during the design of the plant, an economic decision was made to leave the supports in place.

Concern No. 3 - A rigid and snubber are close together. The snubber cannot be ef fective because of close proximity to the rigid and inherent free travel of the snubber.

, Response: If the rigid support doe s not deflec t enough to redistribute load to the snubber , it cannot be overloaded.

Concern No. 4 - A snubber was attached to a valve operator. An sufficient analysis without- the support indicates movement of the operator was not to lock up the snubber. A strut should be specified. ._ 1 Rigid supports in the run Re sponse: A strut has very little free travel.

l line near the valve are designed with gaps. A fixed support point on the l valve ' operator and a gap in adjacent rigid support points on the pipe could potentially overload the operator. Concrete creeps and shrinks for years, pipe shakes down during the first few cycles of operation, etc. These 014167.13

'4  !

g things are not considered in a c omputer analy sis rad resulting movement ~~

calculations are not indefinitcij ccurate. I f the snubber and strut were comparable in reliability and maintenance, c nmbber would be a cicar choice for this application.

Concern No. 5 - Calcium silicate insulation with w tal cover has been installed on relatively large pipe without su f f *t Ient clearance to avoid impac t with the building struc ture during a seirmic event.

Response: Building ctructural steel nembers were obviously suf ficient to crush the insulation or withstand the seismic Icading transmitted by the pipe through the insulation. It is reasonable to eypect that c rushing of the insulation will increase dynamic damping in these pipe runs which will reduce stress in the pipe and lead on adjacent sopports.

Concern No. 6 - There are too many snubbers in the plant. Examplas were sited of plants that have removed hundreds of snubbers. The expressed concern was for radiation exposure to personnel during inspection and maintenance of the snubbers.

Re sp on se :

It has been difficult for experienced piping designers to specify rigid supports that reduce flexibility of piping systems. This desire to maintain flexibility has resulted in the use of snubbers where a rigid support would qualify. As indicated by NUPEG/CR-3718 (Reliability Analysis of Stiff Versus Flexible Piping - Status Report) reliability of rigid systems is still being questioned. It is also apparent, industry wide, that inexperienced designers specified an excessive number of supports including snubbers. However, meaningful relief such as higher damping, elimination of 1/2 SSE as a design consideration, spectrum peak broadening changes, etc. , is now being considered. Removal of snubbers considered in design of the piping is an economic and not a nuclear safety consideration. Many f ac tors enter into the economic evaluation such as age of snubber, type of snubber, operating experience, pending changes in industry practice, etc.

Concern No. 7 - It is industry prac tice to specify a 1/16-inch gap _ between a pipe and rigid structural steel type support. A 1/16-inch gap on each side plus a reasonalle tolerance of 1/16-inch can result in a cumulative gap of 3/16-inch. When two supports are closely spaced and the majorthe part two of the 3/16-inch cumulative gap is on opposite sides of the pipe at supports, load distribution to the supports may not be equal.

Re sponse: PG & E is shimming supports to address this concern. To _

expedite licensing, this approach seems prudent. However, unless nuclear -

power plants are extremely overdesigned for such an unlikely event an an SSE, it is reasonable to expect deformation to redistribute load through a 3/16-inch gap. For noreml operation, the larger gap is preferable.

___ _$04167.13

In sur'ey, taking out supports that are r.ot required but have been considered in the piping qualifica tion is an ec onomic c e".t i 1.! rat ion.

Changi::; co t wubbers with struts is nn eccnonic considera tion. If a _.

sy s tem i.J so conserve: Lively nupported that tavver:ent will not he sufficie^t ,

4 to leal up . caubber, there is no safety concern. Sc.ubber s on valve opera tor s ver sos struts are pre ferred by corac designet s to casure against load s <ite to normal opeia tion. k'ith the possible exception of concern No. 7 thare was clarly no valid sa fety problems obsctv d during the tour.

PG & E I. n.cdifying support gaps in response to concern No. 'f to avoid fu rthe r delays .

O em 014167.13