ML20149M444

Disapproves Rev 0 to Procedure 4000-IMP-3891, TMI Unit 2 Recovery Polar Crane Operations. NRC Will Consider Reducing Review After Experience Gained

ML20149M444
Person / Time
Site:	Three Mile Island
Issue date:	03/10/1983
From:	Barrett L Office of Nuclear Reactor Regulation
To:	Larson L GENERAL PUBLIC UTILITIES CORP.
Shared Package
ML20149M377	List: ML20149M375 ML20149M390 ML20149M420 ML20149M434 ML20149M444 ML20149M457
References
FOIA-87-728, FOIA-87-853 NUDOCS 8802260049
Download: ML20149M444 (1)

Text

-~

~

,. - .y -

ATTAC M NT 5 (

- D!$TRItuffon:

ma mast r/r i TN! HQ r/f I BJ5nyder j - . 14 arch 10.19&3 l .

l

?. l i

t;CTE To: J. Larson .

subject: Disappmval of TM!-t Recover / Operations Procedure .

Mc. 4000117 3831 (Rev. 0) 711 Unit 2 Recovery Operations Polar Crane Operation' i

j W have rsviewed the subject procedure, and we are disapproving it for the

followir.g reasons

4 s. It contaics several references to the auxiliary hoist / book. This portion (

of the polar crane was never suositted to us for use and never approved by us for use in the Recovery Operations Program. Therefore we will not '

ap;,reve a procedure that permits its use.

2. It does not specify that. alt rigging used with the polar enne shall be '

cartified/ tasted as secWg applicehle /JtS! standards.

3. It does not require a osality-control-type check of all rigging for loads i in ancess of tas tons.

4. It does W not s pecify the minipine-aseight load that requires a brake test.

reewveee<

ten tons. l

_ i

5. The procedure should refersnca en engineertog drswing for specifics on the l liuitations of crar,e *, rave). (An alternative would be to specify the no-trave.1 portion of the crane arc on Dhtbit 1.) l

6. It coes not refere'nce a gcVerning precedure (er other dccutwnt) to specify the individual who shd11 be in charge of the evere11 rgeration when l 11ftins Particular loans. ,

until you have atsumulated a history of satisfaci.ory performance with the i refurbished crane, the NRC will interpret Technical specification 6.8.: as  ;

requiring NRC prior approval for WI's or procHures. After experience is ,

pained, we will considar reducing our review.

, +

/u

• 8802260049 080219  ;

PDR FOIA Lake H. Barrett CONNOR87-720 PDR Deputy Program Director &p TMI Program Cffice H , ,

cc 8. Ennga  ;

L. King -

1

- J. Thatsing TMI:Po TMI:P0 {

E' 'I'"' L3 age:wa LHlarrett .

j 3/ /53 3/ /83

- - - ~ , - - - - _---..n, - - - - , _ - . - - , . , . _ , - . _ . . , , _ _ , - , - , _ , , - , _ - - ,-_-,..,.-,n ,_ , -,

\ *:

ATTACHMENT 6 umrus erAres

, ,,(l 'g\ NUCLEAR Rs00LATORY C00eelts@d nameseston. m. c. asses April 7,1983 a '

k*...e) - ,

NOTE TO: James E. 1. arson, GPU .

FIO: Lake H. Barrett, Deputy Program Director TMI Progren Office DISAPPROVAL OF POLAR CRANE TEST PROCEDURE, 4370-3891-83-PCI I

SUBJECT:

My staff has reviewed the subject procedure that GPU submitted to the NRC on March 31, 1983. We have reviewed the technical aspects of the procedure in areas other than those pertaining to the Parks /Gischel affidavits. These issues are being handled by the NRC's Office of Investigation. We are disapproving the subject procedure for the technical reasons discussed below.

1. The GPU safety evaluation states that rigging components have a safety factor greater than 5. Our review of the rigging for the 40-ton missile shield (Bechtel drawing 2-COP-1302, Revision 0) states that the wire The liftrope of slings the missile 'shall shield have aisminimum a 4 pointbreaking ljft. strength of 66.1 tons.It appears that The safety factor would then all four slings are be 5.7, which is acceptable. statically detaminant.However; based on information availa cannot. conclude that all 4 slings;will have the sase load, i.e. sling length tolerance may make this a statically indeterminant lift, and therefore, each sling may not have a safety factor of 5.

Our review of drawing 2-COP-1301, Revision 1 for the load test assembly indicates that you may have made a similar assumption regarding the (8) 1 3/4" wire ropes supporting the full load on the load spreading frame.

l Based on infomation available we cannot conclude that all eight wire l

ropes will have the same load; in other words, that this is also a statically detaminant lift. Therefore, each wire rope may not have a safety factor of 5.

An additional concern in this area is that the load test assen61y drawing

' (2-COP-1301, Revision 1) does not provide tolerances for the placement of the missile shields in the load test frame nor any tolerances for the construction aspects of the frame that also influence uneven sling loading.

2. The procedura references multiple Bechtel drawings that contain safety

! infomation, such as specifics for rigging components. Since the procedure does not stand alone, these drawings must htye the same administrative control, e.g. the same safety reviews as the procedure

itself. In this case, it would include NRC approval of drawing revisions.

4 Please describe your administrative controls to assure the above.

I e w p #.

l

.\..:

2 M 7e N g.E; Larson

3. The procedure should clearly state not only load path horizontal Specifically, limitations but also load path vertical limitations.

Sections 9.5.2/9.5.3 should state that the first brake test should occur at an elevation less than that calculated to structurally fail the floor.

In addition, Section 9.3.1 should clearly state that any missile shield being moved over another missile shield should be at an elevation less than that calculated to cause structural failure of the lower missile shield should the upper missile shield fall. -

4. The hook must rotate freely to avoid non-vertical force components being applied to the load cell. The procedure makes no mention of this issue.

Riggers and operators should be alerted of any resistance in the rotation of the hook.

It is not clear to us if the reactor vessel head lift wi'.. involve rotation of the hook. If the reactor vessel head lift could involve hook rotation, then further assurance of free hook rotation under load conditions in excess of the head lift weight will have to be provided.

5. Sections 5.1 and 6.1 do not clearly state how to interpret the readings from the load cell. For exagle, 5.1 states that the calibration range is from 40,000 to 400,000 pounds, whereas' it should actually state 40,000 to The 440 000 pounds as indicated on pages 23 and 24 of this document.

t procedure does not state that a 400,000 pound reading on the load cell correlates to an actual 440,000 pound load. We suggest a table for the operator that clearly correlatas. load. call readings to actual loads.

6. The procedure should clearly state that areas such as the incore instrumentation terwinal plate should be avoided at all times by the We suggest the drawing be amended as loaded or unloaded hook / block.

necessary to implement this and other load path restrictions. This coement also applies to your general operating procedure, 4000-IMP-3891.01.

7. The procedure does not address any measurement of electric current being drawn by motors during the testing. We reconnend that current measure-ments be taken during load testing to provide additional information on electrical performance. These measurements could be taken outside the reactor building to minimize occupational radiation exposures. This data will provide further assurance that the electrical components of the crane will function.

8. The auximum stress on the crane brakes would be a sudden stop at maximum load at high hoist speed in the dcwnward direction. Section 9.5.4 states that the operator should switch from high speed to slow speed before setting the brakes. It is our position that the operator should brake The the load when the load is in high hoist speed in the downward direction.

settling observation would then be performed and then slow speed would be used to lower the load to the floor.

• 3 April 7.,1983

• J. Eg .1 anon

g. The Unit Work Instnetion' cover sheet does not appear to provide ade administrative controls over valve positioning. assurance th In addition, we suggest that Section 8.1 be revised to delete the This reference to the safety evaluation of the polar crane load test.

j j

procedure should implement the requirements of the safety evaluation.

References like this could confuse an operator.

l 2

! 10. Section 6.2 should also include reference to the visual markers been placed in the reactor building to limit crane travel.

11.

Section 4.1.7 requires the task supervisor to review the load paths.

j-The actual control of the load, however, is performed by the I operator.be required to review and concur with the load path restrictions. i i

12. This procedure does not clearly state responsibility and authority of individuals. For exasple, does the Task Supervisor have authority over i >

j In Section 6.1, what is meant by "test director shall the Test Director? We suggest ,

j consult with the test supervisor for further instrvctions"?  ;

Qis sentence and all similar sentences be deleted or clarified.

13.

Section 9.2.1 should clearly state load path restrictions when moving the load test lifting beams. '

t For example,

14. The procedure does not appear.to' clearly specify speeds. t Section 9.3.2 instructs the operintor to move the trolley, but never Also Section 9.1.1 does not specifies the speed, i.e. fast or slow. It is our i specify the speed for lowering the Internal Indexing Fixture. i position that both speeds should be tested.

Sertion 9.7.2 states that the bridge should not exceed third speed.

However, a caution note states that it can not be operated more than 15 seconds at this speed. This is not clear to us.

Section 9.1.1 states that the Internal Indexing Fixture lift of 6 inche should be performed in the fast mode.

, the slow mode. 'Can a lift of this height be adequately controlled in t fast mode?

15. The procedure is not consistant in stating allowable tolerances when For example Section 9.5.3 establishes a performing settling tests.

quarter inch allowance, whereas the next step, Section 9.5.4 states "

settling." Section 9.1.1 establishes a "no movement" tolerance, and Section 9.3.1.2 only states "if settling... exists."

In addition to reconciling the above, please infonn us of how load

)

" settling will be measured. .

) .

t

) . *

• ApH1 7, 1983

• J. E Larson 4

16. The procedure does not always clearly indicate when rigging should be disconnected. In addition, a general statement should be included in all steps that disassedle Mgging that the operator should be observant for obvious damage. .

17. The reference for the Mgging specified for the Internal Incexing Fixture f (Step 9.1) and the missile shield (9.3.1) is wrong.

18. Please provide the bases for the 90,000 pound and 70,000 pound pull limits ,

for the reactor and pressuMzer missile shields.

WG- '

Lake H. Barrett Deputy Program Director ,

TMI Program Office cc: 8. Kanga  !

J. Theising

3. Snyder T

f 9

-l 3

.i

_ ___-_