IR 05000323/1987039

Insp Rept 50-323/87-39 on 870805-1030.Violations Noted. Major Areas Inspected:Improperly Authorized Flame Heating of RHR Pipe on 870505

ML20236W897
Person / Time
Site:	Diablo Canyon
Issue date:	11/17/1987
From:	Mendonca M, Narbut P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:
Shared Package
ML16341E452	List: IR 05000323/1987039 ML17095A534
References
50-323-87-39, NUDOCS 8712080301
Download: ML20236W897 (9)

Text

_ __ - _ _ - __ _ - - - - - - -- - - - -

---

- - - - -

.

.

U. S. NUCLEAR REGULATORY COMMISSION

REGION V

Report No:

50-323/87-39 Docket No:

50-323 License No:

DPR-82 Licensee:

Pacific Gas and Electric Company 77 Beale Street, Room'1451 San Francisco, California 94106 Facility Name:

Diablo Canyon Unit 2 Inspection at:

Diablo Canyon Site, San Luis Obispo County, California Inspection Conducted:

August 5, September 10-11, October 12-13 and 22 and 30, 1987 Inspector:

%

N -^

~

^

////7/B7 P. P. Narbut, Senior Resident Inspector g Date 'Sicjned Approved by:

it/n/77 M. M. Mendonca, Chief, Reactor ProjectsSection I Date' Signed Summary:

Inspection on August 5, September 10-11, October 12-13 and 22 and 30, 1987 (Report No. 50 323/87-39.

The inspection was a special technical inspection conducted in conjunction with an Office of Investigations inquiry into the improperly authorized flame heating of an RHR pipe on May 5, 1987.

Results of the Inspection:

Two apparent violations were identified (paragraphs 6.b. and 6.c.) regarding failure to implement design considerations in maintenance procedures and use of a work order in lieu of a procedure.

8712080301 871123 PDR ADOCK 05000323 G

PDR

_ - _. -_

__ -

.

.

DETAILS

..

v-1.

. Persons Contacted

• J. D. Townsend, Acting Plant Manager

,

• J. A.: Sexton, Assistant Plant Manager, Plant Superintendent

• J.,M. Gisclon, Acting Assistant Plant Manager for Support Services

.,

,

• C. L. Eldridge, Quality Control Manager

• D. B. Miklush, Maintenance Manager

.

• D. A. Taggert, Director. Quality Support

• T.:J. Martin, Training Manager

• W.lG.'Crockett, Instrumentation and Control Maintenance Manager

• S. G. Banton,' Engineering Manager

'

.

"

• T. L. Grebel, Regulatory Compliance Supervisor

.

S. R. Fridley, Senior Operations Supervisor

• R. S. Weinberg, News Service Representative M. J. Angus, Work Planning Manager

-

The inspector interviewed maintenance engineers, mechanics,' work planners, and welding engineers involved in the RHR pipe heating occurrence.

-

Denotes those attending the exit interview on October 22, 1987.

..

2.

- Background to the Residual Heat Removal (RHR) Pipe Heating Occurrence The pipe heating occurrence was previously described as an unresolved item in inspection report 50-323/87-20 paragraph 4.e..

Follow-up actions of the unresolved item were purposefully delayed to allow full participation by the NRC Office of Investigation (01).

OI has completed their inquiry.

This inspection report deals with the technical aspects of the occurrence.

Unresolved items 50-323/87-20-06 and -07 are

. considered closed.and superceded by this report.

3.

Brief Overview of the Occurrence During the Unit 2 refueling outage, with all reactor fuel. removed from the core (and the Residual Heat Removal system therefore not required to be available or in service), the licensee planned to replace gaskets at a flanged connection downstream of RHR pump 2-1.

The flange was at the location of flow orifice (FE641A). When the piping flange was disassembled, the flange fittings sprung apart.

Minor springing of bolted flanges is common but the amount of springing was excessive.in the opinion of the mechanic, and maintenance engineering assistance wa.s

- consequently requested.

The usual solutions for such a situation involve cutting pipe joints and rewelding to achieve fitup, or a design engineering analysis may authorize pulling the flanges together by springing the pipe if the induced stress values are determined to be acceptable. For clarity, it should be noted here that the maintenance engineer does not have design engineer function or authority.

l l

-

_ _ _ _ _ _ _ - _ _ _ - _ _

,

,.

,

The maintenance engineering personnel involved, however, attempted an experiment to " draw" the pipe into place using. localized torch heating.

This is a common practice in commercial low carbon steel fabrication, but the RHR pipe is stainless steel.

" Drawing" the pipe with heat is not routinely done on stainless steel because it undergoes a metallurgical phenomenon called sensitization when heated-to 800-1500 degrees F.

Sensitization makes the stainless steel subject to grain boundary corrosion and can lead to through wall defects.

The maintenance' engineer with the lead for the job was aware of this phenomenon but instructed the work be done.

The heating of the pipe was done on the evening of May 5, 1987.' The heating was unsuccessful in drawing the pipe, but was done at 1200 riegrees F which most likely caused-sensitization of the stainless steel.

The engineer stated that the heating was done as an experiment and that his intent was to have the pipe remvved and replaced after the experiment.

However, the work order documentation associated with the job showed a work instruction path that could have led to reuse of the pipe.

The improper heating of the pipe was discovered by two separate licensee personnel peripherally involved in the job rind was brought to the attention of licensee management.

Proper replacement.of the affected pipe was performed prior to the pipe being placed in service and licensee management took corrective actions to prevent recurrence as a result of the event.

However, the occurrence

' points out several areas of weakness which are discussed later in this i

report and were not all addressed in the licensee's corrective action.

4.

Sequence of Events 4/13/87 A work order (WO C0011616, Activity 1) was issued to replace the gaskets on the flow orifice FE 641A.

5/1/87 Upon disassembly the flanges sprung apart.

The flanges were out of parallel about 5/32 inches.

5/4/87 The work order was amended to allow heating the RHR piping.

The work order WO C0011616 Activity 2 required that a maintenance engineer direct the work, the " flanges" be heated not to exceed 1250 degrees F, and the " flanges" be immediately quenched with water.

The work order then instructed the gasket changeout be completed.

The work order was prepared and issued l

by a work planner based on verbal information passed from the l

dayshift maintenance engineer (welding engineer) to the swing l

shift maintenance engineer to the work planner.

As issued on May 4 by the work planner, the work order was not reviewed by engineering, but was reviewed by a QC inspector who did not have expertise in pipe welding or heat treating.

No QC hold points or inspections were added by the QC reviewer.

No steps in the work order precluded use of the heated pipe and, in fact, the work order directed reassembly.

f

<

+h:

,~

,-

'

,

77@ w <

g-

.

,

4(

q-

'3'

n, -

-

.

,

.

,

5/5/87 Heating of the pipe took place on swing. shift.. The work order,

~

instructions.were not followed since the heating of," flanges"

'

would not " draw".the= pipe. 'Instead a segment of thel" pipe":was

. heated:to 1200 degrees-F and then cooled three times.

The

.' cooling was done with' carbon dioxide fire extinguishers.twice and with water once (the work order'specified wateriquenching'

only).. Additionally, rigging devices were used to apply. force

,

to the pipe to pull it.into place.

'

.

.

-

5/6/87'

Aftersthe unsuccessful attempt to " draw" the pipe, an attempt.

.

.to bolt the connection was made.to achieve gasket crush. '.The attempt was unsuccessful QC was involved.in this portion'of-the work since they were required to verify cleanliness and'

<-

.

torque for final bolting..The work was rejected for inadequate'

'

gasket crush.

5/7/87 Maintenance. engineering issued an Engineering Work Request-EWR 87-172 requesting a design engineering analysis to

.' determine 1) if the pipe could be. sprung into. place, 2) if, skewed gasket spacers could be used; or 3) if cutting and trewelding would-be required.

The fact that the pipe had'been heated (and would have to be replaced) was not addressed.

.

'5/8/87-A design engineer examined the pipe in'the plant'(in response to the EWR) and observed the darkened area caused by heating.

He also observed that the pipe support spring cans had'not been.

!

blocked prior to disassembly as is normally required.

The

' design engineer notified a quality' support engineer.

The quality support engineer went to the field and observed the j

condition.

I

!

5/9/87 The quality support engineer wrote an action request (AR I

A071961) describing the. heating of the pipe without a detailed.

procedure or documentation.

5/11/87 QC issued a Quality Evaluation (QE) Q 00040% d: a result of the AR written on 5/9/87.~ This process b the normel pene m j

of escalation of a problem to a nonconformance.

i

!

5/15/87 The maintenance manager issued nonconformance report NCR DC 2-87 MM-: 057.

,

5/31/87 On this date final torquing was performed on the repaired assembly.

Repair had consisted of replacing the heated pipe by cutting and rewelding a new pipe section.

L 5.

Corrective Actions Taken by the Licensee in Response to the Occurrence The licensee took corrective action as specified in NCR DC 2-87 MM-N057.

Specifically the action was:

o to replace the pipe;

y l-o:s l

.-

'$

,

i t hi a.

_

._

-____ _ - _ - _ _ _ -

.J

-

,

.to revise the administrative procedure for work orders to provide

)

o clear guidance to the work planners as to when a special process l

procedure is required;

'

to revise the quality control procedure, which describes how QC o

,

reviews work control documents, with a section describing when special process procedures are required; o

to train QC in the proceduta revision; o

to prepare a maintenance procedure similar to the construction procedure " Piping Fabrication, Replacement, Repair or System Alteration" which would address piping considerations not covered in

,

existing maintenance procedures, such as,_ limitations on cold

{

springing of piping and blocking spring can hangers prior to pipe j

disassembly.

(

6.

Inspection Observations and Findings

The inspector interviewed involved personnel and management and reviewed

associated records of the work activities.

The inspector's observations

'

and findings were:

,

a.

Lack of Procedural Controls in Certain Maintenance Activities Early in the inspection process, the inspector determined that maintenance personnel did not have procedural controls to limit cold springing of pipe.

Typical controls are for design engineering personnel to specify in a procedure the amount of cold spring allowed (for given pipe sizes and lengths of run) and to include a caveat that cold springing in excess of procedure limits requires design engineering analysis and approval.

The inspector noted that controls had been established in the licensees construction organization but not in the plant operations / maintenance organization.

Secondly, specific past problems such as unauthorized torch heating of stainless steel pipe (causing sensitization) had been factored in to construction procedures in the form of cautions and prohibitions.

The inspectors noted that such " lessons learned" in construction had not been factored into maintenance procedures.

Pursuant to discussions with the maintenance manager, the licensee included an action in the resolution to the nonconformance report (NCR DC 2-87 MM-N057).

The action was to develop a maintenance procedure for piping that would address piping considerations not covered in maintenance procedures.

At an exit interview on October 16, 1987, the inspector stated that other lessons learned in the construction program may also be appropriate for inclusion in the pertinent plant J

programs and procedures.

k b.

Failure to Implement Design Criteria The inspector noted that PG&E " Specification for Erecting Main System Piping," Specification No. 8711, Revision 20, dated July 8,

..

__ _

__ _-

_ _ _ - -

.

..

l

!

1985, specifies important. constraints on methods of work.

These constraints are assumed _by the design organization (Engineering) to be implemented by the field work force.

The constraints applicable in this event included prohibitions against cold springing of pipe and heating of pipe to induce bending. The constraints were inferred from the specification requirements for heat treatment and for_ Code _ requirements for allowable stress levels.

These design constraints were explicitly implemented in construction procedures, e.g., DCP 500 Revision 1 " Piping Fabrication Installation, Replacement Repair or Alteration;" but had not been implemented in the licensee's maintenance procedures.

l The-failure to assure that applicable design requirements were q

acceptably translated into maintenance procedures is an apparent i

violation of 10 CFR 50 Appendix B Criterion III.

(Item l

50-323/87-39-01).

I c.

Weaknesses in Identification of Root Cause Corrective Action The maintenance engineer, with the lead for the resolution m the misaligned RHR pipe, stated that he considered the heating ot the pipe to achieve adequate flange alignment to be an experiment, and that he always intended to have the pipe removed and replaced.

Having made that decision, he verbally informed the swing shift engineer of his intended heating and quenching actions.

The swing shift maintenance engineer in turn verbally related that information to a work planner.

The work planner prepared and issued a work order to do the heating and quenching.

In accordance with the licensee's administrative procedures, a work order is not a procedure and to attempt such work using a work order in lieu of a procedure is improper.

The licensee's administrative procedure for preparing work orders clearly prohibits using the work order in lieu of a procedure.

Specifically, procedure NPAP C-4053 defines a work order as a planning document and states that it shall not be used in lieu of a written detailed procedure.

All the organizations involved in issuing the work order failed to follow

the administrative procedure.

These included the maintenance engineers, the work planner, and the QC reviewer.

The use of the work order document in lieu of a procedure is an apparent violation of NRC requirements (Item 50-323/87-30-02).

The licensee's corrective action for this problem was incomplete in that the licensee action focused on the controlling administrative document and did not address corrective action for personnel errors.

The primary licensee corrective action was to add a list of specific

!

activities that " require a procedure" to the controlling I

administrative document (NPAP C-40S3) and to the QC procedure for

)

review of work orders (QCP 10.4).

The corrective action taken, of

adding a specific and limited list of special activities which I

" require a procedure" appears to weaken and confuse the formerly l

clear administrative requirements.

Whereas the administrative j

document formally required work crders to invoke a procedure in all

'

cases, it now appears that although not the intent, procedures are

--_

.. - - -

a-

,

'6

.

'

,

(

. >

only required in seven special process cases such as welding, heat

'

treating and nondestructive examination.

At the exit interview, the

'

licensee stated a review would be performed to assure the clarity of L

the procedure change and to issue a change if it was determined to l

be required.

Secondly, the.only training specified in the licensees corrective action was specified for QC and was given only to the four normal QC planners who review work orders.

The training was in regards to the change to QCP 10.4.

The QC individual who had approved the improper work order did not receive. training as indicated on the training record nor did other QC personnel who likewise occasionally fulfill the QC planning function.

No action was specified to retrain

. maintenance engineers or work planners on.the licensee's administrative requirement that work orders invoke procedures and do not substitute for them.

The matters were discussed with licensee management who stated the remainder of'QC personnel, the work planning center personnel, and the maintenance engineers, foremen and mechanics would receive appropriate training to insure against recurrence of a.similar event.

The adequacy of the work' order procedure as revised and of the licensees training actions will be evaluated in the normal course of future inspections, d.

Weaknesses in Work Completion Documentation The inspector noted that.the work order records are primarily written and stored electronically on computer files.

As such many entries are electronically added and " signed" by an entry of name and date.

In the case of the " comments" on the work order (WO C0011616), the inspector observed that the night shift maintenance foreman had apparently made an entry on May 6,1987, that heating was done.

The entry stated:

" Heated and moved probably got.005" total movement.

This did not work.

Heated 3 times twice without alignment pressure once with."

(Name deleted by NRC)

...put it'

"

together.

So we started assembly will need to complete torquing."

This was the only available record that the pipe had been heated.

The inspector interviewed the foreman who apparently made the entry.

He stated he did not make the entry on the computer, but that the entries reflect his hand written notes for that evening which he had left for day shift.

He stated that apparently the work planning center clerk made the entries.

The. corresponding turnover information from day shift to night shift was not entered on the computer, presumably because that turnover was performed face-to-face.

There is no requirement to make the entries but the inspectors found them useful in establishing facts.

Further, the inspector was told by the planning department that the

" comments" section is temporary in nature, is electronically erased and does not become a final retained record.

However, the final hard copy of the work order is that copy which was electronically I

n

.

J

--

.

"

,

%

.

i printed and issued for field use, signed off step-by-step by the I

mechanics performing work and is the copy reviewed for completion.

The final reviewed copy of WO C 0011616 caused confusion because the steps for heating the RHR pipe had "N/A" entered and initialed

(whereas heating had'in fact been done).

Investigation and

interview revealed that the steps had been marked "N/A" by the work'

^

planner performing a document closeout review on June 5, 1987.

The planner assumed the heating had not been done since the work steps had not been signed off complete and since the hand written work summary (accurately) stated in a May 15, 1987, entry that the pipe was to be replaced.

The planner had not reviewed the electronic

" comments" section of the work order.

The inspector discussed the

,

apparent confusion in work completion documentation with licensee

!

management.

Management stated that corrective actions would be taken in the area of the work order execution and clo;ure process and would include-

.

improvements to insure accurate information on work performance and

)

equipment history.

e.

Design Documents Not Easily Available to Plant Personnel

i In the attempt to review applicable design documents such as the i

piping fabrication specification, the inspector determined that a useable controlled copy was not available through the plant document control.

The plant document control did have microfilm available for each of the 20 specified revisions but this was considered to be an essentially unuseable form since the user would have to produce microfilm copies of 20 revisions and then assemble, collate, and discard superceded pages.

The inspector was able to locate controlled copies of the specification in General Construction located outside the protected area.

The indication from the above document situation is that design specification infonnation is not readily available for plant personnel use, including those personnel responsible for site maintenance and modification controls.

At the exit interview, site management stated that corrective actions would be taken to ensure controlled copies of important design documents are available for plant staff use.

This item will be followed up in the normal course of future inspections.

Licensee j

management also stated that PG&E had established a task force in

[

response to the industry initiative of configuration management as I

'

it relates to operating plants.

7.

Exit Meeting

The inspector conducted an exit meeting on October 22, 1997, with the members of the plant staff as indicated in paragraph 1.

During this

meeting, the inspector summarized the scope of the inspection activities j

and reviewed the inspection findings as described in this report.

The

!

licensee acknowledged the concerns identified in the report.

Further

!

!

!

I

~

- - _ _ -.

_

g.

..

-

,..

discussions with licensee management were held in the Region V offices in Walnut Creek, California as a part of the agenda of topics discussed on October.'30, 1987.

l I-l

!

b

,

I i

i

l l

l i

!