ML20053C706
| ML20053C706 | |
| Person / Time | |
|---|---|
| Site: | Washington Public Power Supply System |
| Issue date: | 05/17/1982 |
| From: | Dodds R, Narbut P, Powers E, Shackleton O, Toth A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML20053C691 | List: |
| References | |
| 50-460-82-04, 50-460-82-4, NUDOCS 8206020484 | |
| Download: ML20053C706 (16) | |
See also: IR 05000460/1982004
Text
.
U. S. NUCf. EAR RECUL\\ TORY C0!!MISSIO?T
.
y
RECIO!i V
Report No. 50-460/82-04
'
Docket No. 50-460
License No.
CPPR-134
Safeguards Group
Licensee: Washington Public Power Supply System
P. O. Box 968
Richland, Washington 99352
Facility Name: Washington Nuclear Project No. 1 (WNP-1)
.
Inspection at: WNP-1 Site, Benton County, Washington
Inspection conducted:
February 22-26 and March 2-12, 1982
Inspectors:
C%
f~
[ 72-
P. P. Narbut, Reactor Inspector
6 ate' S igned
R 3Mk
s/wlse
A. D. Tot , Senior Resident inspector
Date Signed
-
S f/l,4l Q,
-
Le
O. Cdhackleton, Invey igator
' Date' Signed
f/4kV
hw. -0
ferd
E. J( Power,A nvestigator
Date Signed
A 2 &&
s/n/ccu
R.'O. Dodds, Chief, Reactor Project bection 2
date Yigned
~
Reactor Construction Projects Branch
Summary:
'
-
Inspection during the period of February 22-26 and March 2-12, 1982.
i
l
Areas 'nspected:
The inspection was an investigation of allegations made
l
regarding the Heating Ventilation and Air Conditioning (HVAC) contractor,
University Nuclear Systems Inc. (UNSI). The inspection involved 288 inspector
j
hours onsite and 34 inspector hours in office by two NRC inspectors and
two NRC investightors.
Results: Four items of noncompliance were identified concerning recording
the wrong welder identity, 'not documenting corrective action for suspect
weld electrodes, changing weld procedure specifications with a memorandum
j
and QC acceptance of butt welds which were required to be fillet welds.
,
These items are discussad in paragraphs 2.c., 2.e., 2.i. and 3.a. respectively,
t
I
i
RV Form 219 (2)
S206020484 820517
i
- - . PDR ADOCK 05000460
G
ll_
'
1
~-
. , . . , - , _
, , . - - -
. , - -
-~
,
-
. . , - - - . . . -
,,y
-._
.m_
,~~
_
'
.
.
.
'
DETAILS
1.
Persons Contacted
)
a.
Washington Public Power Supply System (WPPSS)
- D. W. Mazur, Program Director
- C. B. Organ, Assistant' Program Director, Engineering
-
+*C. R. Edwards, Project Quality Assurance Ma n ager
-
+*W. Root, Assistant Program Director, Construction
.
- F. C. Hood,' Construction Manager
.
- L.'J. Garvin, Manager, Quality Performance and Measurements
.
.+*G. K.,Dykeman, Design Engineering Manager
+*M. E. Rodin, Senior QA Engineer
+*M. J. Farrell, QA Engineer
+ N. S.' Porter, Manager of Systems Engineering
.
+ R. C. Mertens, Senior QA Engineer
+ R. Glasscock, Director of Quality Assurance
b.
Bechtel Power Corporation (Bechtel)
+*D
R. Johnson, Manager.of Quality Assurance
+*J. Gatewood, Project Quality Assurance
+*T. Fallon, Project Quality Control Engineer
+*G. A. Hierzer, Field Constructior Ma nager
+*M. A.' Norm, QC Engineer
+ J. Ruud, QA Engineer
+ E. W. Edwards, Project Manager
c.
United Engineers and Constructors (UE&C)
+*E. C. Haren, Project Quality Assurance Manager
+*G. L. Faust, FSQA
+ K. J. Iverson, Manager Construction Services
d.
University Nuclear System Inc. (UNSI)
_
+*R. Canipe, Corporate QA Manager
j
+*S. Cohen, Project Manager
+*B. Sachs, Project QA/QC Manager
- Attended exit interview of February 26, 1982.
+ Attended exit interview of March 12, 1982.
!
,
..-
_.
. . .
-
-
,_
.
.
.
,
9
,
-2-
,
2.
Allegations against UNSI
On February 1,1982 the regional office received allegations against
the Heating Ventilation and Air Conditioning (HVAC) contractor, University
Nuclear System Inc. (UNSI), Contract 216. The allegations were pursued
during the weeks of February 1-5, February 22-26, and March 2-6 and
.
8-12, 1982.
In addition to the allegations initially received, additional
allegations and items considered worthy of-followup were received
as a result of interviews and investigations conducted. The allegations
and items are listed below with the NRC findings.
a.
Allegation: A foreman was fired because he refused to falsify
documents associated .;ith hangers.
Finding: The conclusions regarding the specific allegations will
De documented separately by the NRC investigative staff.
The underlying issue, that craft foremen were asked to sign off
work without looking at it, was substantiated.
The inspector examined the in-process inspection records (IPI's)
for the control room ducting, drawing 4586-1. There were a total
of 76 IPI sheets generated, with et.ch sheet having up to 8 items
signed off.
50 of the sheets had items which had been signed
off by different foremen within a three day period of January
6, 7 and 8, 1982. Eleven foremen were involved in the sign offs.
Nine of the foremen had a reasonable number of sign offs per day;
that is, approximately 8 to 11 items signed off per day. Two
of the foremen had an extraordinarily high number of items signed
off per day. One foreman had 77 items in one day and 49 items
the next day. The other of the two foremen had 39 items signed
off on one day.
,
An " item", as used above, is a unit of work accomplished. Typical
examples of " items" are:
" Install duct, 9092-175" and "All work
complete, 9805-78." The applicable procedure QCP/CP-27.0 states
the installation foreman is responsible to enter all appropriate
data on the IPI as work is performed.
If the item is complete,
the final column is to be initialed and dated. No IPI is to be
" torn apart" and distributed until the responsible foreman has
signified acceptability for all work performed by placing his
'
initials and date in the IPI work column.
The very high number of items signed off by the two foremen indicates
that they did not personally verify the work items were done, which
is an implicit requirement of the procedure.
The licensee's construction management personnel (Bechtel) responsible
-
for the UNSI contract stated that, in their opinion, the procedure
,
does not require craft foremen to personally verify work is done.
,
?
H
, - - - ,
.
-3-
The construction management's interpretation of the procedure
is basea on the absence of the words "shall personally verify".
The sign offs in question are not classified as " falsification"
because of this procedural interpretation.
The issue of circumvention of the intent of the procedtre was
discussed with licensee management at the exit interview.
The final hardware safety significance of this item is considered
minor on the basis that QC personnel subsequently inspected the
items prior to their sign off. The adequacy of QC inspection
is discussed further in paragraph 3.a. of this report,
b.
Allegation: ~ Documentation.is not being done in a timely fashion.
~
Finding: The lack'of timely documentation is a previously known
problem'(see Inspection Report 50-460/81-09).
No further examination
of.this problem.was performed during the current inspection.
,.
c .,
Allegation:
A~ welder's number was used on unmarked welds even
' -ttough they (the foremen) didn't know if the welder did those
welds. The applicable drawing is DH 4586 Sheet 1 and 2.
,
'Firiding: The al]egation was substantiated. The inspector examined
- all of the IPI's for the control room for drawing DH 4586 Sheets
1, 2 and 3 and identified the work items charged to the welder
-
- in question. Qual sty records indicate that on January 19, 1982
the welder's welding certifications were withdrawn. Discussions
s
with the welding foreman and the QC superintendent indicated the
, elder's work was bad, and that a lot of his work was required
w
to be ground out and rewelded by others. The inspector examined
records of work accomplished January 18-19, 1982 where the welder
was credited with welding in the control room. The IPI's show
him as the welder of record for eight items on IPI 5546. The
eight items involve the installation of duct branch take off stiffeners
per EAF numbers IE 1766 and IE 1964. The inspector examined voided
IPI's and determined the welder had been listed as performing
welding on an additional 17' duct branch take off stiffeners for
a total of 25 assemblies.
The inspector estimated that approximately 100 filler metal electrodes
would ordinarily be used to weld the 25 assemblies, but, possibly
the job could be done with 75 electrodes if the welder used his
partially burned stubs. The welding foreman concurred with this
estimation. Examination of all filler metal withdrawal slips
credited to the welder for the month of January 1982 showed the
welder used only 26 electrodes.
The inspector, therefore, concluded
the welder could not have performed welding on the 25 assemblies
for which he had been credited.
.
. -
-
- -
-
.
.
.
.
,
9
.
.
-4-
,
The inspector interviewed the foreman who had listed the welder
as having performed work on the 25 assemblies. He stated that
at the time his -job was to fill out the paperwork, that he did
not have direct responsibility for supervising a production crew.
He used the welder's number for the 25 assemblies based on conversations
with foremen, the general foreman and the welder. He further
stated that at the time of the sign offs in January, he had not
been trained regarding how the paperwork was to be completed.
The failure of the installation foreman to enter appropriate data
on the IPI as work is performed as required by QCP/CP 27.0 Revision 9
dated M a y 11, 1981, paragraph 6.1 is an apparent item of noncompliance.
(EnforcementItem 50-460/82-04/01)
Related to this item, the inspector learned that the licensee
had received an allegation through their. hotline program on October
20, 1981 which dealt with, among other things, the subject of
" arbitrary assignment of a (welders) name to a weld." The same
information had been provided to UNSI management on. September
10, 1981. The construction manager, Bechtel, conducted the investigation
en November 17, 1981 and concluded that they were not able to
substantiate any irregularities, but, would request UNSI to take
corrective action. On December 11, 1981 a memorandum to.the Program
Director was issued stating " corrective action where found necessary
has been implemented and a daily and/or weekly check will be made
by Bechtel." On December 24, 1981 another memorandum to the Program
D i rector was issued stating "Bechtel conducted a daily surveillance
...and did not detect a generic welder identification problem.
UNSI will revise their procedure...to require independent welder
identity verification, this will strengthen the UNSI welder identity
program and provide a high degree of assurance that proper identification
is achieved."
It is of concern that, despite notification to the contractor
and the licensee regarding this problem and the subsequent investigation,
the contractor's construction foremen were not apparently informed
of the problem and, as of January 1982, continued to inaccurately
record welder identification numbers.
This lack of feed'ack to
first line personnel was discussed at the exit interview.
d.
Allegation: Welds that are obviously bad were accepted in the
1979-1980 time frame. Such welds can be found on elevation 421,
see ET 73831 Sheet 1.
Undersize welds are located on elevation
501, see Drawing 4368 Sheet 2.
Finding: The problem of inadequate weld inspection is a known
condition and has been discussed in previous inspection reports.
This problem is currently being carried as an open followup item
(no. 81-06-08). The licensee has committed to perform a 100%
reinspection of all welds. No further examination of this problem
was performed during this inspection. The current adequacy of
weld-inspections is discussed in paragraph la of this report.
.
'
.
.
-5-
e.
Allegation: There was a weld filler metal withdrawal slip forgery
on December 2, 1981.
F_inding: The allegation was substantiated. The forgery was identified
through the contractor's quality assurance system but the matter
was never fully resolved.
The filler metal slip in question is dated 12/2/81 for welder
number 3680 and shows fifteen E 7018 electrodes were issued at
3:29 p.m. and returned at 8:00 p.m.
The 8:00 p.m. time is lined
out and 7:00 p.m. is written in with the initials "EVB".
During the regular QC surveillance of welding filler metal control
on December 2, 1981, the QC inspector determined and stated in
his inspection report that " Welder ID 3680 had E 7018 weld filler
metal out over the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> maximum by 31 minutes.
Initials of...
the filler metal attendant wer e forged by. party unknown changing
the time from the correct 8:00 p.m. to 7:00 p.m."
The surveillance report was forwarded by a " speed memo" from the
. inspector to the contractor's QA Manager on December 3, 1981.
On February 2, 1982 speed memo no. 14297 was written by a welding
QC-inspector to the welding QC supervisor stating the original
filler metal withdrawal. slips had been returned and requesting
what actiontshould be taken. The memorandum reply stated to file
the filler metal slips.
-
'
fat;thetime~ofthe'currentinspection,theNRCinspectorfound
~
the original weld' filler metal withdrawal slip in the permanent
file with the forgery unco r rected. The inspector found no evidence
~
that the hardware welded by welder 3680 on December 2,1981 had
been evaluated for acceptability.
' The NRC inspector also examined QCRFI's as reported in paragraph
Sa. of this report. One of the requests for information (RFI)
was number'433 dated December 10, 1981 which dealt with weld filler
metal withdrawal slips which indicate E7018 electrodes were not
returned within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />. Fourteen filler metal slips were identified
in a period that covered August 17,~1981 to September 4, 1981, 3 weeks.
The inspector and the contractor's QA manager attempted to. locate
the 14 listed weld filler metal withdrawal slips. Only'one could
be loc ated, weld filler metal withdrawal slip of August 21, 1981
for welder 4537 which showed E7018 electrodes had been out for
54 hours6.25e-4 days <br />0.015 hours <br />8.928571e-5 weeks <br />2.0547e-5 months <br />. The slip showed the time circled but no indication
that the work done by welder 4537 on August 21, 1981 had been
evaluated for acceptability.
The QA Ma nager stated that the missing filler metal slips were
probably filed separately but would not show whether the hardware
affected had been evaluated. He stated it appeared that UNSI had no system
,
.
.
_
-
-
.
'
.
'
W..
- ~
.
-6-
'
-
,
.
.for-correcting filler metal withdrawal slips discrepancies. The QC
inspector'_s request on QCRFI No. 433 had stated in part "A definitive
method of' correction limits and criteria should be forthcoming
to all . applicable personnel to set guidelines for correction of
weld filler metal withdrawal slips."
~
The QA Manager's reply, QA memorandum QA 646 of January 11, 1982
stated in response " Corrections to Quality Class 1 documentation
are allowed as described by QCP/CP 5.1, QA Records, paragraphs
5.6.1 and 5.6.2 and therefore no code or procedural violations
exist."
Paragraph 5.6.1 and 5.6.2 of QCP/CP 5.1 state in part:
"For changes made due to error or oversight, a line shall be drawn
through the original information, and the new information shall
be entered adjacent to the original information and initialed
and dated by the perscn making the change. A written justification
shall be added to each change when the reason is not otherwise
apparent.
If supplemental, new, or corrective information needs to be added
to a QA record, due to error or omission, and there is insufficient
room on the record, a Records Clarification Form shall be used.
In utilizing the Records Clarification (RC) form to complete record
deficiencies, a written justification for each change shall be
prepared where the reason for clarification is not readily apparent.
If the required information is inadequate or has not been provided,
a Document Deficiency Notice (DDN) shall be issued.
If the required
information is not provided within thirty days of the DDN issue
data, a CNCR shall be written against the document deficiency (ies)."
The failure to correct safety related records in accordance with
QCP/CP 5.1 and the failure to evaluate safety related hardware
welded with low hydrogen E 7018 welding electrodes which exceeded the
allowable usage time of 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> is an apparent item of noncompliance
(Enforcement Item 50-460/82-04/02).
f.
A]]egation: Two welders were never qualified. The welder's names
were provided.
Finding: The allegation was not substantiated. The NRC inspector
examined the qualification records and the work performance records
and found the welders were qualified during the periods that they
welded. The allegation apparently stemmed from the fact that
the record copy of the welder certification was not signed by
the documentation supervisor until several months after the certification
was put in effect. The inspector interviewed the documentation
.
,
-7-
.
supervisor who stated he had signed the typed official permenent
record on the basis of reviewing the hand written copy of the
certification which was signed by the welding examiner. Since
the welding examiner had lef t the site, the documentation supervisor
signed for him and dated his signature with the date on which
he performed his signature,
g.
Allegation: Drawings were withdrawn-from the field because base
plates didn't meet drawing tolerances.
-
Finding: The allegation was a correct statement but does not
represent a significant safety concern.
Drawing DH-4500, Sheet 8, Revision 0, had been in use in the field
and had location tolerances (+3") for certain items. Revision
4 of the drawing had been approved by UE&C for contractor use
but contained comments regarding areas of the drawing which had
not been submitted for revision approval. One of these comments
changed the location tolerances for certain items from 3 inches
to lh inches.
Revision 4, with the comments incorporated, was
issued to the field for use and caused concern since much of the
hardware had already been built to the looser tolerances. The
contractor's QA manager had Revision 4 withdrawn from tte field
and wrote a memorandum, 216-BEC-82-0054 of February 17, 1982 requesting,
essentially, that UE&C evaluate the necessity of changing the
location tolerances on equipment which hed been fabricated since
rework was involved.
In the case where approved drawing tolerances were not achievable,
it would have been more appropriate for tre contractor to use
the established request-for-information (RFI) system or nonconformance
report system, rather than issue a memorandum. The use of memoranda
is further discussed in paragraph 2.i.
The drawing control implementation
l
of withdrawing revision 4 of the drawing from the field is discussed
in paragraph 3.b.
h.
Allegation: A memorandum was issued on how to measure welding
machine wire feed speed which violates procedure requirements.
!
Finding: The allegation was substantiated. Quality control request
l
for information (QCRFI) number 462 was written on January 29,
1982 regarding how quality control should verify wire feed speed for the
"Cobramatic" feeder monitors. The general welding standard QCP/CP-22.0
Revision 7 paragraph 6.4.5 requires quality control personnel to verify
wire feed speed on Attachment 6 to the procedure. Wire feed speed is the
rate at which welding electrode wire is fed to the welding torch.
'
l
'
The response to the QCRFI is a memorandum by the welding QC supervisor
(number 9929 of February 3, 1982) which states to measure the
inches of weld deposited per minute and record that as tne wire
feed speed.
i
.
.
-8-
This is an inaccurate method of measuring wire feed speed since,
with the same wire feed speed a welder can deposit widely varying
inches-per-minute of weld deposit depending on the size of the
weld pass he chooses to make. The use of this method is considered
to be contrary to the requirements of the procedure.
The subject of using memorandum to change procedure requirements
is discussed further in paragraph 2.1. bElow.
i.
Allegation: Memoranda are still being used in lieu of procedures.
Finding: The allegation was substantiated.
Background: On January 27, 1981 the Program Director had issued
a letter which was directed to all site contractors (letter WPUE-
-
81-048). The letter stated in part"...it's absolutely essential
that our field personnel provide us the assurance that procedures
are being followed and so verified. There is no latitude for
deviation. Contractor Management is encouraged in a parallel
fashion to seek procedure relief as is appropriate but not at
the expense of procedure adherence." In September of 1981 the
licensee was issued an item of noncompliance because memorandum
were being used in lieu of approved proced"res to perform quality
related work. The licensee's response to this item of noncompliance
(WPPSS letter G01-81-424 of December 23,1981) stated in part
that, "...all site contractors who perform quality affecting activities
are being reinstructed that speed letters or memoranda cannot
be used in lieu of approved procedures, or in advarice of procedural
changes, all quality work must be accomplished in accordance with
approved procedures." Also on December 23, 1981 the UNSI Project
Manager and QA Manager issued memorandum QA 617 to all UNSI supervisory
personnel which stated in part that no work affecting quality
shall be performed to speed memos or verbal directior which does
not agree with approved procedures."
Current Allegation: The NRC inspector examined the memorandum
files of=the contractor's QA manager and the Assistant QA manager
for the period from September 1981 to the most current memorandums.
These files were not necessarily complete since memoranda are
not a contrclled document and individuals file only what they
deem necessary.
,
'The inspector found one additional memorandum which resulted in
deviatio'n from established requirements, The memorandum, from
the General Superintendent to all field supervision, (number 23670
of, January 27, 1982)' states that the material thickness limitations
,
giv'en in the specific weld procedure specifications (WPS's) apply
e
^
to groove welds only.and that fillet welds may be made using any
WPS_for khich "the essential variables apply WITHOUT REGARD TO
MATERIAL THICKNESS."
,
,
,
-
+
e
'
.
-9-
Preceeding that memorandum, the General Superintendent had written
a memorandum on January 19, 1982 to the QA Manager requesting
concurrence to that policy. The QA Manager concurred to tre
memorandum (concurrence not dated) and his response stated in
part "WPS revision is not required."
The memorandum was generated (per the statement of the General
Superintendent) because the WPS's, as written, weren't adequate
for some fillet welding of thinner members, namely the current
ranges weren't adequate. When this problem was brought to the
engineering department for WPS changes, the memorandum approach
was chosen instead, which allows the welder to select the amperage
range desired from a selection of weld procedure specifications
regardless of base. material thickness range. However, this also
allows the foreman or welder to select a WPS for which the electrode
size, travel speed, preheat or heat input may be inappropriate.
In the inspector's opinion, the decision to allow welders or foremen to
choose a WPS which provides an adequate current and for which "the essential
variables apply" places an unrealistic reliance on welder's or foremen's
_
knowledge of the welding code's " essential variables".
The failure to accomplish activities affecting quality in accordance
with documented instructions which prescribe the activities is
' considered a repeat item of noncompliance.
(Enforcement Item
L50-460/82-04/03)
A similar situation where a memorandum was used to change a procedure
requirement is discussed in paragraph 2.h. above. However the
safety significance of that memorandum is not considered sufficient
to warrant an item of nor. compliance.
j.
Allegation: An Engineering Action Form (EAF) was being used
in the field for work on Hilti bolt (concrete expansion anchor)
installation. The EAF violates the Hilti bcilt installation procedure.
Finding: The allegation was substantiated, but is not a safety
concern.
The inspector reviewed EAF No. IE2721 dated January 21, 1982
and the normal Hilti bc.lt installation procedure QCP/CP 33.0.
The EAF stated that Hilti Kwik bolt installation requirements
had been revised per a specification change. The EAF transmitted
the new requirements on an interim basis until QCP/CP 33.0 could
be revised. The EAF stated it would become void upon the revision
of QCP/CP 33.0.
Although the requirements of the EAF differ from that of the
procedure, the action to issue the new requirements is understandable
1
.
.
'
.
'
.
-10-
in terms of getting revised UE&C requirements in the field as
quickly as possible to minimize rework. UNSI can issue EAF's
without UE&C concurrence.
Revisions to QCP/CP's require UE&C
concurrence or a waiting period of 15 working days whichever
comes first before the procedure revision can be issued to the
field. There appeared to be no technical problem with the EAF
since it was interim in nature and invoked a UE&C approved specification
change. This item was discussed with -licensee management at
'
the exit interview.
k.
Allegation: Five to fifteen welders welded after the expiration
of their weldirig certifications.
Finding: The allegation was not substantiated. The NRC inspector
sampled records of several welders in the mid-1981 time frame.
Na records of welding after expiration of certifications were
found. The inspector noted that welder certification records
and welder performance records in that time frame appeared discrderly
and poorly organized and maintained. The records generated since
mid-1981 have greatly improved and are currently orderly and
required information can be systematically retrieved.
+
1.
Allegation: 'There'is no procedure for voiding IR's (Inspection
Reports) and'CNCR'.s (Contractor Non Conformance Reports).
Finding: The allegation was substantiated, but is not a safety
concern.
The NRC inspector examined the irs written for November and December
1981 and January and February of 1982. Although there is no procedure
for how to void an IR or CNCR, all the voided irs reviewed were
voided with an initial and date and in all cases but two, the
reason for voiding was recorded. Voided irs are retained for
the record.
The personnel ir.terviewed stated their intent was
to record a reason for all voided IR's.
The inspector reviewed
the rationale for each of the voided IR's and found it to be reasonable
in each case.
At the exit interview licensee management stated they would consider
changing their procedure to specify how IR's and CNCR's should
be voided.
This item will be inspected further in a future inspection.
(Followup Item 50-460/82-04/04)
3.
Additional inspections related to allegations
a.
Quality Control Inspection Accuracy
.
.
-
,
.
il
-
.
\\
.
-11-
f
The NRC inspector examined a sample of duct and hanger installations
in the control room against the requirements of UNSI Drawing
,
-
DH-4586 Sheets 1, 2.and 3.
Sheet 1 Section "C-C" shows the installation
'
of six inch diameter round duct take offs from the main duct.
Two horizontal lh inch x 1
inch x k inch reinforcing angle iron
-
bars.are shown to be installed to existing vertical members with
,
~
h . inch fillet welds, lh inches long. The vertical location of
'
the round duct take off assemblies on the vertical face of the
main ducts was not provided by the drawing information.
The NRC inspector observed that the horizontal angle iron was
installed with butt welds in lieu of the specified fillet welds.
This. condition was noted throughout the control room.
These welds had been final inspected and accepted by QC personnel.
Examples of final QC' acceptance on these welds are shown on IPI
5546 where'eight assemblies were accepted. Tv,o different inspectors
were involved in the inspection of iteme so the problem does
not appear to be confined to a single inspector. The items signed
off as acceptable and the dates were:
9087A-253 on January 21, 1982
9087A-254 on January 21, 1982
i
9087A-249 on January 21, 1982
9087A-433 on January 21, 1982
9087A-443 on January 21, 1982
9087A-421 on January 21, 1982
9087A-399 on January 19, 1982
9087A-171 on January 19, 1982
The NRC inspector interviewed the two QC inspectors involved.
They stated they had reviewed the drawing prior to performing
the inspection but did not take the drawing to the work at the
time of inspection. They did not note the absence of a vertical
locating dimension for the assembly on the duct nor did they
note butt welds had been installed in lieu of fillet welds.
Paragraph 6.7.1 of QCP/C0 No. 27.0 states in part that the QC
inspector shall assure that each item is installed in the proper
location and is the configuration specified by the applicable
drawing.
The failure to assure that the six inch round duct take off were
installed in the configuration specified by the drawing is an
apparent item of noncompliance.
(Enforcement Item 50-460/82-04/05)
The failure to note the absence of a vertical locating dimension
for the duct take off is not considered to have sufficient safety
significance to warrant an item of noncompliance.
The NRC inspector noted that another QC inspector had observed
that the drawing specified installation of stiffeners with filtet
welds was inappropriate since the weld joint configuration as
.
'
.
'
~
< _12_
drawn was actually a' butt joint and identified the deviation in
an ins'pection report IR-216-69 dated February 9, 1982.
b.
Drawing Control
Paragraph 2 9 of this report discusses an allegation regarding
Drawing DH 4500. Sheet 8, Revision 4 which was withdrawn from
the field because location tolerances had been changed and Revision 0
was to be reinstated. The resident inspector examined the crawing
control implementation of this action and found the following
conditions at different drawing stations:
1)
The UNSI procedure QCP/CP-5.2 Revision 0 dated July 16,
1981 in paragraph 5.3.7 requires that " Drawings in the field
identified as For-Information-Only must be separated from
.,
contract. drawings in a separate file."
It also describes
document control.
,
(a) GSB Elevation 519' Foreman Work Station
A memo (Sachs/Dixon-Bertsch BLS-261 dated 1-12-82,
- 16691) prescribed " reissue DH4500 Sheet 8 Revision 0
to the field. Revision 4 to drawing DH-4500 Sheet 8
is not released and is to be removed from the field."
The acknowledgement on the same memorandum states that
t
" Effective this date, Contracts / Document Control shall
comply with the body of your memo and implement removal
_
of this drawing from all ficld print stations...."
Drawing DH4500 Sheet 8 Revision 0 was available at
the work station.
It was marked " CONTROLLED", but
the drawing set was not identified.
Dr a wing DH4500 Sheet 8 Revision 4 was available at
the work station, but its title block had been torn
off. There was no marking of "V0ID", nor " CONTROLLED",
nor "INFORMATION ONLY".
A partial drawing was available at the work station;
it was marked in pen "DH4500-8".
It consisted of the
"End Plenum Base Dimensions" tabulation, ostensibly
/
a duplicate of that portion of Drawing DH4500 Sheet 8
Revision 4
There was no revision number noted, nor
marking of "V0ID", nor " CONTROLLED"; nor "INFORMATION
3
ONLY".
(b) GSB Elevation 519' QC Work Table
Drawing DH4500 Sheet 8 revision 0 was available at
'
the QC table.
It was marked " CONTROLLED".
,
,
!
i
s
.
_.
_
_
_.
.
.
'
.
-
,
,
'
-13-
.
(c) Central Weld Material Issue Room at Elevation 479'
The Drawing Log dated March 1, 1982 (which is updated
each week).showed a notation against drawing DH4500
Sheet 8, in that " Revision 4 was pulled from field
2-12-82 per memo BCS-261".
The Welding Filler Metal Withdrawal Slips Nos. 6873
and 6874 (dated 3-3-82) showed assignment of weld material
to drawing DH4500 Sheet 8 Revision 4.
The "4" was
circled, and the general foreman for the issue room
said that this had been identified as incorrect and
the responsible work foreman had been notified the
previous day.
(d) QC Document Control Trailer East of the GSB
!'
The QC working file contained DH4500 Sheet 8 Revision 4
It did not contain Revision 0.
The drawing control problems at the foreman work station and
at the QC document control trailer indicated that document control
within UNSI may be a problem.
This item will be inspected further in a future inspection.
(Followup Item 50-460/82-04/06)
c.
As Built Location Tolerances
i
The inspector examined the status of as-built drawings at UNSI.
At the time of inspection, as-built drawings were not yet being
generated by UNSI, since the as-built drawing procedure had not been
issu e d.
UNSI had received contract specification 9779-216 Revision
!
124 dated February 8, 1982 which revised as-built drawing requirements..
The inspector noted that the equipment location tolerances in
the specification revision, which represents the as-built location
accuracy, had greater latitude than the installation tolerances
used in the original erection as specified in the duct standards
.
drawing 604900.
The inspector interviewed the applicable designer supervisor
and was unable to conclusively determine which tolerances would
be used in design verification analysis.
This item will be inspected
further in a future inspection.
(Followup Item 50-460/82-04/07)
5.
Licensee Action on Previously Identified Items
,
.,.
, . _ . , ,
- - -
_
. _ - - . _ , . -
.
.
.-_
..
-
_
. . _ _.
_.
- - - .
-
..
..
.
'
.
-14-
'
a.
(0 pen) (50-460/81-07-06) Followup Item: Quality Control Requests
for Information (QCRFI's) are not answered in a timely fashion.
l
The inspector examined the licensee's action taken on this item
'
up to the date of the inspection. The licensee had not completed
their action on this item.
UNSI had changed the procedure QCP/CP 39 Revision 2 of January 12,
1982 to require the QCRFI's be answered within five working days
of receipt by Quality Assurance. The construction manager, Bechtel,
had closed the item based on the issuance of this revised procedure
requirement.
The licensee's QA organization had initiated an
audit of implementation of the revised procedure, sampling QCRFI 400 of
October 20, 1981 to 4E0 of January 8, 1982. They found that essentially
all QCRFI's were answered in a timely manner but 14 of the 50 sampled
did not have answers filed in the QCRFI file. The inspector examined
the 14 QCRFI's which originally did not have answers in the file
i
and the answers which were obtained. The answers to the QCRFI's
reviewed were considered reasonable with the following exceptions:
1)
QCRFI No. 433 of December 10, 1981 deals with low hydrogen
weld filler metal being returned to the rod room after the
4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> maximum time limit. This item is discussed fully
in paragraph 2.e. of this report. The answer to this QCRFI
was inappropriate in the inspector's opinion in that the underlying
,
problem was not resolved and the weld rod withdrawal slips ard
'
affected hardware were not examined for acceptability.
2)
QCRFI No. 424 of November 24, 1981 stated that a practical
welding problem existed in the field. Galvanized k" thick
angle iron had to be ground to remove the galvanizing prior
to welding resulting in thicknesses somewhat less than the
nominal
In many cases the duct standards require k"
"
.
fillet welds be made to the less than " thick angle iron.
UNSI Engineering posed the problem to UE&C by Contract Waiver
Request (CWR) No. 218 of January 8, 1982. UE&C responded
that, due to design requirements, the weld thickness ceuld
not be reduced and that the weld should be brought out to
full weld thickness.
The UNSI engineering response to the QCRFI dated March 9,
1982, Speed Memo No. 11946, stated that UE&C had not approved
the CWR but that the AWS code states the weld size need not
exceed the thickness of the thinner part.
..
i
, . . .
'
.
'
.
-15-
This answer appears to be contrary to the UE&C answer.
The UNSI engineering answer goes on to state that UNSI submitted
a letter to AWS, dated March 10, 1982, requesting clarification
of the AWS fillet weld details.
Specifically UNSI requested
an interpretation that when a weld throat size equivalent
to a k" fillet weld is obtained with one leg being shcrter
and the other leg being longer than h", then an acceptable
" weld has been achieved.
It appears that the RFI answer is insufficient since, although
it describes actions that may eventually resolve the problem,
it does not instruct the QC inspector on what actions should
be taken for ongoing work, and seems to suggest that welds
less than \\" are acceptable. Additionally, the UE&C answer
to the CWR appears inappropriate since the answer does not
resolve the practical field installation problems and simply
restates the requirement without ensuring a practical solution
will be achieved. An additional example of an inappropriate
QCRFI' answer is discussed in paragraph 2.h. of this report.
QCRFI 462 of January 29, 1982 dealt with a question on how
to measure wire feed speed for a MIG welding machine.
Thisitemremainsopenandwilibeinspectedfurtherduringa
future inspection for timeliness and appropriateness of QCRFI's.
6.
Exit Interview
,
The persons, identified in paragraph 1 met with the inspector on the
. dates indicated in paragraph 1.
The scope of the investigation and
the findings as discussed in this report were discussed.
t
i
i
!
t
.
_
-
_.
- . . .