ML20198C883
Text
,
1.
Allegation Category: QA/QC 2, Documentation Control j
2.
Allegation Number: AQ-72 and AQ-103 3.
Characterization:
It is alleged that unauthorized changes were made to the computerized design change log.
Allegation AQ-72 is that design change documents were lost, that no log was kept on the lost documentation, and that occasionally, when a specific document could not be found by the document control center.(DCC) satellite for craft personnel, the craft personnel would get around this by having the missing design change deleted from the computer log.
Allegation AQ-103 is that design changes that were never to be incorporat-ed (NI) into drawings were deleted from the "open" computerized design change log.
The alleger believed that these changes were critical to in-spection and construction and that work may have been performed without this necessary information.
4.
Assessment of Allegation: Both allegations address unauthorized changes to the computerized design change log, which provides the status of the design changes to site drawings and specifications. The allegations were that the computerized design change log was altered to agree with the doc-umentation in the drawing packages rather than.the packages agreeing with the computer design change log.
In assessing these allegations, the NRC Technical Review Team-(TRT) interviewed site personnel and reviewed proce-dures and instructions (References 3 and 4).
Design changes are revisions to engineering specifications or drawings 2
which affect the form, fit, or function of a structure, -system, or compo-i nent. Texas Utilities Electric Company (TUEC) used component modification cards (CMCs) and design change authorizations (DCAs) to convey design changes.
Generally, CMCs were issued for drawings and DCAs were written for specifications. Design changes listed on computer logs related to these allegations were initiated for released engineering drawings and/or specifications which had been approved for fabrication, installation, or inspection by authorized personnel.
The three types of mechanical drawings used at CPSES are flow diagrams
.(system descriptions point-to point schematic drawings), piping compos-ites (plan views outlining piping and components by room and elevation),
and isometrics (details of piping and component installations). Design changes were evaluated for their effect on or incorporation into all three types of drawings.
For example, even though a design change was written specifically against an isometric drawing, it was evaluated for effects against the higher tier composites and flow diagrams.
If the design change had no impact on these drawings, its status on the design change log was NI (never to be incorporated).
Each drawing, whether a flow dia-gram, a piping composite, or isometric had its own design change log. The design change logs for the isometrics were maintained manually, while the logs for the composites and flow diagrams were computer generated. These design change logs were s.orted and retrieved by design change status and categorized into three logs. Th.e logs were the " Current, Design Change Log," "All, Design Change Log," and "Open Design Change Log."
Access to B605230087 860512 0-59 PDR FOIA CARDE85-89 PDR
the logs was limited, based upon the user's function and responsibilities.
For example, the design change tracking group (DCTG) could access all three logs; DCC could access the " current" and the "open" logs.
Initially the DCC satellites could access the " current" log but after April 1984, satellites were limited to use of the "open" log.
Procedure DCP-3, Rev. 16, "CPSES Document Control Program," stated that the DCC satellites were responsible for maintaining updated compilations of applicable drawings, drawing revisions, and design change documents.
The TRT found that the satellite files were organized by document packag-es, each of which contained a specific drawing and its applicable design changes.
Each package included a design change log which reflected all of the design changes written for that drawing.
The satellites prepared the document packages and distributed them to craft, engineering, and QC personnel. When a document package was pre-pared, satellite personnel checked the computer log or manual log, as ap-propriate, to verify that drawings were current and that all design changes were included.
If they were not, satellite personnel updated the package from its file of documents or requested any missing documents from the DCC.
(See QA/QC Category 2, AQ-16, etc. for details of the TRT's as-sessment of satellite control of drawings and design changes.)
Prior to May 1984, the satellites used the " current" computer log to pre-pare and maintain design change packages for flow diagrams and composites.
The " current" computer log contained the latest revisions of CMCs and DCAs, including NI changes for the affected document.
The " current" log did not include superseded revisions or voided changes. Because the com-puter data base used to generate the logs was a historical file as well as a user file, superseded or voided design changes were not deleted from the data base. Historical data were printed on the "all" computer log, which contained each CMC and DCA written for an affected document, including the NLs, superseded, and voided changes.
When the computer data base was updated, the update was reviewed and a
" flag" was set so that only the latest revision of a design change ap-peared on the " current" log, if appropriate.
However, at times, design changes were occurring at such a rapid rate that there was a delay in set-ting the " flags." Therefore, on occasion, more than one revision of a.
design change would appear on the " current" log.
In such cases, the sat-ellites and DCC were expected to call the DCTG to correct the data base.
DCTG reviewed the Texas Utilities Generating Company (TUGCO) nuclear engi-neering (TNE) independent review documentation for the design change in question, verified its status, and set the computer " flag" to prevent the superseded revision from appearing on the " current" log. Therefore, it could appear to those in the satellites that one day a number of design changes existed against a drawing or specification and the next day those same design changes would not appear.
In reality, the superseded DCAs and CMCs were only removed from the " cur-rent" log, but were retained as historical information in the computerized "all" log which was neither accessible to nor required by the satellites.
It became evident to DCC and to the craft personnel that design document packages contained extraneous documentation, primarily those CMCs and DCAs 0-60
with the status of NI. To alleviate this problem, TUEC and. Brown & Root (B&R) decided in April of 1984 to replace the satellites'_use of the " cur-rent" log with a new log designated as "open," which also excluded NI doc-uments. Again, in comparing the "open" log to the " current" log, it would appear that a number of design changes were deleted from the computer.
Actually, the NI-designated design changes, which had no further impact on the ongoing work effort, were no longer identified or needed by the satel-lites for distribution; however, the NIs continued to be listed on the
" current" and "all" logs.
(A further review of the elimination of extra-neous design documentation by the satellites is contained in QA/QC'Catego-ry 2, allegation AQ-110.)
The TRT reviewed the computer logs ("all," " current," and "open") for four site drawings to assure that the logs accurately listed the appropriate design changes, revisions, and corresponding status.
From the design changes written against the selected drawings, 105 CMCs and 26 DCAs were listed as NI. The TRT determined that these design changes were appropri-ately categorized as NI. The TRT examined the " current" and "open" logs to verify that the correct changes, latest revisions, and proper status were listed, and found no discrepancies.
The TRT found that upon issuance of the "open" log, several nonconformance reports (NCRs) were written stating that the satellite design change log did not list the NIs or distribute them and that insufficient information was available for fabrication and inspection. The TRT found that these NCRs were dispositioned as requiring no action, and agreed with the dispo-sition, since the "open" log provided adequate information. The TRT noted that the NCRs were written and dated May 10, 1984, the second day follow-ing initiation of the "open" log.
The TRT determined that allegation AQ-72 occurred as a result of the " cur-rent" design change log listing more than one revision of a document.
When this occurred, the DCTG would verify the accuracy of the " current" design change log and make any necessary changes.
Based upon the TRT's assessment of the computer logs and their usage, the TRT feels that the alleger misunderstood the interrelationships and usage i
of the logs.
The TRT determined that allegation AQ-103 occurred as a result of the
" current" design change log being replaced by the "open" design change log, which deleted those CMCs and DCAs that were NI. The TRT believes that the packages distributed using the "open" design change log provided adequate design documents for construction, fabrication, and inspection.
After assessing the above issues and concerns, it is the TRT's view that if the DCC/ satellite personnel had been given adequate instruction and training, this allegation would not have been made.
l 5.
Conclusion and Staff Position: Based on the its assessment of the comput-er system used to track design documents, the TRT concludes that allega-l tion AQ-72 was not substantiated.
Allegation AQ-103, that "NI" design l
changes were deleted from the "open" log, was substantiated.
Nevertheless, (1) no design changes were lost, even though that appeared to be the case; 0-61 l
b (2) the deletion of the NI design changes from the "open" log did not re-sult in inadequate information being issued for construction, fabrication, and inspection; and (3) the status of all design changes to site drawings was reviewed, verified, and approved.
On November 27, 1984, the TRT presented the results of its assessment of t
allegation AQ-72 and the TRT's conclusion to the alleger.
No new concerns or allegations were identified.
AQ-103 came f:om testimony of an indivi-dual not identified as an alleger.
(
.5 m% -
7.
Attachments:
None.
Reference Documents:
1.
" Nuclear Power Plant Components."
2.
ANSI N45.2, "QA Program Requirements for Nuclear Facilities."
3.
Procedures:
a.
Gibbs & Hill Project Guides 24 and 29.
b.
TUGC0 CP-EP-4.7, " Control of Engineering / Design Review of Field Design Changes."
c.
B&R CP-EI-4.0-22, Rev. 1, " Design Document Control by TSMD."
d.
DCP-3, Revisions 16-18, "CPSES Document Control Program."
4 FSAR, Chapter 17.1.6, " Document Control."
5.
Region IV Report:
New Issues, D. Norman to R. Bangart, Item No. 4.
6.
AQ-103 Testimony, A-61 Statement, dated May 8, 1984.
7.
AQ-72 Testimony, A-2 Statement, A-5 Interview, pages 42-44, dated April 3, 1984, and A-61 Statement, page 3, dated May 1, 1984.
8.
A-5 Interview, November 27, 1984, pp.95-105.
8.
This statement prepared by:
k d
Avst/
V. Wenczel, 4RT Date Technical Reviewer Reviewed by:
24////L2 dYD H. Livermore, Date Group Leader Approved by:
V. Noonan, Date Project Director 0-62
,