ML20129G215
Text
J
. my w UNITED STATES
.p * * %q':,
NUCLEAR REGULATORY COMMISSION fW
+/ 1 REGION IV i it *'.. ',j ARLINGTON. TEXAS 76012 611 RY AN PLAZA DRIVE, SUITE 1000
'{g-f s
MMf Docket.No. 50-498 50-499 MEMORANDUM FOR:
W. C. Seidle, Chief, RC&ES Branch E 9 W. A. Crossman, Chief, Projects Section FROM:
H. S. Phillips,- Resident Reactor Inspector, STP
SUBJECT:
SOUTH TEXAS PROJECT INVESTIGATION 79-19 l
LESSONS LEARNED Past history, including QA history, is only valuable if it leads to an Jimproved present or future condition.
Team members stated, after the
' investigation, that NRC managers, supervisors, inspectors and investi-gators should be made-aware of lessons learned from the subject investi-gation which started on November 2,.1979 and ended January 24, 1980.
The following are my observations based on 22 years of QA and technical experience, as well as, the past few years working / observing the NRC inspecticn system.
These coments are constructively offered so that the NRC might improve future investigations and so that the NRC might detect QA program breakdowns at the earliest possible time..
Some of the coments are specifically related to the conduct of the' investigation while the QA program coments address more generic issues relative to r
the NRC inspection system.
1.
Lessons Learned Relative to OA procramatic Breakdown The saying that, "An ounce of prevention is worth a pound of cure" may well apply to the STP South Texas Project QA program partial breakdown.
I believe Region IV did the best they could to identify the problems at South Texas Nuclear Project, but were unable to conclusively show a QA programatic breakdown because they did not have adecuate manpower to perfom a lengthy and in depth review.
It was a very difficult task for the investigation team,which spent nearly 3 months on site,to identify the problems. The team also spent considerable time off site after the investigation documenting the results. After extensive evaluation of the results, the parts of the puzzle came together to i
give a clear picture of a partial program breakdown.
It is important for NRC managers to realize that QA programs can only be adequately evaluated by more extensive reviews than have been the practice in the past.
- a. Needed Imorovement:
The NRC does a fair jcb of evaluating the licensee's-establishment of a QA/QC program during the Pre Const-
.ruction Permit (PRECP) inspection phase, however, after that the NRC's inspection system shifts it's primary attention to random 8506070194 840620 gg P
pgt
-a
2 inspection of hareware. As a result tne licensee's and the con-tractor's QA program / systems, including the quality control func-tion, can fail or partially fail and go undetected because of the emphasis on isolated inspections instead of NRC inspections and an annual in depth QA review / audit to evaluate any-trends to detemine if there have been any partial or total QA/QC programatic break-down since the PRE-CP inspections verified QA/QC program establish-ment at the. corporate and site level.
It appears that the NRC has not had adequate manpower to routinely perform these types of reviews and also perfom routine and reac-tive inspections required.,
Recomendation:
A review of each licensee's corocrate and site QA program anc systems which implement the eighteen (18) criteria
.of 10 CFR 50 Appendix B should be perfomed by a special NRC team i
annually. An in depth review should serve as a basis for recommend-ing that construction work, in whole or part, should continue for another year, that is, until the next OA review. These annual reviews of construction activities should continue until the operating licensee is granted.
This would reaffim the PRE CP decision relative to QA program establistynent and implementation.
Assuming this iecomendation was implemented, it would be necessary to provide additional manpower needed by each region to perfom such in depth reviews.. Such reviews would require 4-5 inspectors i
for 4-6 weeks on each construction site. The added manpower would payoff in increasing the probability of the NRC's discovering QA programmatic breakdowns early in the project, thus eliminating situations which could lead to potentially serious construction deficiencies or concerns. This greater involvement should give the public greater confidence that all is well throughout the construction phase.
- b. Needed Imorovement: The NRC placed resident inspectors on sites well after tne project started.
Therefore, the project has not received the same attention as later in the project.
It is a well known fact that good or bad habits are developed early in a project and should be identified early for quickest and easiest resolution by the licensee.
Recomendation: The resident inspector should be assigned at the earliest practical time after the construction pemit is issued because it is critical that the NRC assure continued implementation.
- c. Needed Imorovement: The NRC currently assigns one resident inspector to a cesignated site.
I don't believe one can de the job that should be done. The following factors appear to have received insufficient consideration in detemining manpower:
(1) The construction phase is probably more complex.than the 2
operations phase since the resident has to deal with several different site organizations with 3000-4000 workers which incluce top managers and workers in licensing, quality assurance, engineer-ing, and construction.
Usually there are several different QA programs on site i.e. licensee, and contractors who are responsible for their respective QA/QC effort, (2) Reactive inspection effort can be quite extensive because of 50.55(e) deficiences, allegations,
7..
3 and interviews and will take a considerable amount of the resident's time.
News media interest is also a factor, (3) Licensee / contractor performance can require more or less of the resident inspector's time depending on the frequency and/or seriousness of problems, (4)-Quality and reliability must be built into a plant making the construction phase very important. A plant will realiably operate only as well as it is constructed.
Several other considerations could be stated but these are a few examples which sould make a point.
It would be more appropriate to discuss sucn subjects in a manpower justification report should NRC management choose to do so.
Recommendation:
TheNRCsiouldconsiderassigningadditional engineering / inspection personnel to the site to do a more adequate job. As a minimum, a second inspector should be added to multi unit sites when construction of the first unit is 30-40% ccmpleted.
I have worked in other government agencies / projects and.can safely say that their minimum QA staffing of a project such as this would consist of one senior resident inspector, a discipline engineer l
for each discipline (civil, mechanical, electrical) and a QA spec-ialist/QC inspector for each discipline.
In the case of poor licensee / contractor performance perhaps double this number would be assigned.
Several options exist relative to additional manpower.
(1) Permanent inspectors could be hired at the GS-11 and GS-12 level and work to higher grades.
(2) Highly competent and highly compensated temporary inspectors could be hired with the same 4
understanding of most construction workers, that is, when the project ends the individual would have to find other work if a permanent position were not available.
2.
Lessons Learned on Conducting Extensive Investications
- a. Needed Imorovement:
Some members of the inspection team lacked the desired NRC experience to participate in such an extensive 4
investigation / inspection.
Recommendation:
Each team member should have no less than one year NRC experience when participating in a special investiga-tion / inspection.
Technical expertise not available on the
. team should be supplemented.
For axample, technical experts j
with less experience could participate as needed but should be assigned to a senior team member who would be responsible for directing the effort.
Such technical experts would submit fir. dings to the senior team member-who would be responsible i
for including the finding in the final report.
- b. Needed Improvement:
A great number of interviews were conducted during the investigation and, as a result, the last half of j
the 50 persons' to be interviewed appeared to know the line of questioning.
In other instances, a few of the interviewees appeared to be under intense pressure because they were the last to be interviewed. That is, they waited so long to be interviewed that they tried to read something into being last.
e
- ~-
,,,-r
--~e-
-,--r
-m-,e----,-
n--
wn, w,
-ee-
--.w--
---+
4 Recommended:
Identify this as a potential problem before a lengtny interview process begins or early into the interview process.
Furnish two investigators and assign a technical person for each to shorten the interview time if the interview process is to be lengthy, i.e. more than two weeks.
- c. Needed Imorovement:
I participated on the team.as the Region IV representative out reported to the team leader from another region during this period of time.
I was given the task to be an objective team member reporting to a team leader from another region.
During the investigation I maintained contact with my Regional Supervisor'.
I kept my suoervisors informed as well as-I could; howeve?, I did not speak for the team.
Communication rules were not apparent.
This was an awkward situation.
Recommendation:
My first inclination would be to recommend that the Regional Project, resident or engineering inspectors, not 1
be a part of the team, however, in the case of the resident /
project inspector this must be weighed against the advantage of having their broad knowledge and feel for the project and per-sonnel.
If a Regional inspector is assigned to the team, Headquarters, Region Management, and the Team Leader should establish rules of communication during the investigation that will keep Region llanagement infomed. This would have improved the situation whether Region IV personnel participated or not.
s
- d. Needed Improvement: The team held briefings with the licensee intermittedly and there was a management meeting at the midpoint of the three month inspection.
Some of the team's findings were not fully developed, however, preliminary findings were discussed l
with the licensee. As a result, the licensee began to take cor-rective action on the basis of these discussions and began dealing j
with Region IV, the Resident Inspector, and the Team Leader regard-ing tentative and final findings before the inspection was com-pleted.
This caused several potential problems: (1) an official corrective action response to a briefing was made four months before the licensee received the official report, (2) reacting to findings before evaluations were completed could have led to incorrect corrective action, (3) potential compromise of NRC enforcement was real and, (4) feedback of information to the per-sonnel yet to be interviewed may have caused the interview pro-cess to be less effective.
Recmunendation: Only a final exit interview should be conducted during an investigation; however, activities in progress, such as pouring concrete in an unsatisfactory manner, which makes correction difficult if not impossible would be reason to identify
._ such problems to the licensee during the investigation / inspection.
7,,
g l
- e. Needed Imorevement:
The Investigation Report 498;499/79-19 wnicn cocumentec investigation and inspection results was a large work effort which took several weeks to put together after the investigation / inspection effort ended on January 24, 1980.
Then each participant went back to their respective work locations in different regions where they drafted their part of the report without the benefit of direction.
That is, closer direction would have been beneficial relative to uniform format, style and content.
Also, coordination was not possible relative to notes taken by both NRC personnel conducting the interviews.
Inspectors wno worked in the same areas or who identified similar nonconformances that constituted trends were also unable to coordinate without writing this initial draft.
As a result, I
after the first draft was submitted, it was apparent that the team would have to reassemble to get the job done.
Prior to the Team's reassembling to Headquarters, a preliminary
~
write up of the proposed Appendix "A", Notice of Violation, was furnished to Region IV Construction Branch for comment
}
without the benefit of the report details to suoport the pro-
~
posed Notice of.' Violation.
Solicited comments without full 1
documentation of findings led to less than satisfactory comments i
being furnished back to the team.
Starting off on the wrong foot may have given the impression that the team could not document i
the noncompliances.
The cmunents may have given team members the impression that those cannenting did not understand the full significance of the team findings.
Recommendation:
In retrospect, the team should have remained intact inneciately following the exit until the first draft was completed.
A draft of the Notice of Violation should not have been furnished for comment without a draft of report details.
The draft Notice of Violation could have been submitted without a draft of the details for information only.
H. S. Phillips, Resident Reactor Inspector, STP l
i o--,
,,----__,,,,y
,v.__-
..,-...y
-.--,----,-------.-----v-