ML20234F326

From kanterella
Jump to navigation Jump to search
Partially deleted,marked-up Affidavit Re safety-related Problems at Plant,Specifically,Serious QC Problems & Listed Practices
ML20234F326
Person / Time
Site: Diablo Canyon  Pacific Gas & Electric icon.png
Issue date: 01/16/1984
From:
GOVERNMENT ACCOUNTABILITY PROJECT, PULLMAN POWER PRODUCTS CORP. (FORMERLY PULLMAN, INC.)
To:
Shared Package
ML20234F270 List:
References
FOIA-84-743 NUDOCS 8709230092
Download: ML20234F326 (10)
v • d • e


Text

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

, l I

AFFIDAVIT l I

as '

l My name is I am giving this sworn statement freel , without any threat or inducement, to John Clewett, who has identified himself to me as ,

l is an attorney working with the Government Accountability Project (GAP) of the Institute for Policy Studies (IPS) in Washington, D.C. I have decided to spsak openly about the safety-related problems at Diablo Canyon because I fear that the Nuclear Regulatory Commission may otherwise take the premature action of allowing the startup of Diablo Canyon before it has been shown to be safe to operate.

I L

1

{

I am currently working at Diablo Canyon for Pullman Power Products as a Quality Control Inspector, Level II.

j l

)

1 l

1 I

970914  !

e7092h0092 01A PDR g w ee4-74a t _ _ _ _ _ _ _ _ _ _ _ _

~3 I l 1

5 A.

j As an individual trained and experienced in quality control inspection, I ma l l

have found serious quality-control problems at Diablo Canyon. These QC failures affect safety-related structures and systems that are crucial to the safe opera-tion of the Diablo Canyon nuclear reactors. ,

1 l

My particular corcerns include the following practices: l 4Y\

1. In October or November of 1982, while I was working for Pullman Power Products as a QC inspector, Magnaflux Corporation was brought in by Bechtel/PGGE for a weld reverification program on batt welds on Unit I ruprture restraints using ultrasonic testing. It was apparent after the testing was done, that an unacceptable percentage of the welds were bad.

There were some truly pathetic welds. In one particularly vivid case, I observed a butt weld, which is required by professional welding codes to However, incorporate a backing bar as an integral part of the welded joint.

the backing bar was only held on in this case by the tack-welds meant to hold It was on so weakly the bar in position while the weld itself is completed.

This (no penetration) that a light tap with a hammer knocked it off completely.

weld was on a steel ruptrue restraint that was probably also structural, holding up a pipe rack at Unit 1.

Later, I was told by one of the Magnaflux inspectors that a

-p q

s. . j s

1

,1 j

a PG6E engineer from San Francisco declared that in spite df the ultrasonic l

testing' results, the welds were acceptable, : on a volumetric basis. What this Ii a.

1 neans is that the number of defects in the welds doesn't matter because the re- l maining seceptable weld volume is sufficient. . The problem with this analysis is 1 la i that ultrasonic testing is unable to determine what type of discontinuity exists I

in the weld. For certain discontinuities, such as trapped gas or slag, the But if the weld is internally volumetric size can be a valid consideration.

1 cracked, there is a good chance that the crack will propogate, causing the weld- ]

to fail. For this reason, the acceptance of these welds on a volumetric basis is completely improper,' because the welds could still be bad.

P I am concerned that the specific welding of the ru,ture restraints and pipe supports may not be adequate to insure safety. But beyond that I am concerned that the actions of the Bechtel/PG6E officials who tampered with the ultrasonic testing showed contempt for the very concept of quality assurance, which presupposes reliable testing of components, systems and structures so that they have been honestly shown to comply with applicable principles of quality l 1

assurance.

jh 2. To my knowledge, from personal observation and from discussions with

' others, improperly made studs have been and are still at the present time being welded on to the containment interior liner of the containment buildings for ,

i Units 1.and 2. Because of the crucial safety importance_of the containment liner, one of the few things allowed to be welded onto the containment liner are studs, from which crucial systems are hung, such as the containment spray ring.

^

. w The problem is that many of these studs, which perform a crucial safety- m.

related function, were fabricated on-site using unqualified materials,(ASTM A307 Grade B) that were not designed for the purpose of being welded. As a.conse-1 quence not only have the applicable welding procedures been violated, but the welds attaching these studs to the containment liner are likely to be insuf-ficiently strong and durable to meet design requirements.

The improper studs were made by taking ordinary bolts and cutting off the heads of the bolts, then chisel pointed the end so that it could be welded to the containment liner. Other components were then attached to those studs.

Using these bolts as studs at all is a violation of the plant design and of the concept of quality assurance. Beyond that, however, the welds holding these bolts to the containment liner are in all likelihood weaker than they would be if done with proper materials.

The bolts in question are made of material, namely ASTM A 307 Grade B, that is not qualifiable by ASME 9, which covers the welder procedures for pressure vessels (nor, for that matter, by AWS D1.1, which covers structural welding) .

The procedure used for making these welds specified that studs had to be "P1", a rating of the weldability of the studs. As unqualified material it is dubious that ASTM A307 Grade B is weldable, especially because bolts often contain elements such as sulphur and phosphorus to make them easier to machine, Also, but which would contaminate and weaken any welds done with such materials.

t . .

the bolts are probably much higher in. carbon than is acceptable for weld-ing. Because of this, welding these bolts to the containment liner would create In addition, the l welds- that are not as strong as welds made with P1 material.

~

contaminating elements in the bolts could cause the heat-affected zone of the nr.

containment liner itself to become weaker. ,

Miman buer Mad 5 Even if the material were weldable, 4GE-doesn't have a procedure to weld it, since it is an unqualified material.

The fact that unqualified bolts were used in place of the studs that are required by the design raises serious questions notgonly k p een andabout the. safety of the studs themselves, but also about the integrity of the Quality g Assurance system at Diablo Canyon.

o t

3. Several problems with Diablo Canyon were brought to my attention in y,.'.,L. .'/-

early 1983 by a friend of mine who used to be the assistant manager of the night shift at the H. P. Foley Company until he quit because he was finding so many-The H.P.

problems and so much improper work that had been approved and accepted. 1 l

Foley Company is one of the principal subcontractors on site at Diablo Canyon, and is responsible for the structural, electrical and instrumentation work.

One of the most shocking of these problems concerned the in-core thermo-couples which perform the safety-related function of monitoring the temperature inside the reactor core. Foley's night shift assistant manager related to me how the in-core thermocouple had been installed in rush fashion on Sundays,

6-without the normal quality-assurance procedures required for safety-related NO l hardware. The reason for the rush was that at the time they were installed, PG6E was hurring to finish completing the systems required by the NRC in order to license the plant. In particular, Foley did not treat the in-core thermocouple ma i as Class 1, so that there was no need to maintain material traceability on the work.

When Foley's night shift assistant manager told his management about the

,t was implita%5The das problem, +kay told him to keep quiet about it or else he would be fired.

4. Another safety-related problem that related to the work of the Foley g g L company concerned the " beam clamps" that clamp onto I-beams and support the electrical cable trays, which carry all the electrical cables for instruments- l tion, control, lighting and other purposes.

The I-beam flanges onto which these clamps attach come in two types (M and W),

Because of this, proper one of which is straight-sided and one of which tapers.

practice requires each beam clamp.to be of a type that matches the type of I-beam flange to which it must attach.

However, the H.P. Foley Company bought only one kind of beam clamp.

Therefore, sometimes the beam clamp would fit properly, and sometimes it wouldn't.

l On a walking tour inside the containment, my friend the night assistant manager for Foley showed me a large number of beam clamps that were cock-eyed and ap-peared loose. He said that a lot of the beam clamps could be kicked completely off.

And he showed me that a number of the beam clamps that didn't fit properly This is had been tack-welded onto the I-beams to keep them from falling off.

1 9

, i l

an unqualified procedure; beam clamps are not supposed to be welded. m 2.

Y rough the use of improper materials and unqualified procedures,:the beam clamps'that hold up the crucial electrical cable trays have been installed in as flagrant violation of quality-assurance standards, and are of indeterminate This is a quality. In fact, many of them are probably affirmatively unsafe.

continuing problem - I have seen cock-eyed beam clamps as recently as last week.

' 5. In early 1982, while I was working for Central Coast, I became 4

personally acquainted with yet another instance of a questionable quality assurance I

approach by H.P. Foley Company. They had installed a substantial amount of copper and stainless steel tubing for use in the instrumentation systems of- the plant. One of the procedures used to install that tubing was a procedure that j

joined two pieces of tubing together by placing the ends inside a sleeve, and l

then bra:ing the' joint to assure -a hermetic seal.

At the time the work was actually done, this procedure had not been qualified. After the tubing was installed, Foley came to Central Coast to get the brazing procedure qualified. I was the one who actually did the work of testing the sample supplied by Foley to see if it passed.

Because there was an inadequate amount o# 'orazing within the joint, I re-jected the procedure. Foley's response was to prepare another sample and Again re-submit the procedure for qualification. Again, it was inadequate.

Finally,:

Foley simply prepared another sample and re-submitted.the procedure.

on the third or fourth re-submission of the procedure, I found an adequate joint had been made, and I was able to approve the procedure.

t _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - -

l

a. First of all, the procedure This incident concerns me for two reas mi should have been qualified before the work was done. That way, if the,, procedure failed,. Foley would have been able to alter the parameters of the procedure to i As l increase the likelihood of creating a good joint, before they resubitted it. "

it was, Foley had no choice but to repeatedly resubmit the original procedure, Since it took three (possibly four) f or re-do the work to a better procedure. l l

tries before the sample brazing, which would have been done by Foley's best brazer under ideal circumstances, could pass the test, I have strong doubts that the work on site is adequate. The results suggest that the procedure may only Beyond that, be effective 255 to 33% of the time, even under ideal conditions.

the pressure to avoid having to re-do the work could have tempted the Foley Company, after the first two failures , to submit a sample that was done to a 1 l

different procedure (such as a hotter temperature) to help insure approval.

Central Coast had no way to know whether the sample being tested actually ,

Even if the final system were conformed to the procedure being qualified.

tested, hydrostatically for example, the long-term reliability of the work is dubious at best.

Another problem with qualifying the procedure after the work had been done ders'who have is that all work in a nuclear plant is supposed to be done by w been qualified to that procedure. That cannot have been the case with this important safety-related tubing, that plays a key role in the plant's instruments-tion system, because the procedure itself had not been qualified when the work was done.

N O 6. Another problem, that I have seen recently at the Diablo Canyon plant, concerns the fashion in which pipe support-baseplates are installed throughout e sp .

l the plant.

In the concrete floor are anchor bolts, which secure the baseplates .

To to the floor. Nuts are then torqued onto the threaded ends of the studs.

insure _ level baseplates, steel shims are placed underneath the baseplates, near the studs. There are no procedures detailing the proper size, shape or place-msnt of these shims. After the baseplate has been torqued down onto the shims, grout is added underneath the baseplate.

The problem with this installation procedure is the question of what part 1

l of the assembly is the bearing surface, because calculations of seismic dis-1 I

placement are calculated from the bearing surface.

I From the installation method as it is being done, it is the shims which actually bear the weight. But if the shims are the bearing surface there should If, on the other be specifications as to the size and placement of the shims.

hand, the grout is intended to be the bearing surface, the installation has been done improperly because the plate is not torqued against the grout pad.

This question of which part of the assembly is the bearing surface is far from being merely academic, because I know that at other plants from other The inspectors that work done in this fashion has had to be completely re-done.

fact that there were no procedures or specifications for the location of the shims at Diablo Canyon raises serious questions about the reliability of some of the seismic calculations, and, spparately, raises serious questions about the extent to which principles of QA/QC were followed in the construction of the plant.

s

2. l i

- I m;

These issues are serious examples of QC breakdown, but they are only a few examples of a more pervasive problem. The fact that both the QA/QC department function and the construction function itself are within the same company creates great pressure on the QA/QC department not to make waves. Perhaps due to l

)

these pressures, there have been a number of decisions made that compromise quality. For instance, a memorandum was sent to QC inspectors directing us not j to inspect welds done offsite by vendors, unless there is a crack that can be seen with the naked eye. QC inspectors are, in effect, supposed to wear blinders; if an inspector is inspecting one piece of work, the QC department doesn't want him to notice any problems in other areas of the plant.

I have read the above Ip-page affidavit, and it is true and correct to the best of my knowledge and belief, executed under penalty of perjury, and the foregoing was executed January 16, 1984 at San Luis Obispo, California.

Subscribed and sworn to me tis th9 day of ihm"~ , 1984. De OFF'CIAL  !.63, 9. EEAL 1 t- i WEMTER 

                                                                         ? 'l "       :m x ;:: .ic . cAwcrnA
                                                 ,) ,

f 'C It 'Y.. js r/,N Luti oB;5DO cOU:3 Y f" 't Cu ll g  !. ; 19 ccw r pen un 9.1926 da g ( (, Notary Public My Commission expires d 3 bb - _ _ _ _ _ - _ _ _ _ _ _ _}}

Retrieved from "https://kanterella.com/w/index.php?title=ML20234F326&oldid=3193465"