ML20039D238
| ML20039D238 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry, Zimmer  |
| Issue date: | 08/07/1981 |
| From: | Hofstadter E AFFILIATION NOT ASSIGNED |
| To: | |
| Shared Package | |
| ML20039D233 | List: |
| References | |
| NUDOCS 8112310449 | |
| Download: ML20039D238 (5) | |
Text
.
l
.., c..,
T" t,
..__.__.... m.
=
cc:cc.
pen. rriL c..:..g,g
}
t
--m
'31 OEi! 17 P1 :27 AFFIDAVIT OF EDWIN P. HOFSTADTER S
The stor:r on cable trays starts with Browns Ferry Fire and the NRC report on it.
This report pointed out what happens in a relatively small fire to all the control cables, when they are stacked one on top of the other.
The report concluded by asking C.G. & E. what their plans were, to prevent a similar occurence.
C.G. & E. replied that design work had not started in this portion of the job, but they would advise Sargent & Lundy to follow this recommendation.
Sargent & Lundy failed to do this, with the result that the Zim-mer Plant cable trays are stacked one on top of the other.
This is the same condition that existed at Browns Ferry.
The size of the cable trays was determined through a computer pro-gram.
The program was faulty, the biggest error being in the a-m'ount'of area space required.
This figure should have been doubled.
Trays were ordered 6" by 24" wide.
This should have been doubled, to become 6" by 48" inches wide.
This is the rea-son the cable trays are overloaded.
The NRC found out about the trays being stacked.
Sargent & Lundy and C.G. & E. proposed to make stacking acceptable by fire proof-ing each set of trays.
It was at this point that I became in-Avdo C..T4.
volved.
Husky was working with Aree-on a material applied to a ventilated tray, placed under a regular cable tray.
This ma-terial applied approximately 1/8" thick would expand 300 to 500 8112310449e11h'g.
DR ADOCK 0500 PDR D
r
.g e
-,-ye gre--dg--+-+-9 e
tw+*-p=-
19 -=yg-
--1,-*e.-
-We qN-e-9*+==w--W
- * -+==s9'==0
-wd--39me y,y-**e--eeyv*ewT--v--*g4--4,
+-wC-g 4w#-M'+-
twt e-
--Mr-m-
times at a temperature of approximately 400 degrees.
This materi-al was fire proof and with this much expansion a fire proof blan-ket surrounded the cable tray, providing fire protection for the cables.
Husky was asked to look at the Zimmer cable trays to see if this would provide them the fire protection they needed.
Because of ine stacking of their trays, we could not figure out how to adapt the Husky coated, vented tray to the system.
This is where and when I saw the overloaded cable trays.
Seeing the trays so greatly overloaded, and knowing of the weak material problem, plus the bad welds, I felt compelled to report my observations to the NRC.
I was so naive in this respect that I honestly thought the NRC was supposed to protect the general public.
It was an extreme i
shock and disappointment to find that they are far more concerned with covering up embarassing situations and devising " fixes" to nullify a given condition.
All this accomplishes, in the long run, is to prove the old adage "two wrongs don't make a right. "
In this case the NRC is seeking to prove that many wro'ngs will make one right.
The story on fire proofing these trays becomes nearly fantastic.
fdesed BiP),
A material developed by Babcock and Wilcox was processed.
Bab-cock and Wilcox tested it and had a U.L. man as an observer.who reported what he witnessed.
His observations became a U.L.
report with a number.
Babcock and Wilcox flaunted this U.L.
I T-
. 7Y -
i k
4 3
report number as though it represented U.L. endorsement.
Read-ing the actual report reveals that the " test" was a " farce".
The I
NRC had doubts to the extent that they asked Sandia Laboratories to devise another set of tests.
Sandia came up with still an-f other " farce".
In order to give the test weight, Sandia had U.L.
(ftJ>) f,d}4 perform it under the hjown stipulations.
The material failed so miserably that the U.L. report states, "the KA0WOLL material burnt like the wick of a candle". This'is a direct quote.
The material was approved, however.
While this testing was going on, the NRC, Sargent & Lundy and C.
G. & E. worked out a "fix" on the overloaded cable trays.
The NEI standard for loading of trays is conditioned on area, with 60% of area being the maximum.
Their ingeniously simple fix was to put special 4" added sides to the 6" tray to make it look like 10" tray, thus complying with the 60% space limit.
This com-pletely ignores two important aspects.
First,10" cable tray is made of much heavier material thn 6" cable tray.
Where 6" o
tray can be.094,10" tray is usually 50% more, at.135
- Second, is the factor of load capability.
Zimmer tray was rated at 40 j
pounds to the square foot, with a safety factor of 2.10.
This means the maximum load of the tray should not exceed 82 pounds.
i i
I have obtained cable loading data (which is not completely cur-rent) which shows the actual weight of the cables in the trays to be over 100 pounds per square foot.
c Now we come to the area of welding in this bizarre chain of I will first show the recommended settings followed by events.
the actual settings used.
-4
~
Recommended Setting Actual Setting D.C. Volts 20 30 D.C. Amps.
200 300 Volt Control 40 30 Slope Control 8
12 These high settings produce welds with poor fusion, and undercut-ting the extremely high heat actually " burns" the parent metal, p
causing crystalization of the molecules immediately adjacent to the weld.
This crystalization, once it,has started, continues on, so that in 5, 10, 15 or 20 years the' weld breaks, due to
" metal fatigue".
The only way these welds can be checked with any degree of accuracy is to section the weld and the adjacent parent material, and check microscopically for signs of crystalization.
Just as day follows night, you will find this crystalization.
In respect to welding at Zimmer and on the cable supports in particular, NRC Peoples Survellanco Inspection reports detail numerous instances of welding being done by non-certified welders.
This is bad for two reasons.
First, Ohio State Law requires welders working on new construction to be certified.
Second, in most cases where a welder is not certified, he is not qualified.
Again, an day follows night, you will have many bad welds.
l Also shown in the NRC Survellance Inspection reports is an al-most continuous problem relative to the proper control of weld-ing rod.
Again, inevitably, you will find welds made using the wrong filler material.
Then, the combination of unqualified welders using incorrect welding rod nearly 100% guarantees a multitude of weld failures.
l
5 Most of the above can be substantiated by documents in the Pub-lic Records access files.
On visits to coal fired generating Plants I have found that none exceeds the 60% load by area re9uirement.
I have found that the NEI repuirement is based on safety because of the heat dis-sipation needs of energized cables.
This extreme overloading at l
Zimmer greatly increases the heat dissipation problem.
This could in itself be a critical factor which must be evaluated.
Basically, if you add up all these items and their interrelation-ships, you are forced to recognise a completely unacceptable con-dition now exists.
There is no one or multiple "fix" that can
(
ccrrect these basic " wrongs".
No other conclusion is possible.
The factc we cannot ignere are as follows:
1.
Stacking of trays which nearly caused a disaster at Browns Ferry is basically non correctable.
2.
Cable Tray ordered was only 1/2 the required size, resulting in crowding all the cable into trays 1/2 the necessary size.
This is how the overloaded tray problem resulted.
Over-wc:9ht $J%~,
loaded trays carry more w 4 t than the safety factor allows, also creates a heat dissipation problem.
This is also bas-l ically non correctable.
l 3
The use of special sides to give the illusion of " unused" space, together with the use of totally inadequate fire l
proofing material should be investigated for criminal fraud.
4.
The bad welds and inferior material used are relatively minor compared to numbers 1. and 2. above.
l
1
,s
_g_
These are not opinions but facts mostly taken directly from the Public Records file.
The critical part is the extremely close relationship these items have to each other.
This is a compound-ed problem with the direst of possibilities inherrent.
k E. P. Hofstadter 8 81 O
.m.w-g.g 4
l l
l l
9
. = - n ; _ - -. _
_ _ _.. _.. _., rntrr;_ _.._..
- 72,.,,
._..., _ _ _. _ _