ML20207G790

Phenolic Chemistry:Brief Overview

ML20207G790
Person / Time
Site:	07109019
Issue date:	12/31/1987
From:	Sullivan J External (Affiliation Not Assigned)
To:
Shared Package
ML20207E074	List: ML20207E066 ML20207G790
References
NUDOCS 9903120165
Download: ML20207G790 (34)

Text

..

JM-20-99 ED 08:52 M F M NO.

P. 02/21 i

k PHENOLIC CHEMISTRY:

A BRIEF REVIEW l

t l

JOHN SULLIVAN SCHENECTADY CHEMICALS, INC.

FEBRUARY 18, 1987 ADOCK 07109019 {4~*

9903120165 990210 f

PDR

-g PDR I

JAN-20-99 ilED 08:53 M CAX NO.

P. 03/21 J

(,

Phenolic resins are the products of condensation reactions between phenol and either aldehydes or ketones.

I Phenol / ketone resins are not produced by Schenectady

\\

Chemicals, and therefore, this introduction will be i

\\

limited to phenol / formaldehyde resins.

A.

Functionality and Reactivity Phenols are characterized by the presence of a hydroxyl group attached to a benzene ring.

In

[

general highly reactive phenols form condensation products that cure rapidly to hard insoluble resins, while phenols of low reactivity form condensation i

products thet show little or no tendency to harden.

The major structural feature of a phenol that determines its reactivity with aldehydes is its functionality (FIGURE 1).

Since the hydroxyl group is activating and ortho, para directing, functionality may be defined as the total number of unsubstituted positions on the benzene ring that are in ortho or para position f

to the hydroxyl group.

There are many types of phenol which vary in reactivity depending upon the number and type of substituents and also on the position of those substituents Islative to the hydroxyl group.

I I

i

0 2--

l k-Phencri has a functionality of three:

it will react at the position para to the hydroxyl and at the

)

two positions ortho to the hydroxyl.

The two meta positions will react only under severe conditions not generally found in the resin reactions, and can be considered inactive.

The trifunctionality of phenol allows for a three-dimensional crosslinked structure i

l and an ortho-para directing substituent in the meta 1

position often enhances the reactivity at these points.

Bowever, if a substituent is in the ortho or para I

position, the phenol becomes difunctional and when I

reacted under usual conditions, produces linear,

l uclecules without crosslinking capability.

FIGURE 2 illustrates the structural features of resin molecules l

based on a) trifunctional and b) difunctional phenols.

Another important factor in the selection"of a phenol for resin production is the size and structure of the substituent grcups on the benzene ' ring.

Long l

side chains on a phenol generally change the solubilit'y l-characteristics of and impart plasticity to the resins made from them.

Butylphenol, used in resins for contact adhesives and octylphenol, used in carbonless copy paper atid rubber compounding resins, are examples i

of this.

FIGURE 3 compares some consonly used phenols 1

with phenol itself in terms of differences in functionality, l

reactivity, and properties of the resins produced i

i j

from them.

)

ffg1'mts

,,w' m

++.m g.

---p T-

-i+-7s w -*--

JME20-99 WED 08:54 N1 FAX N0.

P. 05/21 h

. )

Monohydric phenols show very little reactivity toward formaldehyde in the absence of catalysts.

In the presence of catalysts phenols that are not structurally hindered will generally react with formaldehyde.

Although certain other aldehydes may be used, formaldehyde is preferred because of its high reactivity and freedom from side reactions.

FIGURE 4 compares the properties of other aldehydes, used in resin manuf acture, with formaldehyde.

i B.

Resin Classification Phenolic resins are divided into two main classes depending upon the type of catalyst used and.the, rat $o i

of phenol to formaldehyde.

If more than one mole of i

i forwideh5rde is reacted withs one mole of phenol in

~

the presence of an alkaline catalyst, the resulting resin is called a RESOLE.

If less than one mole of formal-l dehyde is reacted with one mole of phenol in the j

presence of an areid catalyst, the resulting resin is called a NOVOLAC.

C.

Resoles With an excess of formaldehyde over phenol under basic conditions, phenol alcohols predominate as the I

initial product.

The reaction mechanism for this

{

first step, ca[ led a hydroxymethylation reaction, is illustrated in FIGURE 5.

As phenol alcohols are formed, they can undergo condensation reactions to form larger molecules linked together by either methylene or methylene ether bridges.

FIccRE 6 illustrates this step.

JM-20-99 IED 08:54 M FAX NO.

P. 06/21 4

k-As the reaction continues, these molecules become sore compicx, increasing in molecular weight, but by cooling, the reaction may be stopped at any desired point short of a final " cured" product.

To resume the

' reaction, the temperature is raised or an acidic catalyst is added.

Since there is no sharp break in the reaction, swh resins have b6an termed the one-stage (or one. step) resins.

Because of the presence of terminal methylois, resins comprised of trifunctional phenols will advance with heat to a permanently in-soluble and infusable state and are thus called thernesetting resins.

D.

Heat Reactive Resins

~

" Oil-soluble" is an old varnish industry term used to identify those resins now cocmonly called heat-reactives.

They differ from other resoles in that they are made with a para-substituted alkylphenol.

As mentioned before, this para-substitution makes the phenol di-t functional and thus makes crosslinking extremely difficult.

With the addition of heat these resins' can advance in molecular weight but generally only in a linear fashion and thus cannot reach the permanently insoluble and infusible state of thermosetting resins.

e

U JM-20-99 WED 08:54 M F M NO.

P. 07/21 L

E.

Novolacs With a solar excess of phenol over formaldehyde under acidic conditions the initial product which predominates in the condensation reaction is the dihydroxydiphenyl-alkane type compound.

These may form chains usually l

from one to five units long and differ from resoles 1

in that they are phenol terminated rather than methylol terminated.

FIGURE 7 illustrates the reaction and i

mechanism.

Their physical structure may be altered with heat (i.e. softened and hardened) but 91thout an additional curing agent, they will remain permanently soluble and fusible and are thus termed therinoplastic.

With the addition of a methylene (-CH2-) donor such as Hexamethylenstetramine, novolacs may,be made insoluble

• and infusible (cured).

An acidic or' basic catalyst may be added, if desired, to speed up ( 2

reaction, Because of this second step, they are some-times referred to as two-stage (or two-step) resins.

FIGDRE 8 summarizes the major differences between resoles and novolacs.

e

l JM-20-99 IED 08:55 M FM NO.

k08/21

~

~

(;

F.

Phenolic Resin Reactions Once formed, phenolic resins are capable of undergoing reactions with many other substrates.

Some j

of the more important ones, from a commercial standpoint, are as follows:

1.

FIGURE 9 illustratos the use of epichlorohydrin l

to prepare epoxidized novolacs and the use of phenolic resins as epoxy co-reactants or " harden-1 ers."

2.

FIGURE 10 illustrates the reaction of an isocya-nate with a phenolic hydroxyl to form a urethane.

3.

Phenolic resins will react with many compounds cotitaining double bonds such as drying oilc, terpenes, and unsaturated elastomers and are, thus, useful as components in adhesives, coatings, and as rubber vulcanizing agents.

Over the years, i

there has been considerable disagreement as to th; axact mechanism by which this reaction pro-coeds and FIGURES 11 and 12 illustrate two of the more widely accepted ones.

At present, the i

ionic chain reaction mechanism is generally agreed to be the most probable.

i l

I

JM-20-99 ED 08:55 M FM NO.

P. 09/21

~ j-FIGURE 1 FUNCT10liAllTY OF PHENOL d

OH I

\\ p- /

i 1

1 V

4 l

i h

j 1

l TRIFUNCTIONAL i

i OH i

i n/

i I

f Y

a h

R

\\

DIFUiiCTIONAL-1 1

l J

r

FIGURE 2 SIRUCTURE OF RFSIli M01FCUIFS l

1 1

OH A

FO o"

O L

\\!

/\\

o\\

O HO N/_

OH CROSS-LINKED STRUCTURE (TR! FUNCTIONAL PHENOL) l I

OH OH OH O

O O

Y Y

V R

R g

O LII1 EAR STRUCTURE (DIFUNCTIONAL PHENOL) i 4

l

'J M-20-99 WED 08:58 M FM NO.

P. 11/21 FIGURE 3 9

l DTHFR PHEN 0lS

(,

COMPARISONS WITH PHENOL 1.

fJtEEDL1 OH OH OH

[O O

d

[O, C H 3

V VCH Y

3 CH 3 ORTH 0 META PARA il ORTH 01 PARA - D! FUNCTIONAL (NON-THERMOSETTlNG) AND SLOWER REACTING

- TRIFyNCTIONAL (THERM 0 SETTING)

META "0!L" SOLUBLE LORGANIC NONPOLAR SOLVENTS)$ND F MORE WHEN USED IN COMBINATION WITH PHENOL:

TO,UGHER WHEN CROS$ LINKED

~

BETTER MOISTURE RESISTANCE USES:

COATINGS, LAMINATES, MOLDING COMPOUNDS

~

~

2.

OTHER ALKYLPHENOLs OH OH OH b

d O

• O O

M Y

CH 1 IT P-T-BUTYLPHENOL P-T-0CTYLPHENOL P-NONYLPHENOL ALL ARE DIFUNCTIONAL AND REACT EVEN MORE SLOWLY AS ALKYL CHAIN LENGTH INCREASESJ Cl; MORPHOUS AND HAVE LOWER MELT PO[NTS

--+ C9 :

RESINS BECOME MORE A OIL SOLUBILITY INCREASES RESISTANCE TO MOISTURE INCREASES USES: -VARNISHES, RESINS FOR CARBONLESS COPY PAPER, RUBBER COM-POUNDING, ADHESIVES, AND SURFACTANTS.

JAN-20-99 ED 08:56 (d FAX NO.

P. 12/21 FIGURE 3 (CON'D)

OTHER PHFiiOIS

\\,

3.

RESORCINOL OH I

O V'O H TRIFUNCTIONAL AND MUCH MORE REACTIVE RESINS SOLUBLE IN WATER AND POLAR SOLVENTE USES:

CORD AND FABRIC ADHES!VES, COLD-SETTING ADHESIVES, WOOD GLUES 6

14.

BJ.Sm1FNot

-A HQ O -c O

os 1CH 3

THEORETICALLY TETRAFUNCTIONAL AND QUITE REACTIVE OFTEN USED IN COMBINATION WITH OTHER PHENOLS USES:

COATINGS, ADHESIVES L

JMF20-99 ilED 08:56 M FM NO.

P. 13/21 FIGURE 4 OTHERALDEHDES COMPARISONS WITH FORMALDEHYDE

(

ACFTAtDEHYDF OH OH H[

MUCH LESS REACTIVE C

Q_

RESINS ARE:

g Ij MORE DIL SOLUBLE CH LESS BRITTLE MORE THERMOPLASTIC ETHYLIDENE BRIDGE USES:

RUBBER RESINS, ANTIOXIDANTS 2.

nuTYRAtDEHYDE

~

MUCH LESS REACTIVE-O,H H

OH RESINS ARE:

EVEN MORE 0!L SOLUBLE A

l Q

C

~

4 I

l EVEN LESS BRITTLE MORE THERMOPLASTIC CH V-2 USES:

RUBBER RESINS, ANTIOXIDANTS CH 2

1 CH 3

l l

3.

FURFURAL FAIRLY REACTIVE H

H OTHER REACTIONS:

t O-FURAN RING OPENING O

POLYMERIZATION THROUGH DOUBLE BONDS OH RESINS ARE FLEXIBLE AND HAVE LOW MELT VISCOSITY UsES:

FRICTION MATERIALS, ABRASIVE BINDERS, ACID RESISTANT COATINGS i

m

JM-20-99 WED 08:57 M F M N0.

P. 14/21 FIGURE S PHEH01-FORMALDEHYDF REACTION UNDER ALKALINF CONDITIONS

(,.

STEP 1:

HYDROXYMETHYLATION REACTION FORMATION OF PHENOXIDE 10N:

g$

e e

l

+

HO Q ++ V i

+UH z

m

.v

~

v-e l

0RTH0 C-ALKYLATION:

O O

O

.o

[ CHg-O d -CH OH 6@

se H

O 2

C H.;- O m

m c

PARA C-ALKYLATION:

6 O

o 9

Al Al A

Se 3e

+l

+ C H - 0 l

O 2

H><CH-Oe Y

cM,on z

JM-20-99 WED 08:57 M F M NO.

Po 15/21 FIGURE 6 l

PHEi10L-FORMAlDFHYDE REACT 1011 I

llNDER AlKAllNF CONDITIONS l

STEP 2:

c0NDENSATION POLYMERIZATION GENERAL REACTION PATHS:

OH OH

-CH - O-CHp 2

+

HO 2

v V

METHYLENE ETHER BRIDGE

'c go,

OH 3

C H,O H a

v

1. o.o OH OH

-CH-d 1

V()

+

H*O

+ CH*O v

METHYLENE BRIDGE l

l l

O

JM-20-99 WED 08:57 M FM N0.

P. 16/21 FIGURE 7

('

STR0]!GACIDCATAlYSIS_

+

FORMATION OF HYDROXYMETHYLENE CARBONIUM ION FROM METHYLENE GLYCOL:

H@

g MO - C H -O H i CH

-OH

+

H g

g 2

ADDITION OF HYDROXYMETHYLENE CARBONIUM ION 'TO PHENOL:

, OH '

OH OH

>d O

.+@CH -OH <

< [CH O H + __ h-C H O H 2

+ H 2

y s

2 FORMATION OF BENZYLIC CARBONIUM ION:

OH OH

,CH OH

@.4

,CH 2

O

+H y

+

HO 2

e REACTION WITH PHENOL:

e OH OH OH OH A

O'

+

O

'CHp 0 v

y v

DlHYDROXYDIPHENYLMETHANE

l JAN-20-99 WED 08:58 M FAX N0.

P. 17/21 l

FIGURE 6 i

.(iENERAL TYPES OF PHFliOLIC RFSINS l

l l

FORMALDEHYDE CATALYST TYPE FORMALDEHYDE q

q l

PHENOL PHENOL ACID UNCONTROLLABLE CONTROLLABLE RtACTION REACTION NOV0 LAC

\\

)

ALKALINE CONTROLLABLE.ONE-HIGHLY ORTHO-STEP RESIN:

SUBSTITUTED NOVOLAC R$50LE (PHENOL) i "H

T REACTIVE ESIN ALKYLPHEHOLS e

4 4

e 4

t i

r i

JM-2()-99 ED 08:58 m FM NO, P. 18/21 FIGURE 9

REACTIONS WIT!! EPOXIES

($

l'.

With Epichlorohydrin C H s.

pcHz

/

O I O l N CH NCH CH s.

bH z

OH oy O

b W

~'

M

~

H c --- C H-C Ha,- c i -

y y

i V

O V

V Novolac.

Epichlorohydrin Epoxidized Novolac '

Resin Properties:

Righ Strength, Stron'g Adhesion, Excellent Dielectric Properties, Good Oxidation Resistance Us as contings, adhesives', electrical applications 2.

With Epoxy Resins OH + H c-cH-cH -o O "e

~

\\/

an. AMIN 65 O

Nc vo: Lac Epoxy Rasin O

o-c" -c"- c" -o. O OM Etherification Uses: coatings, adhesives, electrical applications e

JM-20-99 IED 08:58 m FM NO, P. 19/21

' 'F1GURE 10 RF. ACTION WIT 11 ISOCYANATES l

O H

l O ~ +

o = e= 8 O O -o ee O Monomaric Isocyanate Urcthane Phanbl or Novolac t

[

\\

Addition reaction t

no neutral molecule split off

' reaction is reversible

• I reversion temperature dependent on 1r..iling point of phenol.ic compound 4

IResin Uses:

single component polyurethane coatings polyester powder coating crossilinkers fast curing foundry " cold set" resins 1

.e t

e e

i j

}

'a a.

s a

f r-

JM-2KJ IED 08:58 M F M HO.

P. 20/21 FreunE 11 RrActzew wIrit counts nonos Unsaturated Elastomers, Drying oils, Rosin

~

(.

.1., Quirmne Mathide Mechanism - Van der Meer i

.~

e

. Es c4

\\/

O-OH O,0 0

ye 9

e

+

n

++

e

- Hs0 R.,' V ' c H o H k,,

Cg a.

. g,yscMs g

c f\\

g, g

s Quinone

. Methide i

R s.'*C H3

\\/

,OH 9c l

g >-

seg,-c-g I

R Rs R. a.

A, % c - c H AH c ::::: C H z.

O ' C H 3.'

3 O

i

~N V\\CH[

'N g,

~

R R3 3

Chroman'Rin9 Double Bond Shift e

!JM-20-99 IED 08:59 M FAX NO.

P. 21/21 l

FIGURE 12 RUBBl?R VULCANI1ATION Ionic Chain Reaction Mechanism - Giller

1..

oH.

g.

O y

-CHaOH 1

L w

-cy, p.

+ H SnCla(oH).

+ H Sncli(oH),

+,H.o a

y.

y R

R Lewis Acid Carbonium k

Ion Double Bond Attack Rs I

OH Rt CH OH'

' O C' 0"3 3

\\/

l

-cH,+,

ll g

-cs, c_g c

w I

1'V A

.Y R>

R, R

H 3

g, e

j

'4

\\/

on c

~

- w

-cH - c g

n.,.

Director -

l Office ofNuclear Material Safety & Safeguards February 10,1999

(-

Page1ofI i

GE's Questions to Eco-Pak Dated 2/3/99 and Eco-Pak's Response Dated 2/4/99 i

4

i q

i

(

FAX 3 February 1999 3

j To:

Mike Arnold - Ecopak j

From:

Charlie Vaughan - GE Nuclear Energy i

Subject:

BU-7 Report to the USNRC due February 11,1999 a

f Mike -

\\

l Rick Foleck and I are coordinating the report to the USNRC. At the present time j

we are taking stock of the inventory ofinformation we have with which to construct l

a report.

l At the present time, it does not appear that we have infomution adequate to reach positive conclusions in the report. In going through the technical information that I j

have and that has been provided during the investigation,I have found a number of things which I am not sure about and possibly am missing some infonnation. I j

thought you would be in the best position to answer these questions.

l i

If the information is available,I don't think it should take long to respond. Ifit is

}

not available please tell me. Remember the sooner I get it the more chance we have j

for a favorable report to NRC.

l i

Phone: 910 675 5656 FAX: 910 675 6350 E-Mail: charles.vaughan@ gene.ge.com.

I have the faxed document saved as WORD 6.0. If you have an e-mail address I can send it to you electronically and you can work with it online and simplify the i

process of responding.

i

}

We understand that the changes in phenolic resin began when Union Carbide e

discontinued their BRIr2760 phenolic resia. A review, some foaming test and determination of density, strength and thermal properties lead to a decision that Schenectady International HRJ-2590 was an acceptable substitute. When pitting was observed using the new resin it was traced to chloride - a result of the final neutralization of the resin product. Alternates for the BCI used in this step lead to the use of oxalic acid for the neutralization element. This new resin was designated HRJ-11825 (Iow chloride < 200 PPM). The only tests done en this were chemical tests for chloride content in the resin and in the foam, physical characteristics and thermal properties.

Is this history correct? Have there been any other tests or evaluations done to show that the two resins are equivalent? If so please identify them. Note also:

m

i j

W Was zine found as a major component in the paint? It seems not as well as a e

jk>

couple of other expected components. Formulation was compared to the MSDS, however, the specification for the paint is not mentioned. What is the i

specification?

i Most of the work done on the paint coatings speak of a " zinc epoxy primer" and l

they give an alternative if other coatings provide better protection. The current i

coating seems to be PPG Primer DP-40 Epoxy Primer and DP-401 catalyst.

i l

Law Engineering is reported to have tested this paint for protection in free

)

chiedde shuations but no other conditions were tested. What work was done to demonstrate the PPG pdmer is equal to the zinc epoxy pdmer and what was the i

extent of this testing? What are the important measures to judge equivalency?

i The Law Engineering report includes very conclusive evidence related to zinc e

j forinate hydrate being present. We have a similar finding from the university l

laboratory here in Wilmington. Has Law been advised of the zinc plated bolts and the extensive surface corrosion on them? Have they studied the chemistry of j

the formulation and the chemical atactions and byproduct components from the foaming and cadag operations?

Law Engineering postulates that zine formate hydrate is formed by a reaction of

{

e l

zinc carbonate. The question is, are there other viable mechanisms such as the i

direct reaction of Zn with formic acid? The Law report went on to discuss uses j

of zinc carbonate and formic acid but never drew a parallel to the current situation - is there a reason for this?

The Attachments to the report are not clearly marked to indicate the type of e

analyais and in some limited cases the significant points are not identified as follows: Printout #3 does not iddify the peaks, Printout #4 clearly shows the zinc formate signature, however, there is another signature present that is not identified - what is it?, Printout #5 does not make mention of the formate functionality in the 2000 - 1000 wave number range - why was this omitted?

1-3 i

The specification NCI-BU7, Rev. O calls out the maximum chloride content as j (j

<300 PPM, however, all the information sent to labs for evaluation and used in j

reports uses the value of <200 PPM - which is the correct value?

j We understand that the ori inal Union Carbide surfactant le530 (silicone) was j

h replaced by Union Carbide Y-6663 (silicone). The current NCI-BU-7, Rev. O i

specification specifies Dow Corning DC-193.

i j

Is this history correct? Have there been any tests or evaluations which show the j

equivalency of these compounds for other than the physical characteristics of the foam (density, strength, thermal)? If so we would like references and the results.

4 l

We und2rstand that reagent grade boric anhydride became difficult to obtain, j

In searching for an alternative, practical grade boric anhydride by Eastman l

Chemical was determined to result in good foaming (meeting physical j

specifications of the foam) and technical grade or better boric anhydride was

{

added to the specification.

j Is this history correct? Have there been any tests other than physical tests of j

density, strength or thermal performed? If yes please provide the reference and results.

l!

We understanding that Owens-Corning fiberglass rovings # 805 (HIS) was i

e l

changed to # 833. The current NCI-BU-7, Rev.0 specifies Owens Corning 405 1

or equivalent.

l Is our understanding of these changes correct? What is the traceability for these changes? What tests have been conducted to demonstrate adequacy of the

{

rovings? With regard to "or equivalent" what parameters / characteristics are important to making decisions with regard to equivalency?

Is the HRJ - 11825 product categorized as a liquid resole type resin (base-e catalyzed condensation product of phenol and formaldehyde) or a novolak resin (acid-catalyzed condensation product of phenol and formaldehyde)?

What is the basis for the farmulation listed in the foam specification NCI-BU-7, Rev. 0? What function t oes each component play in the foaming and curing process? What are the byoroducts of the chemical reactions occurring during and after the foaming and curig process?

Your letter of February 1,1999 indicates that the Law Engineering Services e

report is " final". The report itself says " DRAFT". What is the status of the report?

4

a-a-ce s iaisasieia:Ga Ri~iio~~~~ ~ ~~~ ~~~~~~[hi sG tat;'^^~~~~~~~~

i ~~~^

~

i, a,

I

~

1251odsetMy,Ishe A Elmshshion.M $7645 05A Iti:425-H54211 800 2212465 Pas:423 5454007

(,

M "T wwmf,way

%=m ru$ste s

i l

Febmary4,1999 Mr. Charlie Vaughan Nuclear Fuel & Components Manufacturing GE Nuclear Energy P.O. Box 790

)

Wil=i=*= NC 28402 l

Subject:

Response to your 2/3/99 Fax regarding the BU-7 Report to the NRC l

Dear Mr. V-ha.

f i

As previously explained to everyone involved, the testing and technical questions that ESP has oflined to investigate are in process and the response is dependent upon the engineering Brm and lab involved. We are working with a NACE certined conosion engineer, Lakshman Ranranam who is out of the office until Monday, and it is probable that we will not get a few of the answers j

to your questions regarding the paint until that time.

i l

Regarding the foam component questions (the first four bullets), these have all been answered l

previously and accepted by the NRC aAer the August 1,1995 physical testing of a GE package.

j However, here are the answers as follows: 1) The history of the resin is correct, and other tests i

and evaluations were not naadad 2) I believe the history of the surfactant is correct, and I am i

unaware of any other tests or evaluations conducted. 3)'Ihe history of the boric anhydride is I

conect, and I am unaware of any other tests conducted. 4)'Ihese changes are corrtet. It is my understanding that each product was 1/4" fiberglass strand compatible with phenolic resin.

Again, the physical testing was perfonned August 1.1995 on a OE package, and the nsults were i

accepted by the NRC.

3 l

The HRJ-l1825 resin is a base <stalyzed corvindon product of phenol and formaldehyde. For j

i more detailed information, please scre-e CM 4 =-:^='y International infonnation previously j

eent or contact Todd Makenzie at (518) 370-4200.

'Ibe basis for the NCI-BU-7, Rev. O foam fonsulation is the original USAEC SP-9 specification for Hre retardant phenolic foam. My understanding of the function of each g mi-:er is as I

follows: 1) phenolic resin is the structure of the foam; 2) surfactant relates to the texture, i

g ny the closed cell pic, 0s; 3) 6 eon is the blowing agent; 4) oxalic acid is the catalyst; i

5) boric anhydride gives the foam best for the reaction and curing process; 6) fiberglass rovings l

is for additional strength To verify these functions and to obtain the byproducts of the chemical l

rWS. please refer to the Schenectady International information previously sent or contact i

Todd Makenzie at (518) 3704200.

I i

f s

u-c-uu o e son % se.-a. ce.c.==.

3, Mr.Charhe Vughan Pass 2

(;

Febnmy 4,1999 Rick Folock has a copy of the final report fmm Lew Rap =~ing.

The eighth, eleventh and twelfth bullets will need to be answered through Lakshman Santanam at Law Fap=aadng, and this will most likely occur Monday.

PPG DP 40 is a zine epoxy primer, therefore no work was needed to he te equivalency.

s Law Engineering was advised of the atansive surface corrosion of the zine plated carbon steel bohs. They, however, have not studied the chimy of tbs foaming and curing operations in thisparticularincident.

We hope this i xmation is bene 6cial to your report. As soon as we have fader information, it willbepassed mg to both OE and BWX Technologies. Ifyou have any questions or g_ ___ _.a. plee

'".mknow.

Best regartin, h

w.

Heather Little Regulations Specialist cc:

Pruston Foster, BWX Technologies, Inc.

File i

l l

.\\

\\

l; i

l

l Director Office of Nuclear Material Safety & Safeguards February 10,1999

!{-

Page1of1 l

l l

i i

1 i

i i

)

Eco-Pak Letter to NRC Dated 1/29/99, Reporting the Results of Their Inspection of an Unassembled 55-Gallon Drum l

l l

i I

f I

l I

i Eco-Pak 135lodent % S& A ESP Sl:ecialty Packaging Einadon,W 37643 USA j

Dvision of CBC Tel:423-543-4211 800 221-2465 Fax:423-543-6007 intermediate Bulk Containersfor armicals & Liquids

• Y'*

j{'

U and00 Packaging 2

I l

January 29,1999 Mr. Cass R. Chappell Section Leader, Package Certification Section Office of Nuclear Material Safety and Safeguards United States Nuclear Regulatory Commission 11545 Rockville Pike l

Rockville, MD 20852

Subject:

Empty BU-7 Outer Shipping Container

Dear Mr. Chappell:

As requested, we visually inspected and photographed (photos sent via Federal Express) a 55-gallon drum prepared by Packaging Specialties for the 1995 Babcock & Wilcox contract. It had been stored in a concrete block building behind the main plant in Elizabethton, Tennessee. The drum was found without a lid, however there was also a lid stored in the same area to inspect.

There was no discernable rust on the inside of the drum or the lid. However, the inside paint had a rough texture. We have decided to get this examined to determine what may have caused the roughness. It could very well be dust or dirt, since the drum was not scaled.

If you have any questions or need further information, please let me know.

Very truly yours, l

h) b M.

William M. Arnold President l

W M A/hl e

Enclosures i

Cc:

Rick Foleck, GE Nuclear Energ-Preston Foster, BWX Technologies File l

l

r-----

w-

---

w---_-

,.-,m,-.

--m.g,e m, %4 c m

m_.

g s

e 4

'L l

m 8m:

N

. s. gs -

pi[$[f'fd!hff 4ff,$%mm khlltl$,b,$

g 1

\\.

Y 90 %,Y

%5 Y i

w

'E w,2 ~'e - e.

a t. #

a;# u cu'-

\\

.y. \\A N

k

\\

N

\\

i

\\

\\

j

\\

\\

y %..

f 3

{

{$

ii

.4 '

, i l

4 k

2:

_j

' ;-2 h

u-

/

aJ4ehA.m

h. m a 04.Er 4d N e mL4ikhfW

-memm.

--4'.,u--w--so--= = = -

i i.

I h.,

i 4

At l (.

'i 4

g.

...e a

l ar

y..

.g n

3 4

)

p.

_3

~ _

~

w

- ~

e j

aAk-

/

'Q g

IIM y

I

.... ~.

.'f

,s h *,

'a l

s f

3 a

u s ; r g w"t X '.J+3 @

w-1-

. ~

C., $f. "'it p f,, '

ig g

3 55

n. w.v.

m[$g'5g'y p q:Q.q

1. N1/>@ i '

A'

~

t

,n,,a$h'M'D I

.w

. U nt?. N.:n - n.:l.

't.:

q aw..c: :

kp g 3 ; ; % Q ; g e Q g3 q?N:/'

f'jfWW.

.gwww x'/..

'.)., :. -..

en y. Q'3?=iby'w+.

g /. e.Q;l".!;.,.i',k :

n1pA i

- <;-Wf.rr J 4..

,c

. si s.

, k;l.pn. q -: I' '4%

M,}3[b; b,9;.

v.

i / e [f:N. N, y,:-%p--

bin y

.L.,

s,g.,...

,.my f f y..w(b;

1. .,:.N.?,p$r. k, U. kh~nn:h A m$ h..

TF y

.s 4

.,. p. tt;$

,w,:.

c.

..,.4..a.,

p

~

e, ;

. y:.,

.x

--N; y

-f 6,-

j * '4 j

'{

y..

h

/.1

,r'

..f:1.y).. :t.

i iAw

'v t Q.il.,.

l'[ f ' )

5-tw.(5 x

s.-

3,,

,,a s

-)

g...

.';*1

..1.m y,.

3'

\\ M' s

N.

eo N

,b.

u.

.f$; /,. -

.j

?p}.

-. ; nj,. i

,.,' T s

y,.,$, % j.. j

., *b ; a s

m.%

.$-3 c.*-3 1; $. -.

A.

r,w.,

b

.L

?

-.--.w-.--

---

wrw.

.us

--_esm.r ww em aammau.a m a emsweehem wmmeur4w h-

---

4m4.pze._m,"

- ^ ^ - ' ^ -

'mm,rm, i

1kl 4

i l

I k

i 4

I l

4a-.

pywm, - -

<r, :

• W,.

7?.,

.~

"g

,..e

• w 6 s

N x

g.'; _r $N 2

g-I,

,t;-[

r :-

t

,V.'

• a

" lN,. '-

j -

v

}

.j

'5 -

'3..~s n

Qlll: * -

-es s,tay.

/

-- q;. ~

)g"

\\

n,.

• k 1.

I t) p'1 p i Qm... m,o,e -

l i

i 4-i b

3 l

1 i

Ii El.

Director l.

. Office of Nuclear Material Safety & Safeguards

{

February 10,1999

(}

Page1of1 i

T r

r

)

1 l

l l 0 l

l GE's Discussion of the BU-7 Package i

\\

I i

l l

I I

P l

1-l i

i

__._._____.________m l

l l

i l

l b.

BU - 7 DISCUSSION L

i Background Discussion ne BU-7 package was developed around US AEC SP-9 specification foam. This foam had good shock-absorbing and thermal propedies. The resin used was Union Carbide resin BRL-2760. The current foam is to conform to specification NCI - BU-7, Rev. 0 -

(see Attachment 2).

p Several years ago the Union Carbide BRL-2760 resin was discontinued. Nuclear Containers Incorporated (NCI) undenook work to find a suitable replacement as there were opponunities to use the foam design in a number of packages. NCI did a search, l

performed some foaming tests, and evaluated the density, strength and thermal protection.

As a result of this work, Schenectady Intemational HRJ-2590 was determined to be an acceptable substitute to Union Carbide BRL-2760.

l Shonly after putting the HRJ-2590 resin into service, some pitting of stainless steel was observed in packages using that foam. This investigation lead to the conclusion that hcl, used in the final neutralization step of the resin manufacture, was contributing to high chlorides and was responsible for the condition. The resin is a base-catalyzed condensation product of phenol and formaldehyde and therefore it must be neutralized with an acid to complete the resin formation process.

The work by Schenectady Intemational and NCI lead to a new resin designated i

HRJ-11825 (low chloride), it uses oxalic acid to neutralize the base-catalyzed reaction in the place of hcl and resulted in a resin with chloride concentrations < 200 PPM. This foam was again tested for chloride corrosion and the physical attributes of strength, density and thermal. The change in the specification to call out the new resin was

. approved by the US NPO.

The NCI BU-7, Rev. O specification calls out the maxmum chloride content as

< 300 PPM. -

ne original SP-9 formulation utilized a Union Carbide surfactant L-530 (silicone). At a later date it was changed to Union Carbide Y-6663 (silicone). The current NCI - BU-7, Rev. O crdis out Dow Coming DC-193 as the surfactant. He role of the surfactant is two

- fold. (1) It reduces the surface tension of the reacting mass and increases the strength of the cells formed during tne curing process and (2) They provide a much more uniform distribution of the blowing agent during the pour and curing process and to control the open and closed cell distribution. It is not clear that any evaluation of these changes have been made - other than the results of the physical tests en foam generated by the NCI -

BU-7, Rev. O specification which was found acceptable.

Page1of4

_ ~.

i i

(;

At another point in history, reagent grade boric anhydride became virtually impossible to obtain. A search and some physical testing indicated that technical grade boric anhydride by Eastman Chemical resulted in good pourability and the resulting foam met the density and thermal properties required. The specification was changed to call out " technical i.

grade" boric anhydride. The purpose of the boric anhydride is to provide some additional heat of reaction to aid in the foaming process. In addition the boron in it acts as a nuclear poison and adds to the conservatism of the package neutronically.

The Owens-Coming fiberglass rovings, which add to the structural strength of the foam, have also changed over time. The original was Owens-Coming # 805 (HIS). This changed to Owens-Coming # 833. The NCI-BU-7, Rev. O calls out Owens-Coming 405 l

or equivalent. Foam strength does not appear to be an issue in this current evaluation.

l l

Resin manufacturers recommend that when foaming metal objects such as the package here, that the metal surfaces be painted with a corrosion resistant coating so as to avoid chemical attack during the foaming and curing process. The NCI - BU-7, Rev. O calls out such a requirement, and calls out the thickness of the coating. Per the specification NCl-BU-7, Rev. O, it is noted that zine epoxy primer is used now, however, other coatings may be used if they prove to provide better protection. Here does not appear to be any demonstration of a coating that "provides better protection". The paint currently used is identified as PPG Primer # DP-40 Epoxy Primer and # DP-401 Catalyst. As of this time the information on the formulation of this paint and it's acceptability to this service has not been made available.

De BU-7 package conditions observed at NNFD, displayed characteristics not seen in any previous BU-7 packages used by GE nor are we aware of any other reports of such conditions. The corrosion of the zine plated steel bolts, both on th: heads and down the length of the threads is indicative of a significant chemical reaction taking place in all the foamed parts of the package. The deposition of the white powder also serves to confirm this. The wrinkling of the paint and associated corrosion could also be a result of the chemical reactions, however, it could also be indicative of some problem with the paint, the application of the paint or the surface preparation.

. Information to Date GE had our laboratory and a lab in the Chemistry Department at the University ofNorth Carolina at Wilmington (UNCW)look at the white crystalline substance. Based on this work the primary component appeared to be zinc. Anions Cl, NO3, PO4, and SO4 were all low. In a second test the UNCW results identified an indication of Forinic or Acetic' Acid or similar functionality associated with the white powder (see Attachment 4).

Page 2 of 4

-. ~.

l

(

Eco-Pak sent samples of the white residue to LAW Engineering Industrial Services for evaluation. He test report provides a great deal ofinformation. The XRD analysis indicates the white compound is n.sinly zinc formate hydrate, Zn(CHO )2 + 2 H O (see 2

2

, ).

i Zinc formate hydrate has a molecular weight of 191.44, appears as a white monoclinic, i

looses the two moles ofwater at 140 'C and is soluble in water from 5-35 g/100 cc j

depending on the temperature.

)

. Eco-Pak sent samples of the paint to LAW Engineering Industrial Services. Their analysis confirms that the paint is an epoxy with probably a polyamide catalyst.

~According to the MSDS approximately 10 wt.% barium sulfate and 5 wt.% strontium chromate should have been present but none was found. Zinc also was not found indicating that a zine epoxy most likely was not used. Strontium chromate is usually used in paints which are corrosion resistant according to LAW. He zinc epoxy primer would have also included zine (see Attachment 3).

He areas where the paint has been attacked and is bubbled-up, the underlying corrosion if the metal is superficial. No extensive degradation of the metal surfaces were observed.

In discussions with Eco-Pak throughout this investigation, they have indicated that the presence of water vapor and unvented storage could have some way played a role in the problem. In the event that root causes are determined these factors should be evaluated and the results factored into the corrective actions. The package has never required any special venting nor has there been any special environmental limitations (other than 10CFR71) placed on it.

Conclusions The following conclusions are derived from the physical condition of the packages and the analytical work done to date.

1. The corrosion and bubbled-up paint are fairly superficial. ney have, however, destroyed the zinc plating on the bolts and washers and the protective paint coating has been breached opening the way for further corrosion.

2. The most likely chemical causing the corrosion and paint failure is Formic Acid. He timing of formation, exact chemical reactions and release of the formic acid is not known or understood at this time nor is it understood how to eliminate or mitigate it..

3. Little information seems to be known about the formulation of this foam with respect l

to stoichometry, and the sensitivity of the process to varying quantities of the j

components of the formulation and the chemical kinetics that drive the reactions and produce post cure residuals.

4. Strength and density of the foam does not appear to be at issue.

i Page 3 of 4

5. The problems with the bubbling paint could be several fold including: inadequate

k. I paint or surface preparation, or in appropriate application. It could also be the result is formic acid formation and attack.

6. This seems to be a new characteristic of this type of foam, as previous experience does not indicate these types ofproblems.

7. Currently there is not enough information available to determine a root cause(s) for the problems observed nor to identify effective corrective actions.

i k

l

)

i l

Page 4 of 4 1

1 i