ML20062B271

Forwards Comments on BNL Pgx Graphite Test Results,Per NRC .Results Verify That Data Base Presently Used to Analyze Extent & Consequences of Oxidation of Core Support Blocks Acceptable

ML20062B271
Person / Time
Site:	Fort Saint Vrain
Issue date:	10/20/1978
From:	Justin Fuller PUBLIC SERVICE CO. OF COLORADO
To:	Gammill W Office of Nuclear Reactor Regulation
References
P-78175, NUDOCS 7810260205
Download: ML20062B271 (9)

Text

_

-_g q ' 2 d[h j.

PUBLIC SERVICE COMPANY OF COLORADO P.

O.

sox e40 0CNVER. COLORADO 80200 J K. FULLER October 20, 1978 Fort St. Vrain t

v.c.

...o r,a Unit No. 1 P-78175

{~;.

Mr. William P. Gammill Asst. Director for Standardization and Advanced Reactors Division of Project Management U.S. Nuclear Regulatory Commission Washington, D.C.

20555 Docket #50-267

Subject:

Brookhaven National Laboratory Quarterly Progress Report -

Issued August, 1978 Ref.: NRC Letter dated October 4,1978 T. Speis to J. K. Fuller Gentlemen:

In response to your request contained in the referenced correspondence, PSC has obtained the attached comments concerning the BNL PGX graphite test results from Mr. R. D. Burnette of GAC.

If there are any questions, please let us know.

Very truly yours, fw J. K. Fuller, Vice President Engineering and Planning JKF/FES:ers Attachment OY' O

7P/0 16 6 8f 0

1 k,

4 INT L CoXE$PONOENCE e

IN R(PLY i

l FRmd R.

. Burnette REFER To CCB:383:RDB:196 TO J. Kennedy oATE 10-12-78 suanct Relevance of BNL Data on PGX Oxidation

........ -.Summary The results of the BNL tests are in reasonable agreement with GA data ob-tained under similar co'nditions. As such, the BNL results verifies the data base presently used to analyze the' extent and consequences of oxidation of the core sup-port blocks in Fort St. Vrain. The BNL data, like any accelerated laboratory test, should not be used directly to predict behavior in the reactor. This can only be done by _ analytical extrapolation of all critical parameters to reactor conditions.

~

The BNL tcsts were performed at grossly accelerated conditions which ensured neither strictly uniform (global) oxidation nor surface oxidation. Rather, the conditions produced a mLxture of the two types of burnoff along with severe pitting and cratering. Since such conditions do not exist in the reactor direct comparisons cannot be made.

==

Introduction:==

Experiments have recently been performed at BNL for the purpose of determining the effect of burnoff on the compressive strength of PGX graphite (Ref.1). In their tests, cylindrical. samples (1.6 in. dia. x 3.2 in. long) were oxidized individually in a quartz tube furnace at 936 C (1717 F) in a flowing gas stream of He at 1 atm containing 16.4% H and 2% H O at a flow rate of 2.35 L/ min. The samples were 2

2 oriented parallel to the flow and supported with quartz pegs. The ends of the cylinders were protected with hemispheric caps of H-451 graphite.

The samples were exposed for times of from 1 to 64 hours7.407407e-4 days <br />0.0178 hours <br />1.058201e-4 weeks <br />2.4352e-5 months <br />. Burnoffs ranged from 0.25 to 22.5% and reaction rates (based on weight loss + time) ranged from 2.9 x 10-to 16.1 x 10- g/g sec. At the end of the exposures the samples were weighed, photographed and then measured for ultimate compressive strength.

The visual and photographic examination revealed severe pitting and cratering of the samples surfaces. The pitting was non uniform, sometimes occurring on the top, bottom or on one side.

~

^

~ ~ ~ ' ~

~~

~~

The measured compressive strengths of the oxidized samples when compared to the nonoxidized controls showed, according to the BNL authors, a loss of UCS which was greater than expected based on exclusively surface oxidation. The authors reasoned-that the samples must have been oxidized in bulk rather than on the surface. Surface

• g5 4

1

L i

oxidation was predicted to occur, based on an assumed effective diffusion coefficient of 2 x 10 which is the value calculated from an equation incorporated in the

~

~

OXIDE-3 computer code (Ref. 2).

Using D,gg 2 x 10 the BNL authors calculated a diffusion length (or L value) of 0.0192 cm for the " worst case." They claimed that the observed results (i.e., apparent global oxidation) contradicted the pre-dicted surface oxidation.

Analysis by General Atomic It is General Atomic's considered opinion that the BNL experiments were pre-liminary and scoping in, nature. They yielded results which were entirely predictable based on the accelerated conditions of the tests, and the known heterogeneity of PGX

[

graphite. As such, the results should not have been used to make generalizations on what could occur in the reactor. The reaction rate and strength loss data are in reasonable agreement with results obtained at GA under similar conditions.and are closely predicted using the analytical techniques developed at GA.

It is concluded, that, the BNL tests help to verify GA's data base and analytical methods.and in no way change the conclusions of the analyses, that oxidation in the reactor is largely surface oriented. Direct comparisons using accelerated tests must be avoided.

The following specific points will clarify the stated conclusions.

1.

Surface Burnoff vs Bulk Burnoff Strictly surface burnoff only occurs at conditions where the reaction rate at the surface is substantially faster than the rate of transport of oxidant to the surface. This condition occurs at high temperature, at high total pressure and at t

low water vapor concentration. The BNL experimental conditions included only one of f

these critical parameters, namely high temperature. The low He pressure and high

- moisture content caused the oxidation to occur partially in the graphite diffusion regime giving rise to a burnoff profile. A burnoff profile was measured, apparently on a sister sample, exposed to the same experimental conditions, using the X-ray attenuation technique. The profile showed that unreacted graphite was found at a depth of 0.15 in. or 0.38 cm (Ref. 3).

Indeed, this depth indicates reaction con-trol in the in-pore diffusion regime. As will be explained, this depth of attack is predicted to occur at the test conditions but not at reactor conditions.

2.

Diffusion Coefficient

-3 The effective diffusion coef ficient of approx. 2 x 10 stated in the BNL report is correct for 39.ata He..The test conditions were approx. 1 atm.He and the corresponding D,gg is 0.078.

The calculated depth'of attack, or L value, should have been 0.12 cm which is in reasonable agreement with the L. values calculated for l

al'1 the ' tests which ranged from 0.067 to 0.38 cm given in Table.1 and. tihe burnoff 2

.~ -

profile which was measured at BNL. The L values given in Table 1 are, indeed, not indicative of strictly surface burnoff. They do indicate oxidation in the graphite diffusion regime which produces a burnoff profile. It is of interest, that the calculated L values for the BNL tests are quite small indicating steep profiles.

In the reactor, assuming the moisture content was the same, the L values would be

.. reduced by 1//P r about 1/7 giving rise to virtually surface oxidation.

He t

~3.

Reaction Rates

~~

~~

The intrinsic reaction rates, R,, of the BNL samples were calculated using the equation that is similar t'o that used in the GA analyses:

R = S/V LR, T

where R is the observed average reaction rate (g/g sec) reported by BNL, S/V is T

~

surface area to volume ratio, em, L is the calculated diffusion length, cm.

= 0.078 cm/sec. The calculated values The values of R, were calculated assuming D ff of R, are shown in Table 1 along with the comparison to GA reaction rate for H-451 graphite. It is seen that the observed BNL reaction rates are in reasonable agree-ment with the published rates for PGX graphite which typically fall within'100 to 1000 x H-451.

4.

Flow Rate

'The stated flow rate of 2.35 L/ min was insufficient to ensure constant mois-ture concentration over the entire sample surface. The calculated axial moisture depletions ranged from 14 to 77%. Accordin;1y, such low flow velocity virtually assured non-uniform pitting and cratering, depending on the location of the catalyst sites relative to the inlet gas.

5.

Moisture Cor. rent Since the chemical reactivity of the catalyst sites is several orders of magnitude higher than non-catalyzed graphite, at high moisture concentrations pitting is assured i.e. the catalyzed sites were continuously flooded with oxidant.

To simulate reactor conditions, tests aspiring to study surface oxidation should

~

be performed at as low a water concentration as practicable.,

For example the rate of pit growth can be calculated (using equation 7 of the Ref. 4 attached) for different conditions. Equation 7 (Ref. 4) assumes that pit growth is limited by diffusion of moisture into the pit where the chemical reaction at the base is infinitely fast.

l 3

l

~'

T a.

For the water concentration used in the BNL experiments, 0.02 atm, a pit 1 cm deep could occur in just 12.6 hrs, solely by diffusion.

Pits can grow even faster if moisture is transpired to the catalyst site via He flow.

b.

At full power conditions in the reactor at water concentration of 1 ppmv a 1 cm pit could occur in 30 years by diffusion alone. Clearly, in lab tests of graphite oxidation where the reactions are dominated by catalytic impurities, the use of inordinately high water concentra-tion will produce pitting attack which is nonrepresentative of-the reactor situation.-

.6.

Effect of Oxidation on Strength The data given in Fig. 2.17 of the BNL report (Ref. 1) are inadequate for proper interpretation. A larger number of unoxidized control samples should have been taken (a minimum of 10 is recommended). The variability of strength of random graphite samples is illustrated by the fact that the sample with 0.24% burnoff was stronger than the controls. The third strongest sample had 1.72% burnoff.

The BNL data are compared with GAC data in Fig. 1.

The CAC data were re-ported in Ref. 5.

The BNL tests studied the effect of oxidation on compressive strength while the CAC tests were on tensile strength. In both experiments the samples were oxidized at conditions in the diffusion regime which produced a burnoff profile in depth. Under such conditions strength losses similar to that of~ global oxidation are expected and were observed. The BNL tests showed slightly higher rate of strength reduction than the GAC tests. Within experimental uncertainty, however, the two tests are in agreement.

An analysis of strength reduction was performed on three of the BNL samples.

using the met!od of transformed sections which is used in evaluating the behavior of the Fort St. Vrain core support posts and blocks. In this method an oxidation profile for the sample (or support post) is calculated.

(In this case using the observed reaction rates and an assumed D,ff of 0.078.) Then, the sample or post is

(

divided into annular sections, each section having decreasing burnoff with decreasing radius. Each section is then assigned a strength and elastic modulus, assuming the oxidation in the section is global or uniform. The overall strength of the sample is then calculated by integrating the differential strengths of the annular segments.

A plot of the calculated strength reduction compared to the actual strength reduc-tions vs burnoff is shown in Fig. 2.

The calculated strengths of the samples are, within experimental uncertainty, equivalent to that of the measured strengths which in effect, helps verify the analytical techniques used at GAC.

4

~; ;

e.,

e e Summary and

Conclusions:

The BNL reaction rate data are in reasonably close agreement with GA data and with calculations using CA's analytical techniques. The strength loss results

.are in agreement with GA data and with predictions using the method of transformed sections. Accordingly, it is GA's position that the BNL tests offered further verification of the existing data base and analytical methods at GA.

The severe pitting of the BNL samples is explained by the use of high moisture content and low total pressure. Nqn, uniform surface attack would' occur at low flow rates, bec'ause of depletion of oxidant.

Si. ace such conditions do not exist in the reactor, direct ' comparisons will result in grossly pessimistic pre-dictions and must therefore be avoided.

References 1.

Reactor Safety Research Programs Quarterly Progress Report for Period April 1 through June 30, 1978. BNL-NUREG 50883, August 1978.

2.

Peroomian, M.B. Barsell, A.W. and Saeger, J.C., "0XIDE-3: a computer code for analysis of HTGR Steam or Air Ingress Accidents", GA-Al2493, Jan. 1974.

3.

Private communication, Fred Growcock, BNL, Oct. 5, 1978.

-4.

Letter to R. D.,Schamberger from A. P. Malinauskas dated June 13, 1978 (attached).

5.

HTGR Fuels and Core Development Program Quarterly Progress Reports for The Feriod Ending Feb. 28, 1978.

cc:

G. Engle A. Barsell C. Velasquez A. Beavan R. Price C. Hoot R. Vollman W. Craul R. Wunderlich G. Wessman T. Gulden 5

i._ _,

I i

Table 1.

Brookhaven PGX Ovidation Data e Average **

1. C*

Burnoff T

a Calc

• at a

Strength Frac Time g/g sec g/g see L

Surface Rate Sample psi Hrs x 107 x 107 X H-451 cm 1

4684 0.24 1.7 6.7 40 0.17 1.4 137 2

2922 0.5 1

13.8 170 0.08 6.1 583 3

3393 0.53 5

2.9 7.7 0.38 1.4 26 4

3553 1.72 3

15.9 230 0.069 24.8 788 l

3 16.1 240 0.067 25.9 823 5

4473 1.74 6

2634 3.0 15.25 5.5 28 0.196 15.3 96 7

3095 3.08 22 3.9 14 0.279 11.1 48 8

2634 5.12 22.5 6.3 36 0.175 29.1 123 j

9 3393 6.22 24 7.2 47 0.153 40.6 160 10 2190 6.57 15.75 11.6 120 0.097 68.0 412 11 2140 22.33 64 9.7 86 0.112 198 295 4079 0

Controls 4371 0

k.4578 0

Calc R, & L assumes m = 0.01, D12 HO

= 7.8 and P

.02 atm

=

2 Burnoff at surface = R, x sec e

t t

5 6

e f

4

. M fi i

' s 1

i

..h'._..

4_.

, i i

-h-

_~. _._4_._1.

~

i i i

_J

II !

!1

! I _.

4-

._._w'.,_ U i*i

'*I

_ t Ae 8 t

!!I t..

d.

1, T-1 a

f_

""T' T T r

9 v3 i

i i.

i a

i

. _L,

g._.. W.__

f

-.-1.-

!Ii !

. m l_;_.J I

i II

', I l' l

i '

' I -

p' ' {.' _g. C y M.

I!. 4 L._

t

(

_.i._

._L 1.

I ?

j 1

3 p.

s-a mnm i

--t,-

i 6 c4 m.-

o-

.._.O-.

.N f.

i li; -Q ; f

'iI

! _ (3.. p C. _ __d _... Q.

i i I f

f

..g__ue ww I

. ~ ' I T.

3 i i I I

f i c

i 4t'-t l

t 3

M M

M M i

O i

c-umv i

w r3 ~J u

i i

8..

- C# Z ----

1 3 '

g 4

e t t

i

_. _m = ca ___.

i

_1

.o

?!

l'LL_.. !. M-

,'ii i

i

+

1 '

t i lll lil' l

5 i '.

g

-i

+m

.T i

,,i i

...i

!2 g

' ~~~ N, 1t 1.

i l i i

i 7' ' ~li',

- ! i,'

' i ',.

.'i' o

t

. i i i 6..

Ii, i i I i

i.

i 6

. t i. -*-, i i i i i

iii

} I i w

TT -*

• i i

i 6.

E r,'.

j,!

i i i l

!l1 lijl il.

llF l i 1TM; r-t-i T

-i i !i N.

_u i-i !

i ;.

Ti,j

~

ii,,

i r' TTT,-

iii-i, !

p-r--.

i - ! i

' + i i'ii

-c i +.

' i o

~M g

N i+

-4TT sh _. -_ L __. 'a p i.i i

i. i

-t t

.i, R

m k

-- r- +-

14 i i L44 _.,j,,

',; }

a __

rs O

l' o

,6

-f 2 i-t*-

3 i i i r..

5Fg-p g

. i,

i>>-,

i

< i e

_ _a i

i - "

pa ',

L

!a w.

or

.'L. '. '

t

'i 4

1 ii i t u.

.c 22. _;_

. e ____.

M w< _. _. _._ o _.

i ! t

! i_

i I i'>

_i o.

l'!

ill' ll yt_

p l

g 4t

_uL A.

2_

U

._ _ h

- \\

. i i

a_!

__~__E___.

__.:_g E

x.__.2_4 -

m..-__.__

T i !

!i

_; J -

i!! I

'+

e-

. e a

I t

i

+

1 I t y

~

l l l

-f !'

ll'.

l l

'. l

-e---

i e

4

__o_

u

_.__ i

i...

,i,

i i i u-i1

.i 1

g

_m,

. i p

' I e i

i

_ J,_

i i

.ie i.

i e e

t\\

r i e i l e i

e c

7 i\\

O w

- +

i !

6

.n

.t_ t i

u _._ 4 '; 4 i +

o.>

t '

.I t

e t

i

'O

~.

- ]

k' ' I

, a i f iI D

i i

I

\\

. e,

e

' ii.

M t.--__-

i i

i !

_t I,

t 1

a -

,e y

i i

i. !

u a

i i.

\\

y oc i

c

\\.

1

+e g

\\'

i '.

!ii e

\\

! i mt p

i. !

i i

-y

.u

_a ll'l

\\,--

1 U

4-(

i yg

-+-

O Q_'_L i

i 1

+

aa _a' _

Q

- t i y

c

1

,. t i

1 i

I

's 2' -

i

&K p_i L f

g

+

s

\\t__ __.

p 1

1 - '

_a n

n ug

~ ' ~

~

- 'i _

T--- (

y g

\\__

a___. _

O

-~~~^

g V

I '

.l f

• I,

I f

Tiki ' N!

~

M

_q

__w I

a ___

.I ec __

u

_a_.___.

q W. l

.. _,la..

li. -l ' ' l l

l'

' ' ?

L 4l_

J j

f

_; _L J_v ;

s.

i j __ _ _ _ _K_ _._,

L-4

_._.y.__..__._

......j.

.I...._

r ---.__..

-'1,.

__-__.-d i._

d_

l

_..._._4.__-______.4

_.________C Qa_

_.__.O_.

l

.~, a.:__ _,______.o_____ _. _ _..O

_C' ' _, L e. l

{%). p3uara s ' uT-._ _. _. +

{

1 a_. _ n _._

L_._.

e ci

~

t I

1-

___L.

_._.a, l

uopanpag t '

a. a..

J.

t !

.q..,l _L 7_ I

_1.

J _;J -

J_31-._

I 1

_1,

.1 i-_4_a _i L

ve yt 4;.

u_

i,.

m e'

t,:

,3,MJa-++_ 2.: :

1! Ad!

+

6,

,. c D

h I

- ed

.,.6-

,,.4 M

_a_ :

• i

.. i n

p!

1 t

f.

I s.t

. t

- i i i r.

4 i

. i

_L..I i

l.'

f i

4.

l iji

. i.i

~

k.Y M TT T.

i i T-*-tii a

i

, i i i j

i

i. ?

!. 1 i i i i !

i j

b i

i i iii.

1 i i

• iai i i i 6 i i

Ii i

2~

1..

i i

+

i i t

Ii i

aj i 1 TT III i !!

l i _ !

I 6

3 1

'l

! i I

'iI!

i i i

!3 i

d' I' I Ti i e j

T-j1

• e i i

i i i i I I e h

ie

.i

-t%;

I ii*

i i

i i i i. !.

!. t i

1 i

a i ia ii 1

C~2

. i

' i i

I O

N vs

" d.3.,. ; i j,,,!

-'M

-i10-d j ;

j

, i i

i ;

i,

i..

is i z i!.

4 i

i

' i i i

i I i,I i

i i a i ! i 1

i ei i i 9 u 4.s i

u.

gy

, ;3

• C C e i i.

? !I i

ia i

i i i t

!,i i

T'- '*N i i i...

i i

is i

! 6 i!

7 i

i ii i

i i

I i i I

e i T~g I i.

I t

8 1

• re o i -

i !Ii 6

i ! ?

i. i_ i i e i

l i e 1

i ii.

mg i i ii i

i i i

Ei i

e i

i 4,

m

. i-b i

1

-i-I'*- u j j jii i;ji 4 I i

---e4 0 -- ~

~

j j i

iii

~

i!I i

i D i. ~ h ~ -*

"' i e i

~' I-"~

I ' j"-'

I i

, i i !

! t e i'!

e ie j i i

T"'T

' * - * < 'q

, l i.

e i

i i I ) ! !

I i i f i a

i.i T*

i i 9 q

i i r i i i

e i i i e

4 i.

1,

i 1 1

_L,-. 2. ~h e

i i i i, y

^

t i

i go

- -o - t-e -

l

! !l t' l-

! i.

i

, i 1

'~ h_ ;'!i!.!

l'

! l I il ll i

. i i

-h+.. M ~- N!f h-1

,i, 9

l

-i n

m i j;i l

i i i

-t-*-

o 4-414. $ -H-!j - i@ -i---it,,;; ;;jl ll ill

ll llll

_ l.'

CO

, j y ;

.,i 6 9 l j T

4 i

^

l i( 1 iii i. ~

.ie i i i e iii i

i i

.e i

i e i.

I

' a i

e i i i i

i.

.<e

, ie.

i i

i e i.

I i,

e i i i lY-1 1'e i

i i

h ii!

4 i i e

i'

.i I

'i i i ' i

! _4,1 i i!8

. M.,.

i _

_L. J S

__ A 6

8

. t_ n._.

m___

i 1

i l t i t

i i

+ -

i t e i e ie f sc 6 e i

-t~-*-+-

T i,..

i e

i i.

i i

. 6,

i i i

j i j i.

j h.

i i eiii i, i I 'k _h,-,

i,

!i.

i

. T_"_"~1 l

- j i

t

,i T

1,.

6 i a.

_.--+

.i

~

6 ta i.,.

i i i 1i.

i i

I i

e i

1. !

t a

! ! l I i

. e a e i. !

e I

. i e

i

} I e

. e i C

4 i

a I i e

(

i j i i i i.

a E

8 i

i

! ! I '

8 l ? f I. !

i e t

?.

8

) ! _1

- 6 i i_,

I

!i

! 1 I I I I i i i i i l t C'

I at i

a i

!,I l 6 i3 i t!

l !' i

' i C

- i -

i

• g I i I

! ! i 8

i i '

g i i 1 e e t i

i i i

i e

i,

i i s e

i e i t i i

k r

t !.

s i,

t O g' k

i i i e4 7

i i a t

_!i!'

t ?

i

!I t

i. !

i i '

g

+

e i e i p ;_

a

! l. 4 I i i

I i

e I I

! ! i 6 1

i !

e i

i r

e i

e I i l t t

6 i

i. i a

e i 6

. i i.

i 6

i i

8 i,

i t

i i f

f a

+

h j

l I*

t ! i i! !

i 6 i

. !li

! I i

. i t_

lii!

i i e i i

i i

i g

I

+

i !

i i

I s

i l.

g i

i i

g e ! i 8

i i, i y

e C

1 I

i i

o*

C

~.

-=

' ' 1 I

i t

• 6 i ! l l f I

i

- e s

• i

o aiI e

t

' e l!

4

. ' ' I I -

' +

ii

• i 2

j_L_'

O,4 _

i 1 e i 1 i t i i

• i

' +

i i

i ! '

g il i i i

yo

_. _O e

(

.l-Qy i

e i i i

i,4,,,

e

_,q

,Ti,i 4 !

i

. 6 6

6 Is e.

e e*

• i

,. 4

^1

'tc i i

, i

~

Og c) i i

O

%e V_

w p.

i e i t I

i i i I e

i ;

i i i,+

i,

i 2 i i i '

i.

m 8

I l

t 0 4 0

i 0 '

8 k '

• u i LL e

i i 4

i 1

f

, i.

m.

w

.,.,m.

i > e i

e I i Y..

i

_y I

i

_. i 6

i et.as

_.i_,(_

a

.q, i

i e e

a '

I j l 6 i e

edium m

.e i I 4 i i

!+

i

,;g

• sh M

+44

i. 1

.,i,.

6 N

g I

i I

, i..

. I 9 g g l

g

' g 6

1 i

(

4

-_..i l

,e L Li, I If A-_

.g t i e i i j

N.

i>

+. - - _

.m.__

s g

l e emw-ee t

t

+.e ie_.

---r

,ee I

-._.--__g f

J_ A L_L _.

g

, O o ~-

O m W O

G U%

O O

o

-r-p a.e -.-

a9

-l9 o,_..

ag um *

• ql..

'i i

, t

- ' l

+;

a a

i 6

1

$ j 5

p 8-I.

I I

. 4_.

4

!i!

• i.!

i 8 4

-I

^.

1 i i t t !

17ii 6 i

. i.

! 1

.i*

.t

+

i i

4 4

i

.g e

gm