ML20023B073

Forwards Independent Evaluation of Ga Co Rept GA-A16764, Containing Analysis of First Plateout Probe.Reasonable Agreement Exists W/Plateout Probe Measurements

ML20023B073
Person / Time
Site:	Fort Saint Vrain
Issue date:	12/06/1982
From:	Wagner P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:	Lee O PUBLIC SERVICE CO. OF COLORADO
References
NUDOCS 8212210520
Download: ML20023B073 (2)

Text

.

L-DQS /%-Di(#

DISTRIBUTION E6 1982 c aocket File Docket No. 50-267 Local PDR ORB #3 Rdg D.Eisenhut J

temes Mr. O. R. Lee, Vice President Electric Production utzer (3)

Public Service Company of Colorado P. O. Box 840 NSIC Denver, Colorado 80201 E.L. Jordan

Dear Mr. Lee:

P GA-A16 m,

We have completed our review of General Atomic CompanyG 302(P-82419).

June 1982 which you submitted by letter dated September This report eontains the analysis of the first plateout probe from the Fort St. Vrain HTGR, which was removed in accordance with Technical Speci-

)

fication SR 5.2.6., and concludes that the radiochemical concentrations are far below the allowable Limits of Technical Specification LC0 4.2.8.

This conclusion was reached even though considerable differences existed between predicted and measured data.

Our consultants, Los Alamos National Laboratory, have performed an indepen-dent evaluation of the submitted information and have arrived at a closer agreement between the predicted and measured values. Although sources of error uncertainty still exist, we have concluded that reasonable agreement exists with the plateout probe measurements.

Enclosed is a copy of our report for your consideration and, if you deem appropriate, comment.

If you have any questions on this matter, please contact me.

Sincerely.

Original signed by:

Philip C. Wagner, Project Manager Reactor Projects Branch #3 Region IV Enclosure; Plateout Probe Report cc: See next page 8212210520 821206 PDR ADOCK 05000267 P

PDR omerp...Q8.B.Q;.D.L....0RRA3.:D1... car

. 9.RBf 3.;p.L.

suRace >

...P,K,r,9,u,t,z,e,r...PWagneedd... M.I one>

.12/...../.82.....1.Uh/.!!2.... ?...W/4../s2..

5...........

NRC FORM M 00-80) NRCM A0 OFFIClAL RECORD COPY uso es -mem

4 i

e Ft. St. Vrain cc list l

James B. Graham, Manager Licensing and Regulation East Coast Office General Atomic Company _

2621 K. Street, N. W.

Suite 709 Washington, D. C.

20006 Mr. W. Dickerson NRC Resident Inspector 16805 Weld County Road 191/2 Platteville, Colorado 80651

' Director, Division of Planning Department of Local Affairs 615 Columbine Building 1845 Sherman. Street.

Denver, Colorado 80203 Chairman, Board of County Commissioners of Weld County, Colorado Greeley', Colorado 80631 Regio.nal Representative, Radiation Programs Environmental Protection Agency 1860 Lincoln Street Denver, Colorado 80203 Mr. Don Warsmbourg Nuclear Production Manager Public Service Company of Colorado 16805 Weld County Road 191/2 Platteville, Colorado 80651 Regional Administrator Nuclear Regulatory Commission, Region IV Office of Executive Director for Operations 611 Ryan Plaza Drive, Suite 1000 l

Arlington, Texas 76011 t'

S l

O e

4

• 'g# ' j,

9' - (* ** ); -

h" e - +

w g=+-

9 9

e q

• 9

,j'-

.*t

,g ogt e ; Op e,.9""e ' 8 7

• • y

r

- - - - - - - - - = - - - - - - - - - - - - - - - - - - - - ~ - -

s m

A

~

Analysis of the Iodine Fission Product Plateout Probe Measurements from the Fort St. Vrain High Temperature Gas-Cooled Reactor 9

by Clarence E. Lee Department of Nuclear Engineering Texas A and M University College Station, Texas, 77843 v.-

September 20, 1982 Th.is work was performed for the los Alamos National Laboratory under Contract 9-L32-02815-1

,/

g[

O

(

J is

(

C.

o s 1 1 ' G ^ ~

'Q f f t

(."

IV U

=

v r,

. c.c

..s.

..3,.

9

)

~

Analysis of the Iodine Fission Product Plateout Probe Measurements from the Fort St. Vrain High Temperature Gas-Cooled Reactor by~

Clarence E. Lee

. ABSTRACT.

Iodine fission product plateout measurements were performed on the-Fort' St. Vrain High Temperature Gas-Cooled Reactor in late 1981. Evaluation of the Iodine decay study and the performance analysis of the Iodine monitor indicates that reasonable agreement between predicted and measured performance can be obtained. The discrepancy factor of 200, reported by GA in Ref. 1, between predicted and measured Iodine probe plateout activities, can be resolved. The resulting plateout rates have an uncertainity of about a factor three to five, if some simple assumptions.are made on the delay times before the activity measurements. occurred and on probable uncertainities involved in the measurements. Using a self-consistent analysis of a zero dimensional plateout model and measured activities, the correlation obtained in terms of the decay constant, A, is

= e.48027 in h + 0.21705 0

r r p

where r$. is the purificatiog/s, r rate, and r is the plateout rate. At 3

ThiI res.6E10 70% pow r, if r to 1.6% plateoul per.73x10' 1

=1

/s, which corresponds

=

ult is consistent with the pass.

Iodine parent-daughter experiment, the interpreted Iodi data, and the plateout probe measurements with,ne monitor estimated uncertainities.

@ e G

e g

(F "

9 F'

.*P g

,y A

.y.m

- Q*,

,ye

, g +3,,

s I. INTRODUCTION The re'sults of the Fort St. Vrain (FSV)' Iodine plateout probe experiment were reported in the General Atomic (GA) report on the

" Radiochemical Analysis of the First Plateout Probe from the Fort St. Vrain 131 High-Temperature Gas-Cooled Reactor,"

Ref. 1. The measured I

concentration reported is well below the allow'able limits specifled in the Technical Specification LCO 4.2.8 as reported in Ref.1.

A discrepency factor of 200 is reported (in Ref.1) between the predicted and measured Iodine diffusion tube activities.'This discrepency factor does not affect the satisfaction of the Tech'nical Specifications, as

~

cited above. However, if this discrepency were unresolved, the use of simple models in the independent and consistent verification of the plateout activities end plateout rates would be difficult to justify. The goal of this investigation is the determination of. consistent Iodine plateout rates and the subsequent prediction of activities. This analysis is directed towards understanding the 131.1 plateout measurements using the zero dimensional modeling of Ref. 1.

Several factors can contribute to the difference:between the predicted and measured probe activities.

First, if. one treats the the gama scan measurement of the probe in FSV 51 and San Diego as accurate, especially with respect to Cr,' for overall --

~

consistency of all the data, chere is an apparent 35-40 day pseudo-time

' delay before detailed measurements occurred. Probe activity measurements at 10-20 days after removal from the reactor would not appear to be consistent with other data, and the reported plateout activities, if referenced to a precise plateout rate. However, if we assums that the measured activities were undercounted by a factor of two, then consistent plateout rates and activities can be. determined if the pseudo-time delay is about 30 days.

Second, in the determination of the predicted diffusion tube plateout probe act.ivities, although the total plateout and circulating activities due to several different power levels over a three month period wer'e' corr [ctly decayed to sh'utdown time (Ndvember 9,.1981), we find a different result than -

~

131

~

~

given in' Ref. I when the I deposited in the diffusion tube i's correspondingly decayed to shutdown, as it should b.e,',and then to the v

-2

~

v..

~

estimated time of measurement.

Third, a correction factor of three was applied to the Iodine monitor 135 133 'in Re'f."1. If this experimental circulatin'g activity data for I and g

135 correction factor is applied only to the I data, we find reasonable agreement with both the estimated pseudo-time delay befor's probe data measurement and with three (T, 8, and R) of the five tube activity measurements using a plateout rate, r, plus purification rate, r, of about p

g 1.63x10-3/s compared to the, corresponding GA value of 1.6x10-4/s for 131,

g 131 For 10.3s cycle time (70% power) we estimate for I that 1.6% plateout per pass occurs, as compared to 0.25%.per pass for the GA value (assuming r

=

2.7x10-4/s and r = 3.3x10-5/s, as in TABLE A-5 of Ref.1).

(

s Fourth, although the plateout rate expression given in Eq.- (2-9) of Ref.1 is correct for steady state and the inequalities presented therein, a more accurate limiting steady state expression is given by r +r '= (A /A -1)h, s p p g where r = purification rate (/s),

, 3 r = plateout rate (/s),

p A = plateout activity (C1);

p A = circulating activity (C1),

c and A = isotope decay constant (/s).

This formulation permits discrimination against the determination of self-consistent r,+r values that are' smaller than known values of' r in the p

s analysis of 131 1 131 The self-consistent 1 plateout rate was solved for iteratively using the plateout and circulating activity, the predicted T tube activity 133 135 contribution decayed to an assumed measurement time, and the I and g

measurements in a. least squares fitting process on in(r,+r ) versus in(h),

p the representation suggested by GA. The modeling equations of the Ref. I wore used with the modifications indicated above. The final result is 133 135 consistent with the probe measurements, the I and I measu'rement[ and 51 the apparent delay time to ineasurement, as indicated by the measure Cr-133 135 gamma activity. The results are also consistent with just the 7,n'd g

measurements if the experimental correction of a fac. tor three is applied 9

133 only to the I measurements. Finally, if an undercounting activity uncertainity in the probe measurements of,a factor two is assumed, then we find a consistent solution, as above, with almost the same, plateout rates, but with a pseudo-time delay before measurement of about 30 days. The overall uncertain'fi:y in the plateout rate using various different

~

assumptions is estimated at about a factor of three to five over the range of h for 1 to 131 Other possibilities that were examined seemed 132 1

inconsistent with the probe measurements, predicted unrealistic plateout rates, or corresponded to measurement delay times that were probably unrealistic.

In Sec. II we examine the Iodine decay study, in Sec. III the Iodine monitor analysis is evaluated, and in Sec. IV the plateout probe analysis is addressed. In this analysis we assume some familiarity with Ref. I and the general problems being addressed.

II. 10 DINE DEC Y S1VDY:' PARENT-DAU5HTER EXPERIMENT The derivation of Eq. 2-7 of Ref.1 is very easily obtained using the method of Laplace transforms. If N is the amount of the parent Iodine g

isotope, and N is the amount of daughte'r Xenon isotope, then the decay 2

equations applicable to the parent-daughter experiment are

~

dN /dt = -h N,

g g g (1)

~dN /dt = h'g,Ng+h2 N

2 2

(2) where hg = Ag + r i + r g, s

p 2 " 2 + r 2 + r 2' h

2 s

p h'i = BR Ag + r,1 + r g

1 p,

(3) 4_

~

with the definitionsj

= the decay constant (/s) of the ith j

isotope, rg=

th th purification rate (/s) of the i isotope, r,j = plateout rate of the i st th isotope, ' nd BRg = branching ratio of the 1-1 to the i isotope. In this a

simple model it is assumed that there is no source term to isotope 1. namely no further production of Iodine occurs during the time peri'od of interest

'after shutdown.

th The activity of the i isotope, A, is defined by Aj = Aj N, so that j

j the activity rate equations follow from Eq. (1) and (2) as dA /dt = -h A,

g g

g

- (4) dA /dt = g At-h2 2

g A'

2

~

(5) where g ',= (A M )h' t.

g Z 1

' Defining the Laplace transform of Aj as L[A (t)'] = A*j(s), Eqs. (4) and j

(5)become 0

s,A*g - A

= -hg A*g, (6) 0 s A*g - A g A*g - h2 A*2'-

=

(7) 0 th where A g represents the initial activity of the i isotope. Solving Eq.

(6) and (7) for A*j easily yields l

0 A*g = A /(s+h),

t (8) 0 A0 )/[(s+h )(s+h )3*

A*2 = (A 2+91 t

2 (9)

Since Eqs. (8).and (9) have only simple poles at s= -h and s= -h '

g 2

application of the Cauchy Residue theorem gives the solution directly as

-h t 0 A (t) = e t A g

-(10)

-h 8 02 + E9 /(h -h )] [e-h t,,-h t] A (11) 0 A (t) = e 2 A t

2 2

1 2 t

~.e...-..v..,:,,.

n n.....

0 Sq1ving Eq. (11) for A g we find the estimate of the total parent plateout given by 0

t 0 3/[* h t,,-h t23, A

= [(h -h )/g ][A (t) - e 2 3 i

2 y g

2 2

0 (12) where A (t) and A here represent the measured and extrapolated activities 2

2 of the daughter isotope, i=2.

For the experimental parent-daughter experiment, i=1 corresponds to Iodine,, i=2 corresponds to Xenon. The reactor was shutdown for about 19 4

hours (6.84x10 s) before the first measurement was made. As there was no Iodine production (and therefore no plateout) during this period and we are estimating the total Iodine plateout inventory, r = 0. Xenon, b'ing a noble e

p gas, does not plateout. Similar11y, rsi = 0, and rs2 = 1.73x10~5 (/s), as defined in Eq. 2-7 (p. 2.-21) of Ref. 1. Hence, the coefficient.

[(h -h )/g ], in Eq. (12), becomes 2 t g

[h -h )/9 3'" OlXe +

sXe ~ A ] / [BRg k,].

2 l 1

I c..

(13)

Equations (12) and (13) agree with Eq. (2-7) of Ref.1. The parameter values are indicated in TABLE I, where the decay constants, branching ratios (BR), and initial conditions are listed. The time dependent values at 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />' were estimated from Fig. 4-8 of Ref.1.

}

The predicted values for the inventory released are obtained using Eq.

(2-6) of Ref. 1, g = K P f Y R/B (1 - e-At ),

A t

(14) where A g

= act,1vity of released Iodine (C1),

P

= thermal power = 8.42x108 g, f.

= fractional power level, Y

,= cumulative isotopic fission yield, R/B = isotopic release rate-from Xenon R/B t1/2

~

curve given,,

in Ref. 1 Sec. 4, "A

= decay constant (/s)..

.,e E's

'a'

-'a

s**

-4

a t

= time at power (s) g and K

= 0.85 = 3.15x1010[ fissions /W-sy3.7x1010 [dps/Ci].

Using the parameters of TABLE I and Eq. (14) we calculate the values of 133 135 the predicted plateout of I and I, given in TABLE II (labeled LANL),

under the assumption ~that essentially all those Iodine isoto' pes are plated out. We obtain agreement between the observed and predicted 133 135 I and g

values in the range of 13 20%. The corresponding results reported by GA are intherangeof9-37%.TheIodinedecayexperimentcanbejudgedsuccessful.

. TABLE I.-

Parameters Used in the Parent-Daughter Experiment A = 133 A

= 9.3x10-6 /s, BR = 0.976, Y

= 0. M 133; 133 g J. 3

= 1.5x10-6 /s, r

= 1.73x10-5 /s s

". Xe Xe 0

axe (t ) = 8 C1, A Xe = 12.8 Ci o

A = 135 h

= 2.9x10-5 /s, BR = 0.70, Y

= 0. M 135 135 7

g

=2.1x10}5 /s, r,Xe = 1.73x10-5 A135

/s g

0 A,(t,) = 10 C1, A Xe = 48 C1 y

4 6

t, = 6.84x10 s> tg = 2.8x10 s ge 4

9 i.

. e

,e e O G.

9 eJ [

e+

Te e 8 -- *. =

E

[e 1,___._.

3 4

TABLE II.

Comparison'ofCalculatedandObserve[iPlateoutof 133 135 I and g

at 70% Power at the end of cycle 2 (May 13f 198'1, FSV).

(Including information from TABLE 4-4, Ref. I a).

Laboratory GA LANL GA LANL Isotope 135 135 133 133 g

g g

g Obscrved Plateout 94 113 187 169 (C1)b' f

Calculated Plateout 129 128 204 203[h]

(C1) c 1609 l

\\

i l

[Cale./0bs.-?.3(%)

+37%

+13%

+9%

+20%

d 6

Iodine R/B x10 2.9 5.3 5.5 3.9 6

Xenon R/B' x10

-4.0 6.0 Footnotes a-e are from' TABLE 4-4, Ref. 1.

a. Power was 78% for 4 h prior to shutdown.

b. Plateout of Iodine isotopes measured in decay study, calculated using Eq. (2-7) (Ref. 1), Eq. 12 (this report)

c. Plateout of lodine isotopes calculated from Xenon R/B line using Eq. (2 6) (Ref. 1). Ea. (14) (this report)

d. Iodine R/B calculated from actual Iodine plateout using Eq. (2-6) (Ref. 1), Eq. (14) e.R/BofXenonlineagthetghisggort).133 of I and I

f. Using R/B = 5.0x10, measured value from Ref. 2, p. 6-2 at 70% power.

h.UsingR/S=6.0x10gueR/B=4.0x10"6

g. Using the Ref. I va TheLANLparagetervaluesfromTABLEIareusedwith 6

P = 8.42x10 Watts, f = 70% power, t = 2.8x10 s at po.wer.

-8 g

9

III. IODINE MONITOR ANALYSIS.

^

The' circulating Iodi.1e activity in th'e gas' phase measured by the Iodine monitor is calculated using Eq. (2-8) of Ref.1, given by' 2

A = K P f Y R/B [ - e'(

• Is + r )tyjgt, (p,,p )fg),

p c

p (15) where A

= circulating Iodine activity (C1) from monitor data, g

A

= decay constant (/s),

r,

= purification constant (/s).

p

= plateout constant (/s),

r and the other parameters are defined in Eq. (14).

As steady state is approached (t4 oo) the plateout activ'ity, A, is essen'tially given by A. Solving for the ratio of Eq. (14) to (15), in p

j steady state, yields the result.

r, + r = (A /A,

1) A, p

p (16) coApared to the result, given in Eq. (2-9) of Ref.1,

= (A /A,) A.

rp p

(17)-

Clearly, Eqs. (16) and (17) are equivalent only if r, is much less than r,

.and if Ap is much greater than A. We use Eq-p I "-

c in order ~to minimi.te 4

error propagation in intermediate calculations, :ad to permit consistent parameter analysis in the regime where the first inequality is not.

necessarily satisfied even at steady state. This is important for the 131 g analysis.

Finally, the plateout per pass is defined by Eq. (2-10) of Ref. I as i

T'= 1,- e-f t,

pe (18) where F = plateout per pass (fraction),

t

=cycletime(10.3 sat.70% power).

c The measured Iodine monitor data is sumarized in TABLE III.

l a ".*' t M *c

.X*

4

TA8LE III.

Iodine Monitor Dai:a from TABLE 4 5 Ref. la 135 133 g

g Date Cycle Power Cycle R/B Aa R/8 Aa g

b 6

g

(%)

EFPD x10 Ci "x106 Ci (1) 5/18/78 1

65 80

5. 5.

.795' 8.2

1. 839 (2)11/20/78 1

63 139-5.7

.750 8.2 1.'890 (3) 11/27/78 1

64

~143 5.9

.897 8.5 2.34 (4) 9/29/81 3

70 38 2.5

.20 3.0 0.46

a. A measured Iodine activities measured by Iodine monitor. These A c

data have been multiplied by three to reflect gama counting errors.c

b. EFPD.= Effective Full Power Days.

~

III.A ANALYSIS SUlHARY.

The effective full power days (EFFD) in TABLE III imply that 133I and 135 I monitor activities will be in steady state. Using the data of TABLE TII with Eqs. (14) and (16), we estimate the values of r,+r, which are p

summarized in TABLE IV. We assume that Aj = A. The steady state condition p

is well satisfied.

If we' fit the r values given in TABLE 4_6 of Ref.1, on a In-in p

correlation, we find

,,1.0161 in A + 5.5037 7

(19) which extrapolates to r = 1.96x10-4/s for I (Ag3g! = 9.'966x10~7,/s),but 131 p

is inconsistent with the plateout probe measurement analysis. We are not 3

able to reproduce some of the A results of TABLE 4 6 of Ref'. 1 using p

cumulative yields consistent.with those usually associated with 235 g,

-10 l

I e

-. m

.. r -.

,,.,=,

,.gy,..

8

w If we fit the e +f values given in TABLE IV, with all values equally s p weighted, we find 4

,,1.6139 In A +11.999, e p (20) which extrapolates to r +r

= 3.4x10-5/s for I. Such a low value would.

131 3 p not result in a sensible interpretation at 707, power when r = 3.8x10-5,,

7 s

If we average the r,+ r Cycle 1 data and fit the averaged Cycle 1 with p

the cycle 3 data, we find r,+r

= e.6916 in h + 12.906, l

p (21) which extrapolates to r +f

= 3.1x10-5/s for 131 s p 1,.an equally poor result as the previous case, when r, = 3.8x10 5,,

7 Clearly, the factor three corrections to the 133 135 I and I circulating activity (A ) data must be reexamined. If the correction factor is removed g

from both'i otopes, the resultant fit extrapolated to 131I gives an r,+r qf about 10

, but is still not consistent with the plateout probe p

' measurements.

However, if wa assume that the correction factor is applied to only 135 I, then we obtain the in-in correlation fit given by

,,0.64210 in A + 1.827G 7

(22) which extrapolates to e +r

= 8.7x10-4/s for 1311. Alternatively, applying s

the correction factor to o ly 133I and not to 1.iS 4.3x10-5, 7,7 131 I extrapolates to r,+r

=

7 I, which is low for constant power operation, nd i

difficult to resolve with the probe measurements and known r, values.

Thus, we assume that the factor three correction to the A data is to be applied on1y to the 135 g

1 results Casa B, not to both, Case A, as listed in TABLE IV.

These results lead to a change in the extrapolated r,+r values of 131I by a factor (8.7x10-4/3.1x10-5) 28 due tu the correctio factor of three not being applied to.the 133 I measurement of tlie Iod.ine

~

monitor data.

l l

.,. m :.. u. n..

I

~

e J

TABLE IV.

~

Iodine monitor plateout determination

1. Isotope Case A Case.8 I

135 '~

y A

A r +r

~

.p c

s p C1 C1 1/s

.(1) 172.96 0.795' 6.22 x 10~3 6.22 x 10~3 (2)'

173.74 0.750 6.65 x 10 3 6.65 x 10'3 (3) 182.69 0.897' 5.82 x 10-3 5.82 x 10-3 (4) 96.04 0.20 1.34 x 10-2 1.34 x 10-2 133 g (1) 244.14 1.839 1.23 x 10~3 3.69 x 10-3 (2) 236.63 1.890 1.16 x 10-3 3.48 x 10-3

~. -

(3) 253.07 2.34 9.97 x 10-4 3.01 x 10~3 (4) 80.04 0.46 1.65 x 10 3 4.85 x 10-3 O

O e

t se Md 4

O

. e 9

e e e

• 1

a....,...

IV. PLATE 00T PROBE ANALYSIS l

In order to predict the performance of the plateout probe we must decay the plateout activities from the time of production of the fission product to reactor shutdown, and from shutdown to the time of ac.tivitymeasurement.

We estimate the pseudo-time delay from shutdown to measurement assuming that the reported activity measurements are accurate and represent the total plateout. Then, we analyze the probe ' measurement activities for a self-consistent plateout rate. Finally, we estimate possible uncertainities and examine the effects of such assumptions.

A detailed analysis of the fission product deposition in the diffusion tubes was not performed.4The tubes are finite in length and have various i

porous materials at th'e ends, which frequently results in "an ir$ creased f

activity profile near the tu'be end indicating apparent diffusio'n hold-up.

l The actual diffusion process would not be well approximated by the infinite i

open tube slug flow rodel summarized in Ref.1. Analytical and experimental I

comparisons, of Venerus and Ozisik (Ref. 3) for 137Cs in an isothermal laminar open tube gas stream using a wall Resistence model indicate.that e

quite reasonable comparisons are.obtained when the infinite tube

.Jpproximation is valid. Asymtotically along the length of the tube, the deposited material concentration in the infinite tube model 4 creases l

exponentially. Since the diffusion-deposition process in these finite probe tubes !S essentially two dimensional, with the isotopic diffusion f

coef ficients in Helium dependent upon an Arrhenius relationship, the i

temperature variations predicted along the tubes might well contribute to

(

the deposition profile. Development of a such a detailed modeling capability

{

is deferred until the overall activity balance using the zero dimensional,

{

model is understood.

The graphically (semi-log) 1311 plateout activities reported in Ref.1 Indicate that each of the probe tubes was sectioned into about 13-16 one l

inch length pieces for counting of gamma activities. In the T, 8, and R 131 tubes the I plateout activity increases with increasing position (towards l

the "end plug) in the tube. The measured activity of each piece was i

typically in the range of 10-3 pC/in to 10-2 pC1/in over about haTf the

-[

probe tune length, with a marked increased activity to about,0.36 to'0.4

^

pC1/in in the charcoal trap in the and plug. The apparent relative t

i

, s

.l

.a.c..,

...a-

__7

{

uncertainities are ab:ut a factor 1.5 to 2 at activity lovels of about 10~3 pCl/in. The correspondingly measured 131

! activities for the L and C tubes in the range of 1x10~3 to 4x10~3 pC1/in and 8x10~4 to 1.5x10-2{

respectively, in the interior of the tubes. 8,ut, the end plug of the L tube has an activity of about 2x10-2 pCi/in, and the corresponding value for the i

C tube is about 3x10~3 pC1/in. The reported tube total activities apparently include the activ'ity measured in the end plugs. If we were' to delete the e plug measured 131 1 activity from the analysis, we can estimate the possib'io i

effect on the 1311 plateout rate predictions. Likewise, we estimate the consequences of the hypothesis that activities in the 10~3 pCi range might be undercounted by a factor two.

I IV.A TIME DELAY ESTIMATES

\\

If we assume that the plateout probe an'alysis occurred within two weeks l

of the probe being taken out of the reactor, and that the measurements are j accurate, we' do not find consistent results compared to the prediction from l

gama scans and the extrapolation from the Iodine monitor measurements.

Initially,' we assume for this argument that the gama scan measurements ar

,, accurate, and then examine the apparent effect of measurement-uncertainties-later.

l We can investigate the possibility of a longer pseudo-time delay than 10-20 days '/ rom two viewpoints. First, it was reported in Ref. I that the f

gamma scan of the probe at FSV was approximately 5 mr/h, and at San Dieg it measured 2 mr/h. If the major garma emitter was 51 Cr (T1/2=27.71d),

then simple decay implies that approximately 36.6 days (-9, +12 a if a 25% j uncertainity) elapsed using A = A,e-t,

(

Second, both 51 60 Cr and Co emit gamas, and have measured activities in the probe. The 'T1/2 of o is 5.27 y. If we, assume that the gama activity was composed of these two contributors, then at FSV A,

=Ag+A2= 5 mr/hr, i

(23) and at GA in San Diego, some time t later (at least 10 days),

e,

.-~

,.a....

,..,.....w..

.,w

..~-=

dt+A 4t A(t)= Age i

2e 2 a 2 cr/hr.

.(24)

At th'a time T that the T probe tube activity was measured 51Cr Activity = B e-A T = 5.3 uC1, 1

g and 60Co Activity = B e 2T = 0.75 uC1.

2 i

(25) i Assuming the same source to the probe and tubes, and that the tubes received an equal ratio inside and out of activity, then A /A 2 " 0 /8. Using Eqs.

g 1

2 t

(23)-(25), we therefore obtain i

f(T)=b[1+ae$A'A)T3, h t _g,,,(h 4 )(T-tf = 0, l 2 2

g 2 l

(26) whe~re b = 2/5 = 0.4, and a = 5.3/.75 = 7.067 for the T tubes. Since h =

t l

2.89x10~7/s, and h2 = 4.17x10-9/s, we have sufficient data to solve for g

value of T, the time delay before measurement (if we estimate T-t, the i

measurement time) in order that the ratio of 51 60 Cr to Co activities be consistent. The solution to Eq. (26) yields a range of possible or

{

" pseudo-times T at which the measurements might have occurred

. TABLE V has a

(

summary of these results. Although the ratio b=.4 corresponds to the above quoted values, the ratios b =.5, and.6 are included for comparison. If the i

measurements were 2.5 mr/hr and. 4.5 mr/hr instead of 2 and 5 mr/hr, respectively, then b = 0.55. If the measurements required about 5 days to

{

complete, a pseudo-time delay of about 30 days would be reasonable. This l

possibility is investigated later.

An average value of 35 to 40 days for the pseudo-time delay seems not unreasonable, depending on the assumed uncertainity in the gross gamma scan, and in the unknown length of time before measurement occurred.

Averaging the delays, we estimate 38.3 days as a possible approximate pseudo-time of

{

the probe analysis, if the measurements are treated as quite accurate.

Assuming such a delay time yields an estimated r +rp = 1.4x10~3 131 for g,

3 which is within a factor of two of the Iodine monitor extrapolations with the correction factor of three rerioved from the 133 I data, Eq. (22).

~!

l.

l l

.. O NWr "i: 8 6

.. T.

l 1.

TABLE V.

Possible Measurement Times Consistent with 0verall Probe Activity and 51Cr/ 0Co 6

Tube a

T.t b=.4 b=.5 b=.6

~ days T

7.067 0

40.5 30.9 22.9 5

45.1 35.6 27.7 10 49.8 40.3 32.5 (44.8)

.(35.4)

(27.5)

B,R,L 22.35 0

38.2 29.0 21.4 5

43.1 33.9 26.4 10 48.0 38.8 31.3 (42.6' 3.5)

(26.0)

C 10.0 0

39.5

?L1

'22.9 5

44.3 27.1 10 49.0 3r '

31.9 (43.9) 36.6; (h.0) r If we take in account the five day delay between snutdown tal probe removal, then Eq. (23) becomes

-N t A, = Age i o.+ A2*

2*

I (23')

and using Eqs. (24) and (25), the condition corresponding to Eq. (26) is i,

f(T)=b[1+adb-k)E-Y))e g_y_,,@g-hT=0 l

1 t

(26')

where lf = T, t, 7"y = tg. t, with tt = 10 d, t, = 5 d (from Ref.1), and l

y o

~~

T is the time of the measurement after arrival in San Diego. These results

~

for trare ' indicated in parentheses in TABLE V.

i e

e i '. ~

.x

! ?.t '. s

-Js=

I IV.B ANALYSIS RESULTS l

The expected Iodine activity in the diffusion tubes at shutdown is approximated using Eq. (211) of Ref.1, I =(A fx /): V ) (1 - e~

1) e-N,

2 x

g (27) where I = total specific isotope in tube x (uci).

x A = total circulating activity of isotope g

fromIodinemonitor(uC1),

3 f = flow rate in tube.x (cm j,),

x 3

V = volume of the loo'p (cm ),

) = isotopie decay constant (/s),

tg = time at. power (s).

t2 = decay time to shutdown (s).

The values of f are calculated according to TABLE A-3 and TABLE 3-1 of x

Ref. 1. In comparison, we obtain the values in TABLE VI.

TABLE VI.

Flow Rates in Diffusion Tubes 3

fx (cm /s)

Tube T

B R

L C

~

GA 14.6 15.9 13,2

.38 2.1 LANL 14.6 15.2 13.2

. 38'd 2.29 For later estimation of the r +r rate we calculate in TABLE VII the s

p predicted plateout, A, and the A*j, decayed to shutdown, for the' four g

operational time perio'ds prior 'to shutdown.

~

g 8

49 9

J

.s

TABLE VII.

Predicted Circuit Total Plateout.

Dates at

% Power /

t/

A A*j g 2 j

0 Power R/Bx10

'x10

• C1 C1

...___....._--..............................:............. p (1) 11/7 11/9 100/5.0 2.2/0 19.52-19.52 (2) 11/4-11/6 70/4.5 1.7/2.2

.9.73 7.81 (3) 10/9 10/27 30/3.5 16/11 16.59 5.65 (4) 8/S.10/9 70/4.5 56/26 62.39 4.32 1981 LANL 11/9/81 Shutdown A*j = 37.3 C1 GA

. *11/9/81 Shutdown A*j = 45,c Ci

_____.____....____.__...___________..____________,_u.;

A*, values decayed to shutdown.

LARL parameters:

Y131r = 0.0277, h131g = 9.966x10-7/s GA parameters: Ref. 1 Y131g = 0.0320, h131g = 9.7x 10-7/s In order to evaluate the measured I we first need to calculate A from x

c Eq. (15),

A = K P f Y R/B [1 - e4+rs + r )t3 / [1 + (r +r )/h],

p c

p 128) and then from Eq. (27) I, decayed to shutdown, is x

t ),-ht.

I =(A fx / h V ) (1 - e t

2 x

c (29)

The cumulative values of Ix (over the four reactor power periods in TABLE VII) at shutdown are then decayed to probe measurement time, t, after o

j shutdown.

Ix,' total (t ) = Ix, total shutdown.*

o (30)

~

Solving Eq. (28)-(30) for 'r +r requires assuming a value for t. In 3

p o

order to simplify the overall process, we also iterat.e on the ratio 18-

• ', 4 t-.

C*I

/I, total shutdown

• T1 T

the ratio of.131I in tube T during the last rea'ctor 'cperational time period 131 to the total I in tube T at shutdown. This ratio e converges rapidly to about 0.52. Otherwise, we should have to use a more complicated iterative method involving all the tubes at all the power periods.

Thus, assuming that we know all the parameters except r +r, we can use 3 p Eqs. (28)-(30) to obtain an equation for the self-consistent determination of r +r, namely s p f(.a) = [1 'a-(1+a)ht3f(g,3) - b'e-hto = 0, (31) where a = (r3 + r )/h, p

b = c [A V I 3/EIx (1-e

1) K P f R/B Y ]'

T, measured

= 4iT133 10-5 c, 131 using the' parameters for the 1 T tube, h131 9.966x10-7/s,Y1'31 3

g =10 g=

c,3 g

0.0277, fT = 14.6 cm /s (from TABLE VI), V = 2.1x10 T, measured

• 8

.56 uC1. K = 0.85, P = 8.42x10 W, f = 707, R/B = 5x10-6, t,= 38.3 d =

6 5

3.31x10 s, and t = 2.2x10 3,

We use the resultant r +r rate in the in-in correlation with the Iodine 3 p monitor data to obtain the fit of r +r versus h, recalculate the I for all 3 p x

the tubes, and determine the fractional value of c. The results of these 131 calculations for I with c = 0.52 are given in TABLE VIII, for r +r

=

3 p

1.625x10-3/s. The in-in correlation is given by 0.480267 in A + 0.2170476 r r =

p

(.12) 9.

4 -

. -s

.~..._..,_n..__.._

TABLE VIII.

. Diffusion Tube Predicted Plateout-Activities at t, and Shutdown A

A*

A A*g I

I I

I I

g g

T B

R L

C C1 C1 C1 Ci pCi pCi pCi pC1 pCi

. (1) 19.52 19.52

.0608.0608

.306

.319

.277

.0080 0480 (8.37)(8.71)(7.56)(.219), (1.31)

(2) 9.73 7.81

.0383.0307

.122

.127

.111

.0032 0192 (3.35)(3.49)(3.03)

(.088)(.525)

(3) 16.59 5.65

.0128.044

.085

.090

.078

.0023 0136 (2.37.) ' (2.46) - (2.14)

(.062).(.371)

(4) 62.39 4.32

.0383.003

.073

.076

.066

.0019 0114

~

(1.98)(2.06)(1.79)

(.052) (.311)!

i Totals 37.3

.139

. 586

.612

.532

.0154 0922I (16.1) (16.7) (14.5) (0.421) (2.52)

Measured Probe values

.56

.58

.55

.036 045 A*j and A*c are decayed to shutdown, 11/9/81.

I values, x = T, B, R, L. C are decayed to assumed measurement time l

I

=

38.3 d.

I values in ( ) are values decayed to shutdown,11/9/81.

x e*

G D

m.

e g.

,4 4

TI

• s

[_ [ t

,N._

IV.C UNCERTAINITIES We consider the ef fect of possible uncertainities in the probe measurements, and the correlations' of the plateout rates and/or pseudo-time delays that could result. Util! zing the previous observations about the gama scan measurements, we might postulate that the value of IT measured in the charcoal of..the end tip should not be counted as part of the T probe plateout activity. This would reduce che measured value from IT = 0.56 pC1 to about IT = 0.16 pC1. If there wr, indeed a factor,two counting error relative uncertainity in these measurements, then the value might be IT*

l 0.32 pC1. If, on the other hand, the value in the end tip were to be taken as part of the plateout activity, and the tube values were a factor two too low, then the.value sight be as high as. (2 x 0.56 pC1) IT = 1.12 pC1. For thesn'two extreme assumptions we calculate the' r +r values at various g

p pseudo-time delays before measurement that would predict the assumed I T

)

plateout activity result. The plateout ' rates are sumarized in TABLE IX, We 131 observe that if the 1 activity in the charcoal end tip is not credited, then the plateout rate is significantly increased at a particuTar pseudo-time delay. Howeve'r, if the o.verall ' activity rate were.indeed undercounted by as much as a factor of two, then approximately the same

%1ateout rate at a pseudo-time delay of 38.3 days would be valid for 30

~

days. Thus, a factor of two undercounting of the plateout activity would correspond to a decrease of about eight days in the pseudo-time delay between shutdown amd measurement,1f agreement is to be ma'intained with 135 133 measurement. As an example, we fit the I and I Case B data of TABLE IV 131 with the 1 r +r value of 1.76x10-3/s to obtain s p r g = e.473063 in A + 0.136719 0

(33)~

which yields the fit extrapolation for 131I of r,+r

= 1.66x10~3/s. This p

value, applied to a pseudo-time delay of 30 days, yields predicted acti.vities that can be compared to the factor of two undercounting hypothesis. The results are sumarized in TABLE XI. Using this argument, we can fit the plateout activity data at.a 30 day measurement time if" we assume that the probe activity data measurements were undercounted by a-factor two.

1, 9

~

s : z c..Ec w& M : W...

i If that were tha esso, th0n the plateout fraction at 705 power would correspond to 1.7% per pass using Eq. (19). We also observe in TABLE X that the circulating activity is not significantly different at shutdown time from TABLE VIII, only the plateout activities are doubled, corresponding to the assumption made that the activities were Undercounted.

TABLE IX Plateout constants versus t, for 131 Various Possible II values x

3 r,+r, x 10 /s I

1.12

' 0.56 0.32 0.16 uC1 x

t d a

10 21.37 42.8 74.8 149.6 20 4.16 8.33 14.6 29.1 30 1.76 3.52 6.16 12.3

. 38.3 0.86 1.72 3.01 6.02 40

'0.74 1,49 2.60 5.21 50 0.31 0.63 1.10 2.20 ee W

O e.

W i

1 S.

~

4 a.

'*'?

+ -

. +-

,,a

4 e

TABLE X 131

'p Diffusion Tube Predicted I Plateout Activities at t,

= 30 days A

A*i A

A*c I

I I

I I

i c

T B

R L

C C1 C1 C1 Ci pCi pCi pCi pC1 pCi

....-_______.__.____.1__

(1) 19.52 19.52

.0595.

0595

.617

.642

.558 0161.'0967 (2) 9.73 7.81

.0375.0301

.247

.257

.223

.0065.0387 I(3) 16.59 5.65

.0128.044

.175

.182

.158

.0039.0275 o- (4) 62.39 4.32

.0383.003

.147

.153

.133

.0039.0231 Totals 37.3

.135 1.186 1.234 1.072

.0304.1860 2 x Measured Probe values 1.12 1.16 1.10

.072

.090 A*j and A*e are decayed to shutdown, li 3/81.

IE =alues, x = T, 8. R, L, C are decay to assumed measurement time v

30.0 d using r

= 1.6607x10 values in ( ) are {+a10es decayed to shutdown,11/9/81.

I

_x...._________.__.___.._____________________________________.

ee e

e ee e e e

e 23 9

, pe

, =

5

. I' '

~^*1-

~,,,.,,;,

.g* g.

9 c.,m'm.-pg p%

- * + - - - -

~~

p

Vo CONCLUSIONS l

The Iodine decay study of the circulating activity of 133 135 I and g

yields. agreement with experiment in the range of 13-20%, which is judged I

quite reasonable.

The Iodine monitor plateout analysis, based on 133 135 1 and I, yields 1

reasonable result $ when extrapolated to 131 1 provided the factor three correction facto'r applied to the circulating activity data is only applied 135 133. We obtain A = 0.14 Ci for I at shutdown, as shown I

131 to I and not g

j in TABLES VIII and XI. These results are based on the self-consistent solution developed for the plateout probe, using r +r

= 1.625x10-3/s for 131. However, the extrapolation of the modified 33 135 1

and Idata(TABLE

.IV, Case B) yields the vSlue r,+r, = 8.7x10-4. Using this latter value, the

[

value at shutdown of A = 0.14 C1 x 1.87 = 0.26 Ci for the modified Iodine g

monitor data only. This uncertainity of 52% is not unreasonable. The Iodine i

monitor performed reasonably well.

In the determination of the plateout probe analysis, we would argue that

{

131 all of the I expected tube activities must be decayed to s'hutdown time, i

~~

as given in TABLES'VIII, X, and XI. We would clarify the GA argument ~ (Ref.

b;-p. 4-40) that "only the last few weeks of reactor operation are.

important" because of the eight day half life of 131 : Direct computation f

1 from TABLE VIII indicates the reactor operation between 8/5/81' and 10/27/81 contributes to 27% of the shutdown inventory activity and 34% of the' circulating activity, which results are probably non-negligible.

We would estimate that there were probably more than 10-20 days elapsed I

time between shutdown and the measurement of the diffusion probe activities.

[

This view is based upon the gama scans, and the measured 51 60 Cr/

Co ratios I

in the diffusion tubes. The Iodine monitor data extrapolated r,+r rate for 131 p

1 is not inconsistent with a time delay from shutdown of the order of 35-40 days. On, the otherhand, if the activities were undercounted by a i

factor two, a self-consistent solution to the plateout analysis is obtained l

for about 30 days delay from shutdown to activity measurement.

Combining these estimates and assumptions in a self-consistent iterative evaluation, we find th,at reasonable agreement is obtaliied betseen the'

~

predicted and measured plateout probe activities for the T, 8, and R tubes 4

,I,,

,~.4 g

e

as shrwn in TABLES VIII and XII. Tha' excsptions ara tha L (58%) and C (1045) tubes.

We agree with the plausible GA explanation that possible C t,ube plugging contributed to the 104% disagreement. between measurement :nd prediction in TABLE XII. However, both the C and L tube apparent behavior might also be understood if the measurements af the segme.its of those tubes have an uncertainity of 0.01 - 0.02 pCi, corresponding to about a factor 2 at the 10'3 pC1/in level. This would be consistent with the T, 8, and R values which exhi6ft reasonable agreement for the parameters used.

Two sources of error uncertainit.y still exist in this analysis. First, the assumed measurement time delay of 38.3 days, based on the gama scans.

If the probe scans were in error, or were undercounted by a factor two in the probe measurements, then 30 days delay would be more reasonable and overall agreement could still be obtained. However,.since the measurement dates were not reported' in' Ref.1, they might well be under 30 days, which would increase the plateout rate errors. Second, we assumed that the Iodine monitor data should be corrected for 135 133. Other alternatives I and not I.

that were examined did not seem reasonable, but the application of a factor l

three correction may be somewhat arbitrary. Consequently, we estimate that i

the overall accuracy of the estimated plateout ra'tes should only be within

.about a factor 3-5 at best. It seems feasible, therefore, to find reasonable agreement with the plateout measurements, within their. associated I

measurement uncertainities, if we knew them, without having the apparent f

factor of 200 discrepency in the prediction of plateout activities quoted in f

Ref.1.

f f

5

)

-l

)

_25-

..i l

M. ~. : _ _ _ _

'l m. :

TABLE XI.

Comparison of GA and LANL predicted {31I activities at shutdown on November 9, 1981.

I C

A(C1), A(C1) T(pC1) B(pci) R(pC1) L(pci) C(pci) t GA 45

.37 102 112 93 2.6 14.8 LANL 37.3

.'14 16.1 16.7 14.5

.42 2.52 GA values from Ref. 1, TABLE A.5, using r +r = 3.03E.4/s.

LANL values, decayed to shutdown, use r + Ectivity.I.625E 3/s.

1 A = inventsry'"ictivity, A = circulati g c

T B,R,L, and C = diffusion plateout tubes. t 9.___38.3 d

=

TABLE XII.

131 Comparison of GA and LANL 1 predicted activities.

with plateout probe measurements.

v...

T(pC1) B(pC1) R(pC1) L(pC1) C(pC1)

GA 102 112 93 2.6 14.8 LANL

.589

.612

.532

.015

.092 EXPT.

.56

.58,

.55

.036

.045 LANL

+5%

+6%

-3%

-58%

+104%

/ EXPT GA values from Ref.1. TABLE 4-2, Iodine monitor prediction.

Although footnote (a) of TABLE 4-2 states that these values were decayed to reactor shutdown, that appears doubtful.

LANL values, decayed to measu~rement. time, assuming 38.3 days.

EXPT values from Ref. 1. TABLE 4 2.

e.

_,.E__,

m I

REFERENCES

1. R.D. Burnette, " Radiochemical Analysis of the First Plateout Probe from the Fort St. Vrain High-Temperature Gas-Co,oled Reactor," General Atomic Company report GA-A-16764 (June 1982).

I

2. J.R. Wol fer and N.L. Baldwin, " Fort St. Vrain Startup Test B-13 Radiochemical Analysis of the Primary Coolant," General Atomic Company reportGA-015109(1978).

3. E.R. Venerus and M.N. Ozisik, " Theoretical Investigations of Fission Product Deposition from Flowing Gas Streams," Nuc. Sci. and Eng., vol. 26, pp.122-130(1966).

8 -

e e

S G

0 O

ee e

S 6

7

.u;. w,,..

c.

_.e

g. __

.