ORNL-TM-1796.txt

 

OAK RIDGE NATIONAL LABORATORY
“operated by

UNION CARBIDE CORPORATION w
NUCLEAR DIVISION

for the
U.S. ATOMIC ENERGY COMMISSION

ORNL- TM-1794

 

copy NO. 4 52

DATE - March 10, 1967

 

 

v Ll . f".;s
wr, LA SED FOR AENOURCM! CE:71 PRIC
. 5 I8
{¥ NUSLEAR SCIENCE ABSTRAC J ~
ne 3900 MN L 62
-_
.
- THE REACTIVITY BALANCE IN THE MSRE
e 2T = Y
*
J. R. Engel
B. E. Prince
fl‘ht NOTICE This document contains information of preliminary nature
- and was prepared primarily for internal use at the Oak Ridge Naticnal
Laboratory. It is subject to revision or correction and therefore does
not represent a final report. THIS DOCUMENT MAS BrEsl B VIEWED.
“NO INVENTIONS OF /o 3 1 TLREST

10 THE AE.C. ARE gkl THE}EHL
S/ 7
 

 

 

LEGAL NOTICE

This report was prepared as an account of Government sponsored work, Neither the United States,

nor the Commission, nor any person acting on behalf of the Commission:

A. Makes any warranty or representation, expressed or implied, with respsct to the accuracy,
completeness, or usefulness of the information contained in this report, or thot the use of
any information, apparotus, method, or process disclosed in this report may not infringe
privately owned rights; or

B. Assumes any liabilities with respact to the use of, or for damages resulting from the use of
any information, apparatus, method, or process disclosed in this report.

As used in the above, *‘person acting on behalf of the Commission’ includes any employes or

contractor of the Commission, or employae of such contractor, to the extent that such employse

or contractor of the Commission, or employee of such contractor prepares, disseminates, or
provides access to, any information pursuant to his employment or contract with the Commission,

or his employment with such contractor.

 

 

s

-

&
iii

“'4""’- f,kvu

g

f;%i 'CONTENTS
ABSTRACT & a'e 4 o o v o o o oe o e et et e e e e e e
INTRODUCTION & o « o o o o o o o o o « s ¢ a o s o & ;.. e e e e
TESCRIPTION OF THE REACTIVITY BAIANCE . . . . « . o v v ov v o . . .
The Reference Conditions e e e e e e g e e e e e e o o
The General ‘Reactivity Balance Equetion B
~ Control-Rod Worth e s e e e s e e e e e e ; e e e e e
' ExcesséUranium Reactivity Worth & & v 6 6 ¢ ¢ ¢ 6 @ o o o o o
Power Coefficient of Reactivity C e s s e s w e s e e s e e
=Samerium Pbisoning e v er et e e e e e e e s e e e e
Xenon-135 Pcisening . .i; e e e e e e e e e e e e e e e
Density Effects of Circulating Bubbles on Reectivity « s e e
 Isotope Burnout EFFECTE s « « o o o v o o o e v e e e .
. EmamNCEmnmou-meAmummmn;.'............
“"{; Iow-Power CRIculetions . « ¢ v v s o v v o v 4 b v e e e e

T

e * ¢ s ¢ a g »

Complete Calculations

i &

IN‘I'ERERETATION OF RESULTS . . . . R
' Previous Reports of Results -;“.;. . .

Utility of Residual Reactivity « v e

Effects Not Treated -

 Operating Limitations e e e

Conclueions
REFERENCES

-oo'.oo-nc-:'. . e+ 0w

* & = 4 e & e 8 e & s e * + e =

 

'i' "
| This repofl was pre

- of any information, ‘Ppa?
-'ywwaeywmfl’fl o

B. Assumes any liabil

- tractor o
- p}oyee Ol' ei!n
-7 guch employee or €N
“~i giapeminates, or prov
L withthe Commission, 0F

e T

Intermediate Calculetions e e e e e

. 9 . @ - . . .

EGAL NO

pared as an
COmmllfl‘lon nor

ness of
e$ method, of Pl’°°°“ disc

mes wm‘ respect

wcel 8 to,
tde;i . emplo yment with suc

    

TICE
ed work.
saccount of Govemm“n :te :::fn:;’;e Commission:

 

}-"1

o

ViV R R R EEW W
M H O U kO d®IKYEEEE

Neither the United

 
- ‘} |
diz” )

o

\
) ‘r

10,

LIST OF FIGURES

Title

Comparison of Control Rod- Reactivity from Experimental
curves and from Least-Squares Formule.

First Order Decay Schemes for Production of Samarium

- Poisons. in the MSRE.,'

‘ Effect of Vblume of Circulating Gas .on Transient Buildwp

of 135Xe Reactivity. Step increase in power level from

0-to T.2 Mw; bubble-stripping efficiency, 10%,

MSRE BunaNb. 7;—f

Effect of Bubble-Stripping Efficiency on Transient Buildup
of 15¥e Reactivity. Step increase in power level from

0 to 7.2 Mw; volume percent circulating bubbles, 0.10;

MSRE Run Nb. 7.

Effect of Vblume of Circulating Gas on Transient Buildup

of 135Xe Reactivity.: Step increase in power level from O
to 5.7 Mw; bubble- stripping efficiency, 10%;

- MSRE Run Nc. 8.

Effect of Bubble- Stripping Efficiency on Transient Buildup
of 235Xe Reactivity. Step increase in power level from.0
to 5.7 Mw; volume percent clrculating bubbles, 0.10;

MSRE Run No. 8.

Effect of Vblume of Circulating Gas on Transient'Decay of

'135%e Reactivity. Step decrease in power level from 5.7 Mw '

to 0; bubble-gtripping efficiency, 10%;.MSRE Run No. 8

',Effect of Bubble-Stripping Efficiency on Transient Decay of-
135Xe Reactivity. Step decrease in power level from 5 T Mw

to 0; volume percent circulating bubbles, 0. 10,

. MSRE Run No. 8. -

| Effect of Bubble~Stripping Efficiency on Transient Decay of
‘135Ye Remctivity. Step decrease in power level from 5.7 Mw

"~ to 0; volume percent circulating bubbles, 0. 15, o

MSEE Run No. 8. |

| Effect of Bubble—Stripping Efficilency on Transient Decay
- of ¥35%e Reactivity. Step decrease in power level frqm
7.4 Mw to 0; volume percent circulating bubbles, 0 10;

','MSRERunNo 9

;

12

18

20

21

22

23

2l+

[26

27

 
 

Fig, No.

1.
12,
13.

LS
15,
16.

vi

mitle

Effect of Bubble-Stripping Efficlency on Transient Decay
of 135Xe Reactivity. Step decrease in power level from
T4 Mw to-0; volume percent circulating bubbles, 0. 15,
MSRE Run No. 9.

Effect of Bubble-Stripping Efficlency on Transient Deceay
of 135Xe Reactivity. Step decrease in power level from:
T.k Mw to O; volume percent circulating bubbles s 0.10;
MBRE Run Nb. lO. .

' Effect of Bubble-Stripping Efficiency on Transient Decay'
of 135¥e Reactivity. Step decrease in power .level from
7.4 Mw to 0; volume percent circulating bubbles » O. 15 3

.MSRE Run No. 10, o s

Results of Modified Reactivity Balances in MSRE
Results of Complete Reactivity Balances  in MSRE -.
Long-Term Drift in Residual Reactivity in MSRE.

.

28

- 29

30

w
L1
W

o
Ҥ .

 

 
 

[ )

*}

ad ( e}

'THE. REACTIVITY BAIANCE IN THE MSRE

.'J;_R;ZEngeI i - B, E, Prince
ABSTRACT

Reactivity - balances have been calculated for the MSRE since
~ the start of power operation. After an initial period of. manual
- - caleulations while the computer was. being set up, machine calcu-
letions were started which are now routinely performed every
5 minutes while the reactor is in operation. The calculations
- are carried out by an on-line (Bunker-Ramo. Model 340) computer
‘using current values of reactor parameters such &as temperature,
- power, and control-rod positions. All the known factors that
_have significant reactivity effects are computed and a residual
reactivity required to keep ‘the reactor just critical is evaluated

R Early results showed that the 135Xe poisoning in the MSRE
(~ 0.3% 8k/k at 7.2 Mw) was lower than was expected and results
~ during xenon transients were used to construct a model to de-
- scribe the xenon behavior. Subsequent results have been used
to monitor the reactor operation for the appearance of anomalous
. reactivity effects. After the equivalent of 95 days' operation
at maximum power, the residual reectivity is + 0.05 + 0.0L4%
~ 8k/k. This indicates excellent agreement between the predicted
- end observed behavior of the reactor. No significant anoma-
~ lous effects have been observed.? -

 Prior to the start of reactor operation, a limit of * 0.5%
| Bk/k wes imposed on the residual reactivity as & criterion for
critical operation of the reactor. This 1limit has not been
approached. ' e '
" INTRODUCTION

- The'availabilityfof ah'oneline'digital*COmputer‘for the purpose of =

'{'data logging and routine computations for the MSRE has made’ feasible the
_: continuous monitoring of the important reactivity effects associated with ‘
: power operation of the reactor. Steady power operation requires that &
'd:balance be maintained between the rate of production of neutrons from
"vrfission and their rate of disappearance due to absorption and 1eakage to
-fthe surroundings.' The reactivity is a qpantity introduced . to describe
' physical situations in vhich these rates do not balance. It is convenient'

 
 

 

to express this quantity as the algebraic fraction of:the production rate
which equals the net rete of accumlation (+) or depletion (- ) of neutrons

- in the entire reactor, i.e.,

Total Production Rate —-Total Depletion Rate

Reactivity = Total Production Rate

- In one sense, therefore, the reactivity makes its appearance physically
only when the reactor ‘power level is changing. At steady power, the reac-
tivity must be zero, and any attempt to ascribe separate reactivity con-
ponents (both positive and negative) to the steady state is merely a ;
convenient bookkeeping device. If we use this device to monitor the ‘re=
actor ‘operation and find that the algebraic sum of the calculated components
1s not zero, this may mean either that the calculations of the individual
known effects are in error, or that there are unknown, or anomalous changes
occurring in the neutron reaction rates which are not accounted‘for in the
calculations. Power operation of the reactor is a complex situation where
many effects are slmultaneously influencing the neutron reaction rates,

The device of separating the effects according to & reactivity scale
allows individual experiments or computations to be used &s .an aid,in
interpreting the whole process. Thus, continuous monitbring'of the com-
ponent reactivities serves both to test our confidence inaindividual
measurements and, potentially, as a means of detecting and interpreting
anomglous changes in the remction rates during operation. ,

As an illustration of these general considerations, we describe in
the following sections the basis and approximations used. for the reactivity
- balance calculation for the MSRE. We emphasize at the outeet that the
methods and quantitative results of amelysis of MSRE operation to date are
still subject to possible future modifications. 1In discussing‘the.results,
wherever possible we will attempt to indicate the level of confidence
‘in present celculations of the individual reactivity effects.

. .r
f)( W

 

 
-m(- »

-};

)

 DESCRIPTTON OF THE REACTIVITY BATANCE

The Reference Conditions | .., - - b_, L

- If we are to monitor changes in component reactivity effects during ;
operation, it .1s advantageous to choose a starting, or reference condition
which can be defined by experimental measurement with relatively little

- error or ambiguity The reference conditions chosen for the present work

correspond to the. just critical reactor, isothermal at 1200°F, with fuel
circulating and free of fiselon products, and with all,three_control rods

'withdrawn to their upper limits (51 inches). The uranium concentration for
‘these conditions, as well as the increase in uranium concentration required

Vto compensate for a range of control-rod insertions and isothermel tempera-

ture changes. was established during a. program.of zero-power nuclear experi-~

'ments carried out in the summer of. 1965 (Ref.-l) ~In this progrem, inde-

pendent measurements of control-rod reactivity worth. (period — differential
worth experiments and rod drop: integral worth experiments) were used. to

 determine reactivity equivalents of uranium concentration changes and

isothermal temperature changes. ..
The’GeneralfReactivit Balance Equation

' The equation describing the general situation when the reactor is

operating at some intermediate steady power level includes terms repre-

senting, relative to the reference state, 7
,V_.l,'; the total excess uranium added before increasing the power,

rf'é, l‘the poisoning effect of the rod insertions, and

3. the power and time-integrated power dependent effects of

- \changes in fuel and graphite temperature levels and spatial

_'pdistributions, uranium burnup, and fission product buildup

o ' "(135xe 14QSm 151Sm, ‘and non-saturating fission products)
This list includes the most important effects of substantial power genera—
tion There are, however, other known effects of smaller magnitude arising
from isotopic burnup which must be added to this list. These include-"‘_

1. ‘the burnout of the small amount of 1ithium—6 present in ‘the

clean fuel salt,

 
 

 

2.  burnout of'residualfboron-lo‘from”the unirradiated graphite

moderator,

3. production of plutonium-239 from absorptions in uranium-238

- and _ . :
L. ' Changes in the concentrations of uranium-23h and. 236 in the
fuel salt due to neutron sbsorption. O

There are, in addition, other known reactivity effects which . cen be shown

to be insignificant in the MSRE, such as photoneutron,reactions in the
berylliun in the fuel salt, and several high-energy neutron reactions. *

This completes the list of ‘component reactivity effects only if we assume
- that the structural configuration of the graphite stringers and the associ-

‘ated matrix of fuel-salt channels undergo no significant changes during

the poWer-generating?history of the core. If changes in the fuel-moderator

 geometry-are induced, for example by nonuniform temperature-expansion
effects or curulative radiation-damage effects on the graphite, this

could eppear &s an anomelous reactivity effect, not explicitly accounted

| for in the reactivity balance.

‘ There is substantial evidence that another special reactivity effect
is of importance in the operation of the MSRE. - This arises from the en-
traimment of helium-gas bubbles in the circulating fuel salt, through the
‘action of the xenon-stripping spray ring in the fuel-pump tank. These

minute, circulating helium bubbles would be expected to affect the reac-
tivity in two ways, by modifying the neutron leakage through an effective ;

reduction in the density of the fuel salt and by providing an additional
- sink for 13SXe, thereby reducing the effective xenon migration to the .
graphite pores. (This will be discussed in greater detail in a. later
section.) o | - |

We can summarize the preceding discussion in the form of & general o

equation for the reactivity balance. By using the symbol K(x) to repre-’li

sent the algebraic value of the reactivity change due to component x and
'.grouping terms which can be treated similarly in the calculations, one .i
obtains- | | | S

L

/

~¥

 
4\\(“ ”

.

e

4

0 = K(Rods) + K(Excess 25)) + K(Temp.) + K(Power) + K(Samarium)
+ K(xcnon-135) + K(Bubbles)
+ K(Isotope Burnout) - |
+ K(Residual) "'_" e o o Q)

'The final term on the right hand side of the above equation includes any
- small residual effects known to occur which are not explicitly accounted

for in the calculation (such as 1ong-term effects of gadolinium burnup
on the control-rod reactivity), effects of any anomalous changes in the
graphite-fuel salt configuration, permeation of the graphite by salt, or
changes in fuel-salt composition.' If, in addition, we consider each term
in Eq. 1 to represent our best estimate of the individual effect, rather
than the value we could compute with perfect information, the final term

in Eq. 1 will also contain any residual reactivity corrections due to

errors in calculating the other terms. In order to make this report

'reasonably self-sufficient we'will give a brief review of the basis of

calculation of each term of Eq. l, in the order given.

antrol-Rod Wbrth |
‘Of the terms in Eq. 1, ‘the rod worth, the 235U reactivity worth, ‘and

the temperature-level resctivity effects [K(Temp.)], are based on zero-
power experimental measurements. Because the uranium and temperature reac-

| tivity effects are inferred from the control-rod calibration experiments,
~and also becsuse the magnitude of other known power-dependent reactivity

effects are evaluated: according to ‘the. “time variation of the. control-rod
position following & change in power 1evel, accurate knowledge of the rod .

_worth is vital to the successful interpretation of the: reactivity balance.\'
| :The control rods were. calibrated by means of rod bump-period measurements :
| nmade with “the reactor at zero power (1. e., with negligible temperature

feedback effects), and with the: fuel- circulating pump stopped. These were

| made during a period of uranium additions sufficient to vary the initial

- :critical position of one rod (the regulating rod) over 1ts entire length -

 of travel. At three intermediate 235U concentrations, banked insertions
" of the two shim. rod required to ‘balance specified increments of withdrawal

of the regulating rod were measured..-In this way, various ‘combinations

 
 

 

 

C
%) <A

_of shim- and regulating-rod insertions equivalent in their reactivity
poisoning effect were obtained. Rod-drop experiments were also performed
~ at three intermediate 35U concentrations. 1n'these'experiments, the
e@uivalent integral negative reactivity'insertion of the rod!"falling from
its initial eritical position to its screm position was measured.lz
Agreement between the integral of the differential worth measurements and
the integral reactivity obtained directly from the rod-drop experiments oo
ves found to be within 5% of the total negative reactivity insertion in- |
volved in each experiment , ’ _ o

| The reactivity vs position calibration curve for the regulating rod,
and the results of the three experiments measuring equivalent shim- and .
| regulating-rod combinations were next combined with a theoretical formula
for the reactivity worth of an arbitrary shim-regulating rod configuration.
The theoretical formula contained several parameters which were adJusted_
s0 that the formula provided a 1east squares fit to the experimental _'
measurements. Derivation of  the formula for the rod worth and discussion
of its application &are given in Ref. 3. The result of this analysis is
shown in Figure 1. Here, the solid sample points are taken from smooth = | v
curves through the experimental data. As Figure 1 indicetes, the smoothed
data could be fitted very closely with the theoretical rod-worth formula,
except for small errors at the extreme positions of the.rods‘(full,inser- 3
tion or withdrawal). No important restrictions in the use of -the formula
srise from these errors, since its purpose is primarily for interpolating:
for therreactivity worth of intermediate shim-regulating rod combinations
not specifically covered in the three groups of experiments described -

 

 

above. It provides a convenlent ‘means of rapidly calculating the reac~ :
tivity equivalent of the rod configuration during reactor operation, by o
means of the BR-340 on-line computer. One restriction in the practical
use of the formule on which Figure 1 is based should be noted, however. -
It should only be applied in regions of rod travel and excess reactivity -
covered in the zero-power calibration experiments (i.e., magnitude of =
reactivity less than or equal to the worth of a single rod, moving. through
51 inches of travel). Modifications of the least-squares formuha would be

»—x‘ i

required to cover a larger reactivity range. . I o :’!'L\'x ksj

¥

 
ifl( N

”?t

-)

N

- ORNL-OWG 66-4758

T T T L

 

 

 

 

 

2.2
20 |- L
[ POINTS OBTAINED FROM ROD-CALIBRATION [ T e
EXPERIMENTS | L /
1.8 [~ e CALCULATED FROM LEAST-SQUARES , —— L
I - FORMULA ' : /.
161  (ROD REACTIVITY NORMALIZED ol | A

 

. 74 kg 2-’*-"u IN LOOP)

 

 

14 |

 

 

 

1.2 f—

 

 

 

 

1.0

 

 

 

0.8

 

 

 

 

0.6

INTEGRAL REACTIVITY WORTH (% A/k)

 

 

04}

 

 

0.2 |

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20 24 28 32 36 40 44 48 52
POSITION or ROD MOVED Cin, WITHDRAWN) -

°'F1g 1. Comparison of Control Rod React1V1ty fram Exper1mental Curves
and fram Least—Squares Formula.i',Q' , S

 
 

Excess~-Uranium Reactivity Worth

Relative to the reference conditions'defined,in the'preceding.part
~of this report, the total excess 235y i1s equal to the amount added during
the zero-power experiments, minus the amount burned'during power operation
of the reactor, plus the amount added to re-enrich the fuel salt when the
burnup becomes sufficient.” Corrections must also be introduced for rela-
tive dilution effects each‘timefthe‘reactor,fuel.loopfis drained. and mixed
with the fuel salt "heel" remaining in the drain tanks du:ring operation.

The reactivity equivaelent of the excess uranium was determined from
‘the zero-power experiments by measuring the amount of control-rod insertion
required to balance each addition of 235U then using the independent cali-
bration of reactivity vs positionrto ‘determine the_incremental reactivity
warth of the 235y. 'Results of these measurements,l expressed in terms of &

- o - | - .
concentration coefficient of reactivity, gave 0.223% increase in reactivity

for a 1% increase in 235U concentration. This was within approximately 5%
agreement with the theoretical calculations of this quantity..‘ '

Temperature-Ievel Reactivity EffEct

When the core temperature is maintained spatially uniform; a change
in this temperature can be related both experimentally and theoretically
to the core reactivity in an unambiguous manner. The method used to _
measure the isothermal temperature. coefficient of reactivity during the
zero-power experiments consisted of varying the external heater inputs and
allowing the just critieal reactor to ecool slowly and uniformly while
measuring the change in regulating-rod position required to maintain a
'constant neutron level In these experiments, the fuel was circulating
and the system temperature was taken to be the average of a preselected

set of thermocouples distributed over the circulating system. ‘The change .

 

. | - S
, At the time of writing of this report, no further capsule additions
beyond those of the zero-power experiments have been made. :

*In the ensuing sections we will often use the normal symbol, Bk/k
to represent reactivity.

-
L.

 

. A

e e i mitran —eiap et T o

 
C”

-

)

-y

X

-in rod position corresPonding to the temperature change was converted to -

*
reactivity by again using the rod calibration curve. These experiments
measured the combined effect of a uniform change in fuel and graphite
temperature. The- mesasured total isothermal temperature coefficient of

reactivity was -T.3 x 1075 ( F-l)

Experiments were also performed to separate the component effect of
fuel and graphite temperatures.?  This was done by stopping the fuel circu-
lating pump, raising the temperature of the circulating coolant salt as

‘well as the stagnant fuel salt!in:the heat exchanger, then restarting the

fuel pump to pass fuel salt that was hotter than the graphite through the
core, The resctor was maintained eritical with the power level controlled

'by the flux servo. The change in control-rod position and the output of a
- _thermocouple in the reactor-vessel outlet was logged digitally at quarter-

second intervals. The value of the fuel coefficient obtained from these
experiments was -L4.9 x 10;5'(°F’1), about 20% higher in magnitude than the
calculated fuel temperature coefficient. TheSe:experiments,rhowever, con-
tained a relatively large band of uncertainty due to the inherent diffi-

' culty in introducing proper time dependent corrections,

Power Coefficient of Reactivity | f

At power levels higher than about 10 kw of heat, nonuniform nuclear
heating of the core .oceurs, and produces_spatial‘distributions_of.temperaff
ture in the graphite and fuel salt_ charac_terietic of the power level., The
reactivity change, relative to a"fixed uniform temperature level,iis no -

| 1onger simply related to a single physically measurable temperature (or
 even the average of several measured temperatures) in the circulating
',syetem. Rather, the reactivity is a cumulative effect of the entire
'temperature field in the core.: Thie temperature-distribution reactivity

 

| Interaction effects, 1. e.'effectsrof the temperature change on the
total rod worth, were estimated from theoretical considerations to. be

- qui‘te sma.ll

 
 

10

effect, or steady-state powercoefficient of reactivity,* is somewhat - o
difficult to estimate reliebly for the MSRE, because it requires-accurate
knowledge of the locel heat deposition and temperature distributions in - .
the graphite and salt and the contribution of these local effects to the
neutron reaction rates. An approximate way of treating this problem in- -
volves the use of & "nuclear average temperature,” as described. in Ref. k.
In this method, the local temperature changes are multiplied by a weight-
ing (imporfance) function which measures’theirreffect on the net reac- -
tivity, then integrated over the reactor core. 'Even if we assume that the -
tempereture distributions in the fuel and core graphite can be calculated
aecfirately, one should note that the weighting procedure described in ..
Ref, 4 may be intrinsically in error, when applied to a small reactor core
such as the MSRE. Here the principal temperature reactivity effects arise
from chenges in the neutron leakage. Although non-uniform‘temperature‘e

- changes induce expansion in fuel salt and graphite which do affect the re-
ectivity according to the weighting procedure described above, they also
influence the thermal neutron spectrum in & more complex, non—lbtal'f‘

manner.

“ture was measured during the approach to power, by”helding the reactor S
outlet temperature at a preset value with the gervo controller, and measur-
ing the control-rod response to the chenge in steady-state power level.
Since the reactivity response to the change in temperature distribution is
essentially instantanecus, this effect can be separated from the slower
power-dependent reactivity effects such as xenon-135 and ssmarium-149. The

 

The temperature distributions in fuel and graphite are determined by
the total power level and the mode of temperature level control (the reactor
outlet fuel temperature is servo-controlled in the MSRE). Since the power
level is an input variable to the on-line computer, it is convenient to
relete the reactivity effect directly to the power level.

**To the suthors' knowledge, the theoretical problem of neutron
thermalization in a moderator with a ‘non-{iniform temperature field has
not yet been completely resolved. _

The power coefficient of reactivity for a fixed reactor outlet tempera-

-\

,
oo ™

 

 

L
 

N

11

 measured power'coefficient was +0.001% reactivity per Mw compared to a
.. calculated value of ~O.OOT%sVZThe'observed coefficient correspondsitosa ,

difference of 22°F between the nuclear average temperature of the graphite
and that of the fuel at 7.2 Mw; the calculsted temperature difference was
32°F. This‘sensitivity of_the power coeffieient to changes in core tempera-
tures results from the fact*that'it;represents s small difference between

two larger values (the positive reactivity effect of the fuel average
‘temperature and the negative effect of the graphite average temperature).

As with the other terms in Eq..l in which experimental results can be
applied directly, we emphasize that the measured power coefficient is
used in the overall reactivity ‘balance equation

_Samarium Poisoning

The direct flssion production-decay schemes for the high-cross-section
isotopes 1495m and 15lSm are shovn in Figure 2. The numerical values of
effective removal constants due to neutron absorption, d ¢, given in

| TFigure 2, are normalized to l Mv and corrected for the time the fuel spends
" in the part of the circulating loop external to the reactor core. In

principle, the chains shown in Figure 2 should be connected by neutron
absorption in lS.QSm;.:other-indirect routes for the production of 14°Sm.can
also be considered.5 However, for the relatively low neutron.flux and

fraction of uranium.burnup‘engendered.in the MSRE, these corrections can be.

neglected. For periodic calculation with the ER-340 on-line computer, the
‘differential equations describing ‘the production and decay schemes in
'Figure 2 were’ converted to- finite difference form. The. form of the equations
”used for computation in. both decay chains are: .

N (t + At) N (ti)(l -2 At) + ClP(t ) At . - (2)

(t + At) =N (ti) [1 cra-d;P(t- ) At]'+i‘NP-(ti) .;x Atf - - (3) |

”P(t)+P(t‘+At) R

 

where N(t ) is the atomic’ concentration of the. isotOpe in the fuel salt
at time t At 1s the time interval between calculations of the .concentra-
tions, and P is the average power level during this time interval, The

 
 

 

 

 

12

| 151 Chain -

149 Chain | |
Yield: O. uu%

Yield: 1.13%

  
  

pnl4® _Pmlsl__
»(0.0130 hre’) x(o 02u8 hrs-l)
Sm149-¥——h-5m15° - - " - Sm151_-____p3m152
g0 (8.56 x 10~ hrs™® Mw"1) . <'__o ¢ (6. 305"6 ‘hrs~1 Mw-l)
 Semartwm-l9 SamarinmflSI
' Figure 2 _ L
‘First Order Decey Schemes for Production
of

Samarium Poisons in the MSRE

coefficient C, is the product of the direct fission: yield and the average.
fission rate per unit volume of fuel salt, normelized to 1 Mw. MNumerical
velues of Cy for the MSRE are 6.32 x 10! and 2.46 x 101 atomS”pechm?
salt per Mw for the samarium-lh9 and samnrium 151 chains, respectively.
The conversion of the samarium concentrations to a,reactivity effect

requires knowledge of their average cross sections for absorption in the -
- MOSRE neutron spectrum and the reactivity coefficient-for-a:nnit-absorber

of this type, uniformly distributed in the MSRE fuel salt. - In the MSRE
spectrum, all but & small fraction of the absorptions in the samarium

~ 1sotopes take place below about 0.9 ev. At 1200°F, the peak of the thermal
spectrum is at approximately 0.09 ev, and 0. 876 wss chosen -as the effective
cutoff for the thermal energy group, for reasons of convenience in theo-
retical computations. The ebsorption cross sections, averaged over the
neutron energy spectrum below and above this cutoff energy can be esti-

~ mated with a fair degree of reliebility with presently available computer -

codes. Effective absorption cross sections can then be obtained which give

 

 
-L-}(’" »’

W0

s

n

13 .

the correct resction rete when multiplied by the thermal-group flux.
Theoretical calculations of the: coefficients ‘which convert the samarium
abSOrption rates to reactivity effects must also be employed since a

direct measurement of the- reactivity-change.due.to ‘s known amount of = .-

thermal ebsorber, uniformly distributed,in the MSRE fuel salt, was not

obtainedtfremIthe_zero—power’experimentsd It may be noted, however, that

considerable indirect support of'the theoretlical reactivity coefficient

for thermal absorption-iswgiven;by~the.close comparison between the
measured and calculated #>5U concentration coefficients of reactivity,

mentioned earlier in this report.

Xenon-135 Poisoning

s Early estimates of the magnitude of xenon—l35 poisoning were based

. upon the assumption that, at equilibrium, a relatively large fraction of
~ the xenon produced in the reactor would diffuse into the pores of the

graphite moderator and undergo radioactive decay and neutron absorption
there. Continuous removal of some of the xenon from the. fuel salt would
be accomplished by circulation of a small bypass stream of salt through
the sprey ring in the fuel-pump tank, which contacts the salt with &
stream of helium gas. Estimates of the efficiency of removal of fission

gases by this stripping apparatus, and also of the expected.mass transfer

of xenon to the graphite pores, were based on experiments performed prior
to the nmuclear operation of the MSRE.6 Although it was recognized that
the presence of any circulating voids (undissolved helium gas) would
drastically affect the xenon behavior, this aSpect of the prdblem was ‘_
first. neglected because there was no evidence that circulating v01ds ___’
would be encountered in the operation of the reactor , , '
During the zero-power operation, several tests were performed to

'evaluate the response of the system reactivity to changes in. overpressure._,

 

"1

Most of the suggestions and ground work: to prov1de an interpretation -

,of the xenon behavior in the MBRE are due to R. J. Kedl of the Engineering
| DeVElopment Growp. o ,

 

 
14

In these tests the system.prESsureiwaslslowly increased hy-ahoutflotpsi
and then rapidly reduced to the normel value. - If circuleting voids had
been present, their expansion when the pressure was reduced would have -

expelled some salt from the core and reduced thefnuclear_reactivity; In

'addition, the gas expansion in the entire loop would have raised the salt
level in the fuel-pump tenk. There was no evidence of undissolved gas in

the tests performed with the normal salt level in the pump tank. However,

vhen the salt level was reduced to an abnormelly low value, the seme ex-
periments did indicate 1 to 2 volume percent of undissolved gas;* We con-
cluded from these tests that circulating voids would not he a factor in
the xenon poisoning during normel power operation. o

Soon after power operation of the reactor was started, it became
apparent that the magnitude of the xenon-l35 poisoning was much smaller
then hsd been predicted on the basis of the sbove considerations. At

this point the attempts at on-line calculation of the xenon poisoning were'

'.suspended and the reactivity-balance results were used . to measure ‘the

actual xenon poisoning. Exasmination of the steady-state results showed ; |

thet the low poison level could not be accounted for'with,reasonable .

paremeter values within the assumption of no circuleting voids. Inad-

dition, the system response to small pressure changes now indicated a
small circulating void fraction at normal salt levels in the pump tank.
Another set of pressure-release tests was ‘then performed which showed
significant pressure effects at normel conditions.t If all of the observed
effects were attributed to circulating volds (es was done initially),

volumetric void fraction of 1 to 2% vas indicated. waever, ‘the pressure-

release tests do not necessarily indicate the presence of this amount of
circulating voids prior to the pressure release, they only indicate that
they are present afterwards. That is the observed response could be ex-
-plained by a stagnant void of fixed volume from which expanding gas could f
escape to the circulating stream,when the system pressure 1s reduced. Such
. & void could be anywhere in the loop so long as its volume‘is-unaffected

by pressure (e.g. volds into which saltfcannot»penetrate'hecause'of'Surfacel

tension). Further analysis of the pressure-release tests showed that most

 

 
Ay (“ »n

.ok

9 (w

.15.

 of the -excess gas that was in ‘the. 1bop*after=the pressure release was: .

removed very repidly; gas- stripping efficiencies of ‘80 to 100% were -

"calculated

‘In view of the new evidence for- circulating voids, the steady-state
xenon equations were modified to include bubbles and were reevaluated.
As expected, the steady- state xenon poisoning was quite sensitive to both -
the volumetric vold fraction and the bubble-stripping efficiency. However,
it was found that the steady-state xenon poisoning as a function of- reactor

ipower could be: described by a variety of combinations of void fraction and

bubble- stripping,efficiency, Therefore,_the equations were rewritten to

;include'the‘time‘dependence;which would permit a comparison of calculated

and:observed;transient“135Xe;poiSQning,effects (as‘determined,by the

_change In the criticel position of the regulating rod during the 48 hours

following & change in -the steady-state reactor power level). The purpose

| _vas“to,attempt 8 separationrofjthose;parameterTeffects that‘could,not be
‘separated in the steady-state'correlations.- The mathematical model used

_,to calculate the time behavior of the 17SXe poisoning is described in
References T-and 8. - In the present section we will give. only a quali-

tative description of ‘the main aspectsnand assumptions in the model.

- Further refinements of the model: for the xenon bebavior may also be re-
'quired‘ianuture'operations These'refinementsshould;not‘affect.the-major
- “eonclusions: ahout the overall reactivity behavior. '

~ In the model: chosen, we have’ assumed that 81l the iodine-l35 produced

from fission remsins. in circulation with the salt. After decay to xenon-135,-

the. xenon.migrates to the accessible pores of the graphite at +the boundaries
of the fuel channels and also to mimite helium bubbles distributed through

~ the circulating salt stream...An effective mass-transfer: coefficient was .
- used to describe the transfer of xenon frcm solution in. the circulating
‘,salt to the interface between the 1iquid and ‘the graphite pores at the

: channel boundaries.. Equilibrium Henry s-law coefficients were used for

‘the mass transfer of xenon between the 1iquid phase at. the interfece and
the. gas ‘phase in the graphite pores. The rumerical value used for the
'mass—transfer coefficient between the eirculating salt and the graphite ;f

L

 
 

16

-were based on krypton-injection experiments with flush salt éirculating in-
the fuel loop, performed prior to nuclear operation of the MSRE.® - ‘

Similer considerations were assumed to apply to the mass transfer of
‘xenon from 1iquid solution to the gas bubbles, The coefficient of mass
transfer from the liquid to & small gas bubble, of ‘the order of 0:010 in. -
-in dismeter, moving through the main part of & fuel chennel, was estimated
from theoreticel mass-transfer ‘correlations.® The equilibrium 135%e
poisoning wes shown to be relatively insensitive to the bubble dismeter and
mase-transfer coefficient, over a reasonable range of fincertainty\for these
parameters. ) ' | ST e fA . L _' ,,7 AP

Different efficiencies of removal: by the external stripping apparatus
of xenonfdissolved,in-the salt and that contained in the gas bubbles were .
provided for in the computational model., 'The efficiency of removalv |
(fraction of xenon remofied'per'unit circulated throughithelspray rihg)coftf
xenon dissolved in the salt fias estimated to be between-lo~and*15%,ibased'=
on some early mock-up experiments to évaluate-the'performance'of’the:xenofi ’
- removal apparatus. - - ' '

The conversion of the calculated 13SXe concentrations in salt, gas
fiubbles,,and graphite pores to the corresponding\reactivity poisoning
effect follows from considerations similar to those described in'the,prewf‘
ceding section for the semarium isotopes. -Here, however, there is‘qne-rrt
special feature which should be accounted for which is not present inthe
case of semarium., This is the non-uniformity of the spatisl distribution i
of the 175xe 1h'the‘graphite pores. In the graphite region, the 1°%Xe -
tends to assume & "dished" shape, governed by the bufnout ofwthefxenon‘in-'
the neutron flux. The concentration is minimum near the~centérof.the'ré-.
actor and maximum near the boundaries of the graphite region. . This:in--
fluences the het reactivityjeffect, since these regions aésume diffefent"'
importances in determining reactivity changes. ' The calculation of this .

"shape correction" factor is described in Reference 7. ' oL T

A computational study-based on the theoretical model'described.fibove
vas first performed "off line", with the aid' of an IERM 7090-progmm. These
theoretical calculations:weré-compared_withthe‘fiataflloggéd_bythé=BRF3h0,>'
The apparent transient 1°“Xe poisoning was determined by subtracting all

 
Ay ( %

»

-other'knownjpower-dependent'reactivity effects from the reactivity change

"pQWer,level. This off-line analysis was the most efficient method of making
‘& first-round analysisiof the l1"’1"7’:){‘63'beh,avior because the many Other‘usage
‘requirements of the data lOgger limit us to a relatively simple "point" .
ikinetic model for on-line computations, end also because a wide parameter

.. to show the effect of this parameter on the xenon poisoning.. A.51ngle |
'f'bubble-stripping efficiency (eb) of 10% vas used for this figure. This';
| ,relatively 1low efficiency is equal to- the efficiency estimated for the
_sstripping of xenon dissolved in the salt 1t was used as & first approxi- -
hmation because at the time there was ‘no basis for assuming a higher value
for the bubbles. The effectiveness of the circulating gas . in reducing the.
_Cpoison level is due to the coflbined effects of the large overall gas- 1iquid
T surface area for mass transfer to the bubbles and of the 1arge xenon-
| storage capability of the bubbles (because of the extreme insolubility of
_xenon in molten salt) Thus, the bubbles compete effectively with the
' graphite for removal of xenon from the liquid and xenon in the circulating

17

¢

represented by movement of ‘the regulating rod after a step.change in the

‘study can best,be performed on & larger machine.
In Figures 3 through 13, -we have compared some of the transient reac-

 

_tivity curves obtained from this analysis with some experimental transients,

in the chronological order in which they were obtained. In each of these

: figures, the solid curves represent the- calculated behavior and-the plotted

points show the observed response from reactivity-balance results. A |
measurement of the l34Xe/135X2e ratio in a semple of the reactor offgas

| taken.at Te2 Mw with the xenon in steady state gave an independent value
'for the magnitude of the 135XE poisoning which agreed well with the
| reactivity-balance results. At this date, only a few relatively clean
‘experimental transients corresponding to step changes in power level (for

which the T090 program was devised) have been obtained. -However, several

characteristics of the 135XE behavior are indicated from these curves.

_These will be discussed by considering the figures in order.

Figure 3 shows the calculated. and observed xenon transients for a

step increase in reactor power from zero to T. 2 Mw. The calculations

(solid curves) were made for a variety of circulating void fractions (a%)
 

 

 

N E C L - c i ORNL-DWG 67-1077
1 - 1T 1T | [ ap, VOLUME PERCENT

12
. & EXPERIMENTAL DATA, OBTAINED
T u - DURING MSRE RUN NO. 7

-

10
09
08
o7
06
05
04

" REACTIVITY MAGNITUDE (% B3k/K)

03

0.2 100

Ot

r

 

0 : _ .
0 2 4 6 8 10 12 14 16IB2022242628303234363840424446.7

TIME AFTER INCREASE IN REACTOR POWER LEVEL {hn

Fig. 3. Effect of Volume of Circulating Gas on Transient Buildup of
135%e Reactivity. Step. increase in power level from 0 to ’T 2 Mw; bubble- -
stripping eff:.clency, 10%, MSRE Run No. T. ,
n (' "

FR

i)

n

19

fluid is a less effective polson than that in the graphite because about

~ two-thirds of the fluld is outside the core at any instant, The plotted
points represent the observed 135%e reactivity transient at the beginning
of Reactor Run No. 7 (July 1, 1966). The data indicate that the low
‘apparent xenon poisoning-at-steady state could be'explained by a large

. void frection (between 0.5 and 1.0 vol%) and & low bubble-stripping
‘efficlency. However, the transient buildup is not closely fitted by these

parameter -values, _
"In Figure L4, the curves 4indicate the calculated effect of increasing

fthe bubble-stripping efficiency for a fixed, relatively small (0.1 vol%)
, circulating void’fraction.A The plotted ‘points are for the same ‘reactor
' xenon transient shown in Fig. 3. A comparison of Figs. 3 and 4 shows
- that the steady-state xenon poisoning is: described as well by a low void

fraction with a high bubble stripping efficiency as it is by & high void
fraction with a low stripping efficiency. However, the shape of the &
transient is described much more closely by the parameter values in Fig. k.
Figures 5 and 6 show. the calculated and observed transient buildup
of +35xe poisoning after a step increase in power from zero to 5 TIMw in

~ Run No. 8 (October, 1966) The>ranges of values of a% and Eb used in
~ these calculations are the same as those used for Figs 3 and h Again,

‘the shape of the observed transient is matched more closely by the calcu-

lations which assume & low void fraction and a high bubble- stripping
efficiency. Thus, it appears that the initial assumption of & low . stripping
efficiency for the bubbles was incorrect - The higher stripping efficiency
not only fits the xenon transients better, it is also consistent with the

:_ rates of excess ges removal observed in the pressure-release experiments.
'fSince the latter experiments do not define the void fraction unanbiguously,

_ ;-the low void fraction that must be associated with 8 high stripping

* __efficiency is also in agreement with all the data.

- Figures 7 and 8 show the calculated and dbserved 155Xb re&ctivity :

transients for & power reduction from 5. 7 Mv to zero, with the 135xe
- *
'initially at equilibrium. Comparison of the results in this case provides

{

 

The reactor was made subcritical ‘before the complete xenon transient

'could be recorded.

 

 
 

ORNL-DWG 67-1078

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o7 ' ' ; \
_ P ——— T . BUBBLE STRIPPING
* oommomsrern Nzl L [* errciency e
06 — - ‘ - . et . ®
' ' A ] b

< 05 1~ : —T /f -
2 1 1 B ] / :
2 , I B 11t a0
sl Lo L

04 - < , [ /
E : L~
8 | T ,
3 03 , // L~ _ 50-
l-é- /. / /// Cefo fo |O.
< - : / o al e : '
w 02 o=
@ - . . ./—-" - ! .

//' LT o (
| I
01 ///’,r::: ‘
o Vi

 

o

N

4 6 B8 10 12 14 16 18 20 22 24 26 28303234363840424446'
' TIME AFTER INCREASE iIN REACTOR POWER LEVEL (hr)

Fig. L. Effect of Bubble-Stripping Efflciency on Trans:tent Bulldup of
135Xe Reactivity. Step increase in power level from O to T.2 Mw; volume
percent circuleting bubbles, 0.10; MSRE Run No. T. e e
u( "

£

N

21

ORNL-DOWG 67-1079

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TIME AFTER: INCREASE IN REACTOR POWER LEVEL (hr)

L ey ' e » - - { ap, VOLUME PERCENT
Ce -e EXPERIMENTAL DATA, OBTAINED - ‘ : -
DURING MSRE RUN'NC. 8 , _ _ R CiRCULATING BUBBLES
05 : -
X 1 L~
2 04 —t— ' _ : ,;{‘ 4
- . : ‘ =
r 1 /// B | - 050
.Eoz b '. ’ / o /l : 'k _-——". .‘..‘.. ® . .
W . / / - // ' '._._._LO""‘.
a1 AA /f_:fi—.{. -
' _ ///./"'. e T
// oo ® !
19
LA |
0 2 4 6 -8 710 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46

Flg. 5. Effect of Volume of Circulating Gas on Tran31ent ‘Buildup of
135Xe Reactivity. Step increase in power level from 0 to 5 1 Mw' bubble-

strlpplng eff:l.ciency, 10%, MSRE Run No. 8.

 
 

 

 

ORNL-DWG 67-1080

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.6 ey '
‘| o EXPERIMENTAL DATA, OBTAINED | ; <n'BUBB‘-E STRIPPING
: DURING MSRE RUN f} 0. 8 . E"""'C'ENCY (%)
| 140

0s fre
r o k_ / ]
= b L~ ,
© I/- -] 1 - '
e 04 20 +— : )
3 ‘ | LA
E - ]
> /1 =TT - ’
- / y et : 1 . ,

302 ' _ , - ‘ : e tee® ¢ Le 100}t _

W . // ' 1 e ® ¢ : T ' - -

01 //,// /4{:. . . 1 . 1 _ N

L
o.
z,‘/fo"
o -]

 

| | | ]
0 2 4 6 B 10 12 14 16 18 20 22 24 26 28 30 32 3 36 38 40 42 44 46
: TIME AFTER INCREASE IN REACTOR POWER LEVEL {hr)
Fig. 6. Effect of Bubble-Strlpplng Efficiency on Transient. Buildup of .
135¢e Reactivity. .Step increase in power level from 0 to 5.7 Mw, volume -
- percent circulating bubbles, 0.10; MSRE Run No. 8. -
7y

u"

0 . "
06 [ -‘\-\\\ _
% os | PN L L , ‘
& -\\ .|~ | ® EXPERIMENTAL DATA, OSTAINED
o 1N |" | OURING MSRE RUN NO. 8
g 04 1 _ 1
& . T TN
% 03 ~L LI
& | el L N \N \
¥ 0,2 : L S_o-o .“ ‘ S - »\\ .
T ~e | | T b T™NL | ay, VOLUME PERCENT
| \\\ TN | TSLCIRCULATING BUBBLES
o1 Pl 1™ - o100
; —t %
Tl | T~———025
. Pre—.. .
050
11

23

ORNL-DWG 67-1081

 

r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 ‘ i ;
. 0 2 4 6 8 10 12 14 6 18 .20 22 24 26 28 30 32 34 36 38 40 42 44 46
' : - TIME AFTER DECREASE IN REACTOR POWER LEVEL (hr) . '

Fig. T, Effect of Volume of Circula.ting Gas on. Tra.nsient ‘Decay of
I35](e Reactivity. Step decrease in power level from 5. 7 Mw to 0, bubble-

“ strlpplng efficiency, 10%, MSRE Run No..:B. -

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N
\
ORNL-DWG 67-1082
07 -
06 ///- e \
Zos—— \\\ - —— rh —1-
K 4
& . \ - o EXPERIMENTAL DATA, OBTAINED
5 /w—'—-\ . \ | DURING MSRE RUN NO. 8
04 .
e \\\ . \\\ ;
: 03 — : »
E L ® e e ® o o . - \\ \\ ) \
, g 02 / ry o _ S \ \‘\ -
) oo ,
R " - , BUBBLE STRIPPING
\\\\ \\\\ *5* TEFFICIENCY %)
' T 20
[ — —— "
- ) . L\\\ 'gg
o L 1t

0O 2 4 6 8 100 12 ¥4 16 18 20 22242628303234363840424446
TIME AFTER DECREASE IN REACTOR POWER LEVEL (hr}

. Fig. 8. Effect of Bubble-Stripping Efficlency on Transient Decay of
135¥e Reactivity. Step decrease in power level from 5.7 Mw to 0; volume
percent circulating bubbles, 0.10; MSRE Run No. 8. S o ~
" (‘ »

(34

o ‘.25

about;the'same‘information about‘the bubble parameters'as the earlier
"xenon“buildupftransients. ~The calculated curves also reveal an lmportant

‘characteristic_of the transient xenon behavior which is due to variations

in the overall xenon distribution that result from the cholce of values for

o, end €. If the circulating void fraction is low, most of the poisoning
effect 1s due to xenon in the graphite and only & small amount of xenon is
in the circulating fluid. Xenon that is produced in the fluid from iodine
decay continues to migrate to the graphite for a period of time after the

',power has been reduced. This produces a shutdown peak in the-xenon

poisoning. Eventually, the stripping process. reduces the xenon concen-
tration Iin the fluid so that some of the xenon in the graphite can escape
and be stripped out. This results in a more rapid decrease in xenon

'poisoning than simple radioactive decay. As the circulating void fraction

is -increased, & 1arger fraction ‘of the xenon inventory (or poisoning) 1s
associated with the bubbles andrthere 1s less xenon migration,to ‘the
graphite. In this case the shutdown peak tends to disappear. This effect

_makes the shape of the shutdown transients more sensitive to-changes in

the values assumed for the bubble parameters and thus- facilitates the
process of fitting the observed data to the calculations.

For this same limited decay transient Figure 9 shows the effeet of
the bubble- stripping efficiency with a slightly larger volume fraction of
circulating gas bubbles (0 15 vol%) Although ‘the data for this particular

transient are- somewhat scattered the combined results from Figures 3

through 9 suggest that ab and eb might be bracketed between 0 1 and O 15

~ vol%, and 50 t0 100% resPectively.,?'

A second ”Xe stripping out-decay transient Observed during

- Run No. 9 (November, 1966), following reduction . in the power level from
7.4 Mw to zero, is plotted in Figures 10 snd 11. Again, the approximate
;h_ranges given above for. a% and e are in agreement with the experimental

fobservations of the shape of the transient and the data show clearly the
,small xenon peak expected for these parameter values.

| Finally; in Figures 12 and 13, we show the most recent shutdown transi-

_ent obtained at the termination of Run No. 10 (January 1l, 1967). In this

case, the apparent 135ye resctivity transient was recorded for more than
 

 

26 : ' ' , | r

ORNL-DWG 67-1083

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.67
05 .
g I
3 ™\
£04 S ' = —— —
TN [ |||
— : URI 8 . , o
= .. [T N .
> 4 S g a0, . 7 B ‘ \ \\ .
— ) )
=02 208 oo ™~ N\ |
E T~ \\ : [~ P b BUBBLE STRIPPING }
: T~ T ~~] T~ EFFICIENCY (%)
o1 ‘---..,,_:\__-_ ~—~— \ 10
' e —— 20
0 - 1 fOO 1

 

0O 2 4 66 8 10 12 14 16 1820222426 28303234363840424446
TIME AFTER DECREASE IN REACTOR POWER LEVEL (nn)

Fig. 9. Effect of Bubble-Stripping Efficiency on Transient Deca.y of
135%e Reactivity. Step decrease in power level fram 5.7 Mw to 0; volume
~ percent circulating bubbles, 0.15; MSRE Run No. 8. . :
 

»‘.- ( "

ORNL-DWG 67-1084

 

08

 

 

VT TN

 

o

 

 

 

 

 

 

 

 

 

N
1/
7

o EXPERIMENTAL DATA, OBTAINED
DURING MSRE RUN NO. 9

N

 

 

 

 

L8]

 

 

 

 

REACTIVITY MAGNITUDE (% BK/K)'
o |
H
/f'

 

N\

/

/
A~

R 5
.
[/

[ A

- .\\ N ¢, BUBBLE smu:pme

o1 - e e ~ "~ L \\\ 20

; et o o - . 50

| L1

_'07246810121416!82022242628303234363840424446
L ~TIME AFTER DECREASE IN REACTOR POWER LEVEL (hr)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 10 Effect of Bubble—Stnppmg Eff1c1ency on Transient Decay of
135Xe Reactivity. BStep decrease in power level from 7. h Mw to 0; volume
percent cuculating bubbles, O 10 MSRE Run No. 9. :

 
 

 

28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

07
) ?;
_ | .
T \\\ :
= N
.o \ ) & EXPERIMENTAL DATA, OBTNNED
9:‘: DURING MSRE RUN NO. 9 :
W , ANNE
o TS
§ ‘ N \
§oa '\\ AN —
S e N \
2 e~® '
202 /-___ | ® ~l_ ] - .
. _ [ . ' ' '
P — . m , BUBBLE STRIPPING
: ‘ \-._‘:'\\ ‘ \\ \\ fb EFFICIENCY (%)
01 1 ™ ~ 10
‘\::\N"‘-—-—._ 20
T 130
L1 1.y

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 :
0 2 4 6 8 10 12 14 6 18 20 22 24 26 28 30 32 34 36 3840424446
’ TIME AFTER DECREASE IN REACTOR POWER LEVEL (he)

Fig. 11, Effect of Bubble~-Stripping Efflciency on Transient Deca.y of -
135%e Reactivity. Step decrease in power level from T.4 Mw to 0, volume
percent circulating bubbles, 0.15; MSRE Run No. 9. IR _

" ORNL-DWG 67-1085’

- ar
29 -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

b
3
. ORNL-DWG 67-1086
08 | _
; / \
- 07 " -
06 —— \
. r 3 NG \ | ® EXPERIMENTAL DATA,OBTA!NED
8 1IN L DURING MSRE RUN NO.10
% 04 . 4L N A
: % N A B I N \\ 1
= 03 |~ - - — h 1
- g2 e0 g g N 1 1 \ 1
) w e '_'.'uo...\_‘\}\ \\ \\ o
02—+ 20 e b IS L S ¢, BUBBLE STRIPPING
e A7 je .,N\\ S NG| N EFFICIENCY (%)
1 * o I T~ . \\\ 10
. e ¢ 5 \ ™~ i - ) .
. S o g T — \
01 — - oo '\i""-i... - 20
: ' . 4 '~;~_r_.-1~\ 50
- . e ¥~ 100
o A L f - | 11 |
0O 2 4 6 '8 10 12714 7167182022 24 26 28 30 32 34 36 38 40 42 44 a6
L R } TIME, AFTER DECREASE N REACTOR POWER LEVEL (hr)-
o Fig. 12 Effect of Bubble—Stripping Efficiency on 'I‘ransmnt Decay of
135Reactiv1ty Step decrease. in power level from 7 4 Mw to 05 volume per-
_ cent circulating bubbles, 0. 10 MSRE Run No. 10 '

1re

 
 

 

30

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

REACTIVITY MAGNITUDE (% 8K/}

 

 

 

 

 

 

o7 { \ - - ' , L - ORNL-DWG e7-1087 -
os A — NG _ , —
, : N | @ EXPERIMENTAL DATA, OBTAINED -
N : 1 \ | |]. DURING MSRE RUN NO.10 ,
cobat =L L LN w
. s .'.""-‘.. . . \ . ) ‘
0.2 1 “}}ug.‘ \c\' '\c\ ‘ : _
T ~~ . |, BUBBLE STRIPPING
L , \"'\-.'.__ ‘\i$ ; \\\\'\ o "EFFICIENCY {o%) .
_ 1 . \\. . . - \ .
0Ot ‘ . — i-'.t.-.- P —— 170 .
- - - R "’--‘(_..h-‘-._.‘-.“ \ 20

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

| - | a ‘ """-o '

0 2 4 6 B8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46
' TIME AFTER DECREASE IN REACTOR POWER LEVEL (hr) '

 

- Fig. 13. Effect of Bubble-Stripplng Efficiency on Transient Decay
of !35fe Reactivity. Step decrease in power level from 7.4 Mw to 03 .
volmne percent circulating bubbles, 0.15; MSRE Run No..lO. S
n

[3]

31

Lo hours after'the reduction in'power level. These results indicate
strongly that the tentative conclusions reached from the earlier compari-
sons. are essentially correct |

- Although substantial progress has been made in interpreting the xenon
behavior in the MSRE, the experimental date which have thus far been accumu-
lated’for the transient behavior of the 135xe poisoning are as yet insuf- -
ficient to provide a rigid test. of our model for analysis. As one example,
1t should be noted that, if gas bubbles are continuously being ingested
into the main circulating stream as the evidence indicates, the volume of
gas in circulation is probably not. constant but rather is a slowly varying

'aquantity depending on the -level of the 1iquid in the fuel-pump tank and
~the transfer rate of salt to the overflow tank. This dependence is as

yet not well understood, and future operation is expected to shed further
light in this area. - ' o

We should also mention that least-squares methods can be employed to
determine the unknown paraemeters in the theoretical model for the 125Xe

'.behaviorwhich.providefa'closestjfittolthe experimental transients. - How-

ever, these methods contain several pitfalls (primerily relsting to the .

runiqueness,'andh'h_enceato”the_;in;cerpr_etation of the reSults)mwhen two or

more parameters in;the,differentialfequations describing.thegprocess are
to be determined;simultaneouslyfi ,Theirrsuccess,is best_assuredoifeground-
work 1s first'completed by a brOad'parameter'study such .as that summarized
here. We are now.at the point where least-squares techniques will be |

~useful in further refining the conclusions.;.,

Based on the results of the off -1ine analysis with the IEM-7090,

t_approximate equations and . parameters were determined for the:ER-340.

on-line calculation of the 13sxe reactivity effect. Similar to the case-

'-of the samarium poisoning calculation, these are finite difference equa-

"tions, of the form.given below--'

?J.Ii-*’s"(’ti-»- -At)‘.i-f;-'iifis(%g"(,i“--aoméaf(timi )

 
 

32

xfis(t + At) Il’sct ) [ - aaht - asP(t ) At]

+<341135(t1)-fl¢.+~asx§?5(ti)-+ agfi(ti);em - (6)

.xéfis(ti + At) = _xéf’s_(ti) [1 - azat - aefi(ti)ét] + 8gX1?9(t,) .A‘_c. o n

830

HI5(hy + &) = ————— XS(s, + ) o e

8y, + a-12*1’(‘t‘- )
RPNl = FR(e). ()

~ In these equations, I35 is the concentration of iodine-135 in the
circuleting salt, and X*35 iz the concentration of xenon-135, with sub-

seripts s, g, end b representing the components in solution, in the graphite

Jpores, and in the circulating helium bubbles, respectively. The parameters
~ ag through a,, are determined from the analysis described. in-theipreceding
-peges, and depend on the fission yields, radioactive decay constants, mess-
transfer coefficients, bubble: characteristics, and external stripping
‘efficiencies. The factor F is a shape correction factor for the component
of the 1°5Xe polsoning in the graphite. Although this is actually a time
dependent quantity, in the BR-3h0 program we are presently using a constant
value, equal to the correction calculated under equilibrium conditions -
‘(F~08atP 75Mw) | ‘ |
As further experience is accumulated from operation of the MBRE,-'
efforts will be made to refine the analysis summarized in this section.

Y

 
»n

@)

33

| Density EffEcts of Circulating Bubbles on Reactivity

"hi In addition to its indirect influence on the reactivity through re-_

,Aduction of the 135%e poisoning, the entrainment of undissolved helium in B
~the circulating salt also directly affects reactivity by increasing the
;nautron leakage from the reactor core.; This "fuel-salt density coefficient
-of reactivity was estimated earlier as part of the analysis of core physics
‘characteristics summarized in Ref.. b, The value obtained wasr-O 18% reac-

tivity for one volume percent of circulating gas bubbles.
Measurements ‘were made during the zero-power experiments to. evaluate

-the reactivity effect due to fuel circulation. At that time there was no

evidence of circulating voids and the measured reactivity effect was -0.21%,

in good agreement with the calculated decrease due to the loss of delayed
neutrons. This measurement was repeated in October, 1966 after the

analysis of the 135 poisoning had indicated a circulating void fraction
of 0.1 to O. 15 vol% ‘Prior to the start of circulation, the fuel salt had
been stored in a drain tank for ll weeks so it should have been free of

| undissolved gas. The observed reactivity change between no circulation
| and circulation at steady state this time was -0. 23 to -0. 25%, an increase

of 0.02 - O. Oh% If the amount of gas normally in circulation is approxi-

ymately 0. l to 0. 15 vol%, this means that the density-reactivity effect

would be in the renge of -.02 to -.03% reactivity. Although.this result
does not prove the existence of c1rculating voids, it is at least con- o
sistent with the xenon results. | BEcause the actusl amount of gas in cirouf-

“1ation appears to vary somewhat during operation (see also. later section
~ deseribing cperational experience at’ the . MSRE), the magnitude of this'f_
'_wreactivity effect 18 not well enough established to be. included as an /37
'fexplicit term in the BR-3hO on-line reactivity balance calculations. HEncey
;it is. included “in the residual reactivity in the experimental results prerr
j“~sented in the later sections.rwraf

*IsotOpe Burnout Effects

" We have. already'mentioned in an’ earlier section that changes 1n the

iisotqpic concentrations of lithium-6 uranium—23h -236, and -238,

plutonium-239, and nonesaturating fission products, all in the. salt and

 
 

 

 

3L

residual boron-10 in the grephite cen be lumped together &s a single cate-
gory in terms of their effect on the reactivity of the core. Mbst of these
effects manifest themselves as Very slowly developing positive reactivity
changes, dependent on the time-integrated power, or energy generated. The
exceptions are 236y (for which there is & slight increase in concentretion
resulting from radiative capture in 23577) and the buildup of non-saturating
fission-product poisons.a _ .

~ In the MSRE, the 235y consumed per year s operation at 7 5 Mir is |

56'kg, or approximately 5 percent of the. initial fuel cherge. Because L
this represents a relatively 1ow fractional burnup of the fuel and he- |
~ cause each of the reactivity effects mentioned above is a small correction
in the net reactivity'balance, we can make convenient first-order appfbxi-'
- mations in calcubating these effects. For this purpose, we have assumed
that the megnitude and energy spectrum of the neutron flux remain- con-'
stant during operation at a glven power level, and have used calculated
effective cross sections for neutron reactions in this spectrum With
these assumptions, it is a straightforward exercise to obtain the solutions
to the differential equations governing the first-order changes in isotopic
concentrations with exposure to the neutron flux. We will omit description
~of the algehraic details of these calculations. For ‘all isotopes but
boron-10, account has to be taken of the "flux dilution" effect of the
time the fuel spends in the section of the loop that is external to the
core. Thus the celculated volume-average thernal flux® for the entire 7'
fuel loop 15 0.665 x 1012 n/em® sec/Mw; whereas the average thermal flux
over the graphite-moderated region of the core is 2.0 x 1012 n/cm sec[Mw
The boron concentration initially in the MSRE graphite was estimated from -
‘Ref. 9 to be about 0.8 ppm. In the calculation of the boron burnout we
have neglected a correction factor accounting for the spatial dependence
of hurnout in the graphite region, since the total effect is quite small

 

~ Neutron flux below an effective thermal cutoff energy of 0.876 ev.

 
»

1]

35

Table 1 lists the effective cross sections used in these calculations.
The. effective cross sections ’ multiplied by the thermal fluxes » give the

total reaction rates per atom for neutrons of all energies in the MSRE 7

spectrum _ _ ]
- Since-the formulas for the reactivity changes corresponding to each
of the a‘bove ‘terms are. algebraically smilar, it is possible to develop

an approximate formula for a single 'pseudo-isotope to represent the net
reactivity effect of this group in the ER- 340 calculations.- .The equation

we use is-

. K=4p +4A; PT-+-A2e"blF:Ij + Age P2ET A4e'b3'?Tr | ,(10)

- The parameters Ao through Aé, and 'bl through b= in this formula depend

on the cross sections and . initial isotopic concentrations » and are o'b- )

tained from the analysis outlined in this section.

 

 
 

 

 

 

36

_ Table 1
_ Effective Cross Sections and Reactivity Effects
Due to Isotopic Changes(a) ’

 

Effective Approximate

 

. ' . - Cross Section Reactivity
- C in-MSRE Thermal - . -Effect at -
o - Spectrum &t 1200°F 10* Mwhrs
‘Isotope - -~~~ (parns) - (% ak/k)
ea® 457.6 N .017
Boront®) - - 362l 007
B4y o2k .001(®)
236y S 435 . -.003
megle) 2.9 .0k
239py (abs.) - wsu3 e
%Py (v x fission) 2496.7 - © ,051 (net)
Nonsaturating fission products(f) 43.1 (barns/fission) -.005
Total _' - .072

 

( )The reactivity effect of burnup of 2357 is not included in this 1ist,
since this term 1s treated explicitly in Eq. 1.
(b)

Cross section for the reaction SIi (n, o) >H using the initial
611 concentration. |

(C)Netural enrichment boron (19.8% 1°B)

(d)Includes reactivity increment due to both depletion of 23*U and -
production of =5y, a

™~

)Eurnout only. 7
(£)Egtimatea from Ref. 5.

 
o C ?

»n

*)

37
'EXPERTENCE WITH THE ON-LINE CAICUIATION

.Reactivity-balanceCalculations have been performed.for'the.MSRE>since
the start of'reactor"Operation'atisignificant'poWer,/ During-theHVery;early
stages of the operation, manyfoffthe calculations were done menually while
the'computer*program wes being checked. out. Such. calculations'were'feasi-
ble at that time because the terms which depend on integrated pOwEer. were .
negligibly-small. ‘Subsequently ‘the on-line computer was used to execute -
modified reactivitylbalances;to provide data for evaluatingatheixenon-“

vpoisoninggterm.‘hAt-presentpithe_complete-reactivity balance.is calculated

routinelyiby:the{computer"every’s minutes and the'results‘are'uSed‘without

~-further modification during'normal operation.»'waever; it is still neces-

sary to manually - calculate ‘the ‘dilution effects that occur when the fuel -
loop is drained, Since shutdown’ operations may involve a variety of fuel

- and ‘flush-salt transfers,.eachfshutdown must be treated as a specisl case.

;p Power Calculations

The first operation of the MBRE after the zero-power experiments and
hence, the first opportunity to apply the reactivity-balance calculation
occurred in December, 1965, and Jemary - Februsry, 1966, during & series
of low-power experiments., (The intervening period, July - December, 1965,
weg spent in completing those parts of the system that were required for

“power operation.) The reactor was operated at a variety of powers up to

1 Mw and a total of 36.5 Mwhr of fission energy was produced in these. tests.
| During the control-rod calibration, capsules of enriched fuel were .

‘added to the 1oop with the sampler enricher, and at the end of the zero-
lpower experiments, the 235U concentration in the primary loop was about 10%
'f'greater ‘than that in the salt heel which remained in the drain tanks. Thus,
':pwhen the reactor was drained in.July, 1965, a substantial dilution occurred |
_;which had to be accounted for in the reactivity balance ' |

Since the computer program for the on-line calculation was not ready

ffor service during the low-power tests, ‘manual calculations were performed.
FHoweVEr, the analytic expression for control-rod poisoning and the various

reactivity ooefficients that were being incorporated in the computer L
 

 

 

38

-program.vere applied. Since very little integrated POWer wes produced,
the xenon, samarium, burnmup, and other- fission-product terms were
neglected. . “ ' :

‘At low power these calculations: provided a test of those terms in the

' belence that do not depend directly on power operation, i.e. control-rod

‘poisoning, varietions in operating temperature, and changes in 257 con-
centration.‘ They aslso gave some indication of- the inherentwaccuracy_of
the calculation under the simplest conditions.” -These calculations gave a
residual réactivity of +0:01 % 0.01% 8k/k. This residual wes attributed
- tofiuncertainties in the physical inventory in the-systemvand.was'elimi--
nated from subsequent reactivity balances. [That.is,.the reference con-
dition for the remctivity balence was established as the system condition

~ just before the start of power operation. In addition,to‘verrfying-the_ .
”zero-power"nreactivity-balance,:the'calculations:at;lwa gave-an:early -
indication thet the powerrcoefficient of reactivity.was less negative than
had been calculated and that the xenon poisoning would be less than.-we had-
expected. (See also pp 9 - 11 and 13 - 32. ) As a result of these and
leter findings, experiments were performed to evaluate these two terms.,

Intermediate Calculations

Operation of the reactor at powers and for times that produced sig-
nificant fission-product terms ‘began in Apri1, 1966. This operation soon N
showed that the xenon term was inadequately. treated and that part of the
celculation was temporarily deleted from subsequent computations.' The
calculation results from the other texms in the reactivity balance were
‘then used to aid in the development of an adequate representation of ‘the
xenon poisoning o | fl T

" In order to use the reactivity balance to evaluate xenon poisoning, it
was necessary to assume that there were no other unaccounted-for reactivity

effects. This assumption was not completely valid for ‘the early calcu~'

lations because of long-term effects that were neglected but it was valid

for the relatively short times involved in the xenon transients. Since
most of the data for the xenon calculation were developed from the reac- B
tivity transients after the reactor power was raised or lowered

N

 
»

39

(see pp 13 - 32) the.early errors in the long-term reactivity balances
were of ‘1little consequence. ,
Figure 1b shows the results of reactivity-balance calculations without

xenon for ‘all power operation of the reactor between April and July, 1966.

The reactor power 1s shown with each reactivity plot. for reference pur-

_ poses. The reactivity transients associated with the buildup and removal

of xenon due to changes in power are clearly displayed. The apparent .
steady-state xenon poisoning et meximm pover (~ 7.2 Mr) 1s 0.25 to
030%8]:/1:. o R x

"~ The large negative-reactivity transient on June. 18 - 19 was caused

by the development of a large circulating void fraction in the fuel loop.

It was known that if the fuel-salt level - in the pump tank were allowed
to decrease below & given value, the emount qf gas in circulation ‘would
increase significantly. ‘This condition was reached on June 18 and the
accompanying - decrease in average fuel density prodnced ‘the reactivity
decrease, The reactivity recovered rapidly when the normal pump-tank |

 level was restored. and the excess gas was stripped out. The response of

the reactivity balance in this event 1llustrates the sensitivity of this

method for detecting minor anomalies under otherwise normal circumstances.
The reactivity ‘belances calculated,for the period shown in Figure_lh

were not completely'corrécted“for_longfterm isotopic change effects or

for flush-salt dilution. This'is'illustrated.by therapparent increase

in the residual reaetivity at zero-power when there . was no xenon present,

(the especially the results on April ll, May 9, June 13, and.July 1 and’

121-23 ) Corrections for these factors were subsequently applied to the "

zero~power balances to evaluate as accurately as possible the 1ong-term

,drift in the residual reactivity.fifi L

Comnlete Calculations

The complete reactivity balance calculation, including all known

,effects, vas first applied to the period of reactor operation which began
:ain October, 1966 Figure 15 shows the power history and residual reac-

tivity results on & day-to-day basis for the next three runs (the reac-

‘tivity scale in Figure 15 is expanded from that of the preceding figure).
 

 

  
   
 

   
   
    

 

Lo
ORNL-DWG €6-9082
. 0 nE . o s _ Ao oo -~ i ol
2702 | [NeTREACTIVITY
-0.4 V 11
& ~ [POWER
: 4 '
0 iy b Pt ] R4
SN, 13 1 1w 18 20 235 25 27 29 © N 13 % 7
B | APRIL,1966 © . MAY, 1966 .
o
-o.z | |NET REACTIVITY
-04 '
-0.6

 
 

 
 

9. 21 23 25 27 29 13 15 7 19 2t 23 25 27
‘ MAY, 1966 - JUNE, 1966 '

R " INET REACTIVITY
q 0
:®
-0.2
8
o POWER
$ 4 ,
0

T 3 5 7 8 n B B w7 1© @21 25
| R JULY, 1966 o
" 'Fig. 1b. Results of Modified’ Reactivity Balences in MSRE.

[l
   

. ORNL—DWG 65-41944R

 

REACTIVITY

«—-DRAIN

   

8 ‘10 92 - 44 6 - 8 20 .22 24 2. 28 30 . .1 9 T 3. 45 4T 49
TS A | .. OCTOBER,#966 - o ~  NOVEMBER, 1966

REACTIVITY

TH

%COMPUTER OUT
" OF SERVICE .

 

Mw

  

DECEMBER 1966 ' o 1 . V JfiNUAHY, ‘967 ‘

 

F1g. 15 Results of Ccmplete Reacuvity Balances in MSRE.

A
ka2

During steady-state operation the results show only minor veriations in
the residusl resctivity. However, in October and November there is still
some indication of a disagreement between the calculated-and:actuai_xenon
polsoning, both in the absolute magnitude of the term and in the trensient
behavior. The results for December, 1966 and,Jannary, 1967 show better
- transient agreement but still some difference in the magnitude of the
xenon term. o - - D

The larger spikes in residual reactivity'can a1l be accounted forrby
‘abnormel reactor conditions which are not covered in the reactivity balance.
For example, the spikes on October 10 are associated with special experi-
mente during which gas bubbles were circulating with the salt Fuel-salt
circu;ation was Interrupted for 2-1/2 hours on October 16 and nosxenon
stripping occurred. When circulation'and'pouer operation were resumed,
the actualnenon-poison level was higher than that calculated in the re-
‘activity balance which assumed continuing circulation and stripping while
the power was low. On October 23, the salt level in the pump tank wes at
an sbnormally high level for a brief period. The xenon stripping was'nnch
less effective in this condition and the xenon—poison 1eve1 rapidly built
up to a higher value. When & more normal salt level was. restored, the
xenon.poisoning returned to the normal value.

The perturbations in residual reactivity during the November operation
‘resulted from failure of the calculation to adequately describe the xenon
transients. - During this run it was necessary to reduce the power on
several occasions because of conditions imposed by the reactor offgas
system. In each case the observed xenon behavior was about.thersame,'indif
cating a longer time constant for xenon stripping than wes caleulated in
the model, This disparity in the time constants produced the cyclic be-
havior that was observed. o | _

Considerable difficulty We.s encountered in the operation of the on-
line computer during the last period of operation shown in Figure 15. As
& result, ~substantial gaps exist in the complete reactivity-balance re-
sults. However, the gvailable results are in good agreement with the ex-
pected behsavior. .Again, the spikes on December 23 and 24 and.January 12
reflect abnormel reactor operations which resulted in circulating voids.

¥

 
 

43

~ The smaller variations (see, for. example, the period from December 30 to
Jamiary 5) appear to be related to.variations in the. fuel-system over-

~ pressure, They mey reflect changespin the circulating voidtfraction or

| '_variations'in the net xenon-stripping'efficiency."Additional-detailed
analyses will be required-to identiiy'the.cause of these small'variations.

'—iong-Term Residual Reactivitv

| The long- term drift in residual reactivity can best be seen in the
calculation.results,whererthere is no xenon present. In order to make
,thisicomparison,‘representatireresults.of this kind have been-converted -
to & common basis using current'values for all coefficients, The major
corrections that were applied to earlier results were to compensate for
:'long-term isotopic-change effects that had been neglected and for flush-'
salt dilution effects. Each time the fuel loop is drained a small heel
of the salt that vas circulating'remains'in the loop. This salt then
:‘mixes with the material that is- next introduced into the loop. . When the
reactor is shut down for maintenance the fuel loop is normally rinsed |
with flush salt to remove as much residual radiocactivity as possible
‘Then, when the loop is refilled with fuel salt, the remaining flush-salt
heel produces_a dilution of the fuel.,  Some additional intermixing occurs
because a common fill-and-drain'line_is used for the two salts. The extent
of the salt intermixing was determined from the amountlof uranium that
- has appeared in the otherwise—barren flush salt. Chemical analyses of

© the flush salt indicated the amount of fuel salt that was picked up by the

flush salt in various- operations._ We then assumed that a similar volume
of flush salt is added to the. fuel.h The net result of & flush-salt £111
- and drain followed by a fuel—salt fill is to reduce the system reactivity
*f7by sbout 0.05%. CELEL T | |
fl The corrected reactivity-balance results at zero power with no xenon
‘present are shown as’ a function of integrated power' in Figure 16. It

'rfshould.be noted that the reactivity scale is greatly expanded and that the

*average residual reactivity is only about +0 05 Bk/k. There’ appears to |

_ have been a positive shift of ebout +0. ol to +0.05% 6k/k early tn the o

0peration with insignificant changes occurring subsequently. -

 
 

0.08

 

" 'ORNL-DWG €7-1076 = -

 

004 — 8

oo|

 

 

RESIDUAL REACTIVITY (% Sk/K)
o

-0.04

 

 

 

 

 

 

 

 

 

_

 

 

. -0.08.
o o

’ . " INTEGRATED POWER (Mw-hr)

| Tig. 16. long-Term Drift in'Residual Reactivity in MSRE.

)

2000 4000 6000 8000 10000 . 12000 14000 16000 ____,'3000;.

e

 
 

uqzra,

L

L5

Through the end of Run No. 10 (January, 1967) the reactor had pro-

”;duced 16,450 Mw-hrs, equivalent to 95 days' operation at maximum power
: and substantial changes had occurred in many of the reactivity-balance

terms ~ Table 2 shows typical values fdr the various terms in the reac-
tivity balence at the start of power_operation and at the end of Run 10.
The vealues given represent zeroépouer operationywith no xenon present to
emphasize the long-term effects.” The estinated_accuracies'of the various
terms are included.in the table for later'consideration'(see'pp'hT i8).

' This table shows. clearly the current value of the residual reactivity

of 0.05%.

INTERPRETATION OF RESULTS

lgrevious'Reports of Results

Therresults‘presented in this;report represent'our;currentleyaluation'”
of the reactivity behavior of the MSRE during the first year of power

" operation. In the course of this year the accumulation of data and experi-

ence has resulted in a number of changes in the calculation of various

terms as well as in the interpretation of the results.' Because of ‘the
-'interest in the performance of the MSRE and the value of the reactivity

balance in assessing that performance, intermediate results have bean

| reported from time to time (see ‘especially Reference 10) even though 1t

was recognized that further analysis was required for an accurate inter-

- pretetion. Some of “these results suggested the possibility that the posi-
'.'tive residual reactivity was gradually 1ncreasing. This apparent increase
ruwas due to an inadequate treatment of 1ong term changes in minor salt o
”fconstituents and to a misinterpretation of conflicting data on the circula—”

ting void fraction." . 3 _ : f o
. Tt is to be expected that additional modifications will be made in

:iour treatment of the reactivity balance as more operating experience is
eccumulated. - However, we’ feel that -any future refinements will have small
:'effects and that the current evaluation is reasonably accurate.

 
 

 

L6

Table 2

" Values of Reactivity-Balance Terms in MSRE
- /Y,r-..a.t‘ - .
zZero Power

 

Value (% 8k/k)

 

 

8 Start of | | ~ Estimated
: -Effect | Power After =  Change ~ Uncertainty
Term . Described Operation . 16,450 Mwhr (% BX/k) (% dk/k) .
KROD  Control-rod N - - e
KU235 Excess =57 ’ e | |
| concentration 1.785  +1.355°  -0.430 = +0.011
KTEMP Reector outlet : | - o
~ temperature -0.073 -0.073 0 ==
- KPOW  Temperature . c B S
-~ - aistribution 0 0 0o .-
. -polsoning 0 -0.534 =0.534% . 0.027-
KXE  Xenon - . 4 g : . el e i
- ~ poisoning 0 | 0T - -
KB Circulating - L
~ bubbles e ' - - ==
KFP  Isotope - | S D -
: burnout 0o - +0.116 +0.116 = *0.006

KNET - Residual 0 +0.047 +0.047  t0.04

 

‘a. Change from reference condition.
b. Includes dilution by flush salt.
~c. Value at 7.4 Mv is +0.007% Bk/k.
d. Value at 7.k Mv 1s -0.27% 8k/k.
e. Not currently included.

 
M(‘ l

Yy

IUtilitv of Residual Reactivitv

| nuclear stability both at steady power and during transients, a comparison

k7

‘The residuel reactivity as ‘determined from the resctivity balance o
cennot - be used by itself as an absqlute indicator of the reactor per- ,
formance. Because of the experimental nature of the MSRE and the variety
of unknowns associated with the reactivity behavior, perticularly in re-
gard to xenon poisoning, it was necessary to use the reactor behavior as
a tool in developing the reactivity balance. During this development it _'
was necessary to assume that no anomalous reactivity effects were present.
This assumption was supported by a variety of other observations the

 

of predicted and directly observable nuclear characteristics, chemical

analyses of fuel-salt samples, and examination of in-core irradiation and
corrosion specimens.. Even after its development the reactivity balance_
must ‘be used in congunction with these other observations to insure that |

- no neglected but otherwise normal reactivity-effect is interpreted as

an anomaly.

| The - reactivity balance is potentially one of the most sensitive indi~
cators of changing conditions in a system like the MSRE. waever, there
are certainly 1imitations in both the precision and absolute accuracy of

“such calculations. At steady reactor conditions (constant temperature,

pressure, and power) the variation in consecutive reactivity balances is
only about O. 01% ok/k. This is associated primarily with variations in

the temperature and control-rod-position readings from . the computer and,

- ,therefore, probably represents the precision limit of the calculation.
» It is airficult to provide a reliable estimate of the confidence o
"limits of the calculations summarized in. this report Tb a. large extent
| refinements in the analysis to. include effects found to be: significant
.'together with reinterpretations of measurements, have to. be performed
‘cfsequentially as reactor operating data are obtained. The measurements
?of reactivzty effects important in. Operation are often interwoven, 80 _
rthat operational data taken in connection with one particular effect have 5
t’shed further light on earlier measurements pertaining to: other effects.;-r
_~,'This process is expected to continue.
 

48

Beceuse several of the most importent terms in the reactivity-balance; ,
(control-rod'worth, excess 35U, temperature levels) are based on measure-r
ments made during the zero-power nuclear experiments, a rough basis for
‘discussing the accuracy ‘of these terms is provided by those experiments. -
As mentioned in an earlier section, independent measurements of the j“ |
control-rod worth (by means of period-differential worth experiments and ;
rod drop integral worth experiments) were found to he self consistent
~ within 5%. 'Also, the interpretation of other reactivity effects (2350 ,
concentration coefficient overall temperature coefficlent, and delayed-
Vneutron 1osses) based on the rod calibration were within 5% of the calcu~'"
lated values. Thus, reasonable confidence limits are probably . 2.5% on |
 terms for which experimental measurements are available and 5% onfterms |
for which only calculations are available. Application of these 1imits to
the changes in reactivity leéads to an uncertainty of * 0.04% 51:/1: in the
residual reactivity at zero power ‘with no xenon present. (See also_'p
Teble 2, p 46.) | | | | D

The very small uncertainty in the residual reactivity mekes this a
Vivery sensitive monitor of conditions in the MBRE. By comparison, sta-
tistical analysis of the results of chenmical analyses of fuel-salt samples |
gave & change in 25U concentration of -0.025 + 0.013 wt% hetween the |
gtart of power operation and 16,450 Mw-hrs.}' This corresponds to &
reactivity change of -0.36 + 0.18% 8k/k which can be directly compared
with the reactivity-balance value of -0.43 * 0.01% &k/k in Teble 2. - Thus,
while both the reactivity balances and the chemical results indicate normal
behavior the reactivity balances are somewhat more accurate in this par-fl
ticular‘application and_are continuously avaeilable during-reactor operation.

Effects: Not Treated

Several effects have been mentioned which have not been explicitly
included in the reactivity-balance calculations, and for which cognizance
should be taken. These include the production of photoneutrOns-through
(y,n) reactions in the beryllium and neutron-absorptions'in'the.products
which result from reactions that are evaluated. Since only changes in
reactivity relative to the reference condition are cbserved in the ré-
activity belence, one may show by approximate calculation that the magnitude

 
»*

ko

of these effects should have negligible direct effect on the reactivity

"balance in the MSRE.

Of potentially greater significance among the effects known to be
present but not accounted for are (1) the slight changes in the structural

J_configuration of the graphite stringers and salt channels due to neutron

irradiation demage to the graphite, ‘and (2) the cumulative effects of

irrediation on the control-rod worth (through burmup of the gadolinium).

'_'Both these effects ‘should appear as slow changes in the residual reactivity.

, Radiation damage is expected to cause the graphite to shrink, thereby |
.reducing slightly the axial dimensions of the core, increasing the ef-"“‘h

fective graphite density; and causing some bowing of the stringers due to
the radial gradient in the neutron flux. Tt is @ifficult to provide a

'_precise estimate of the change in core reactivity associated with this

effect, but’ a reported estimate which should be on the conservative side
(larger than the actusl magnitude) 1s about +.07% 8k/k per Mv yr.12
Althoughfthis is in the range'whiCh might be detected in the residual re- -

'activity; ‘no consistent,’ slowly increasing change of this magnitude has

been observed 1n the reactivity.
~ In the second case, above, ‘rough calculations supported by comparative
bservations in the reactor, have indicated that the effect of burnout of

" the gadolinium on rod reactivity 18 of negligible significance in the.MSRE

operation to date. waever, corrections for this effect should properly

B be accounted for as operation:continues into aisubstantialifraction-of the
core life. A thorough enalysis of this effect 15 planned in the. Immediate

In the MSRE operating authorization, the USAEC recommended that .

; "allowable limits on reactivity anomalies should be established and docu- .
';mented.before critical tests begin and 'should be adhered to during all

operations.fl13 This wes, . and is beingb done. The qurating limits on the

. MSRE include this one.—rWAt no time during critical operation of--the; re-
"ac‘tor will 'the reactivity anomaly be aiiowed to exceed O. 5% Bk /i, "4
 

50

The 1imit of 0.5% was set in consideration of the consequences of &
very pessimistic hypothetical incident involving separated uranium. _It Y
was postulated that uranium separated from the circulating fuel by some
unspecified process and collected in the lower head of the reactor _'
vessel. . Then something caused part of the uranium. to be. resuspended and
sucked up through the central channels in the core in a single blob,_;(The
velocity in 22 channels near the center is 2.0 ft/sec,.almostpthreeitimes
“the velocity in the 940 channels covering the main body of the core.)
| The computations were done as follows. 1s The shape of the reactivity
transient due to movement of & blob of uranium up through a central chennel
was computed. - Then the transients in power, temperature, and core pressure
wvere computed for various amounts of added reactivity and different initial
power levels. No account was teken of rod scram, only the shutdown pro-
vided.hy-the negative temperature coefficient of reactivity.; The computed
power excursions were brief, producing sharp but momentary increases in _
the temperature of the fuel in the core (but little change in graphite
~ temperature) and pressure surges in the core due to fuel-salt expansion.
A tolerable excursion (one which mould not be expected to cause damage) )
was defined as one in which the'pressure surge was less than_BO-psig'and
the peak fuel temperature was less than 1800°F. The limit was reached by
incidents in which the reactivity addition peaked at 0.7% 8k/k. éhee__ p
amount. of excess uranium that would give this reactivity was. computed to
-be 0.8 kg (neglecting self-shielding in the blob, which would increase the
amount of uranium required). |
The next step was to decide what fraction of a uranium deposit might
reasonably be pictured as becoming detached and passing through the: core

 es postulated, In HRE-2 (an aqueous fluid-fuel reactor where fuel separa-

tion could and did occur), deposits could be dispersed by movement of the
‘loose core-inlet screens in the turbulent flow, or by steam formetion, and

 

}

There is no known mechanism.by which such a separation could oceur

under the conditions maintained in the MSRE.

—F

 

 
.‘(‘ ‘

»y '

ah

51

_the dispersed material was soluble. - Even under these conditions the
| -largest sudden recovery'of uranium was less than 0.1 of the existing de-
‘posits. In the MSRE, on the other hand, deposits of uranium.as V02 should
- be quite stadble so the probability of resuSpension of any significant
ufraction should be quite small. Therefore, we. considered that-an assump-'

tion of sudden resuspens1on of 10 percent of the separated uranium was

quite comservative. < - - e

. With the foregoing pessimistic assumptions, ‘one computes that the
separation of 8 kg of uranium is the maximum smount ‘tolerable. If this

" much were to separate from the circulating fuel and collect in a region of

low nuclear Amportance, the reactivity would decrease by O. 5% Bk/k. This
was set then as the maximum allowable reactivity anomaly |

' Conclusions -

Several conclusions can be drawn from the experience with the

_greactivity-balance calculation during the first year of power operation of
the MSRE. The calculation has provided an invaluable tool for evaluating
,the perfbrmance of the reactor system, particularly in connection with the
i xenon-poisoning problem.' The results have been accurate and precise enough
-to permit a detailed analysis and evaluation of mechanisms which would
rotherwise have been largely indeterminate in the reactor. -In addition,
rthey'have provided the operating staff with a real-time monitor of the

' condition of the reactor system.

fi Possibly the most. important conclusion. is that the reactivity balance

| has shown, within very narrow confidence 1imits, no anomalous reactivity
; behavior during this first year of power 0peration. The long-term change

that has occurred is: lower than the allowable anomaly by a factor of 10 -

:and there have been no unexplained short term deviations. This experience
- shows with considerable confidence that the reactor has performed as.ex-
'pected in all respects that could affect the nuclear reactivity
 

 

 

12;_

13.

1k,

‘15.

52
REFERENCES

-.Ps N. Heubenreich et al., Summary of MSEE Zero Power Physrlcs
Experiments, USAEC Report ORNL-TM (in preparation)

. Oak Ridge National Laboratory, MSRP Semiann._ Progr..Repte Aug. 31, 1966

' USAEC Report om\rL-ho37, PP 88-9h - ~

Oek Ridge National Isboratory, MSRP Semiann. Progr. Rept. Feb. 28 1966
USAEC Report 0BNL-3936 D 82—87.

P. N. Haubenreich et al. s MSRE Design and Operations Report Pa.rt III,
Nuclear Analysis, USAEC Report ORNL-TM-T30, Oak Ridge National
Isborastory, Feb. U4, 1964, pp 41-L8.° |

L. L. Bennett, Recommended Fission Product Chains for'*U’sfé’ih ’Reactoi»*
Evaluetion Studies, USAEC Report. ORNL-TM-1658, Oak Ridge Nationsl
Ieboratory, Sept. 26, 1966. |

R J. Kedl and A. Houtzeel, Dé:‘reloment' of a Model fdr-Com:pfi'bing" .

-135Xe Migration in the MSRE, USAEC Report'ORNL-hO69 (in preparation).

‘Oek Ridge National Isboratory, MSRP Semiann. Progr. Rept. Feb. 28, 1966,
USAEC Report ORNI~3936, pp 87-92. ' L - -

‘Osk Ridge National Isboratory, MSRP Semiann. Progr. Rept. Aug. 31, 1966,
USAEC Report ORNI-L037, pp 13-21. U

Oak Ridge Nationsl Ieboratory, MSRP Semiamn. Progr. Rept;
‘July 31, 1964, USAEC Report ORNL-3708, pp 373-389. ~

Osk Ridge Netional Isboratory, MSRP Semiann, Progr. Rept. Aug. 31, 1966,
VSAEC Report ORNI~4037, pp 10-13. R

‘Ozk Ridge National Iaboratory, ‘MSRP Semiann. Progr. Rept. ,Fe_'bv.' 28, 1967,
(in preparation). S -

P. N. Haubenreich et al., MSRE Design and Operations Report, Part ITI,
Nuclear Analysis, USAEC Report ORNI.-']M-'T3O , Oak Ridge’ National
Isboratory, Feb. 4, 1964, pp 183-18k. | |

-Ietter from H. M. Roth to A. M. Weinberg, May 19, 1965,
Sub,ject- MSRE Operating Authorization.

S. E. Beall and R. H. Guymon, Opera.ting Safety Limits for. the MSRE, |
USAEC Report ORNL-TM-T33 (1st revision) Aug. 3, 1965.

S. E. Peall et al., MSRE ‘Des:l.gn and Operations Report, Part V,
Reactor Safety Analysie Report, USAEC Report ORNL~TM-T7 32, Aug. , 1964,
PP 21’+ 216.

 
C

T

| 1k,

. -* ‘.

O OO~ A\ =W N

—
o
-

11.

12,

15.
16.
. 17-

18.
o 19. -
- 20.
21,

22,

=
oh.
25..

26.

o7,

28.
- 29.

30,

© 3L,
32,

- 33,
3k,
35.

36,

37. -
T f 390
ko,

hl-hs

& .
~

R.

. Internal Distribution

- MSRP Direétof's Office'_
) m. 325, 920’4‘"

K. Adams - |
M. Adamson

G. Alexender =

H. Anderson
F. Apple

. Mo Baker L

P. Barthold

F, Bauman
E. Beall .

. Bender . -

L. Bennétt
S. Bettis:

F. Blankenship ) |

Blumberg

- B. Briggs .
. Go'; BOh]mann '_; -
:J. Borkowskl -

S. Carlsmith

. Cantor

L. Carter

- I. Cathers

D. Cheverton
L. Compere
H. Cook

T. Corbin
B. Cottrell :.

L. Crowley

G. Davis

o J. Ditto
G. Donnelly

A, Doss

23

- W. Craven, Jr, : :'. L

‘L. Culler =
M. Dale o T

E. Eaison - .

- P. Epler

- 80.
81,
82,
83.
8k,

86.

87.

890

wwwbzm?ongBes

]

ORNL-TM-1796

Ergen ~
Ferguson
Fowler

Fraas -
Friedman
Frye, Jr.
Gabbard

., Gallaher
- Gift

Grimes

. Grindell

Guymon
Harley
Haubenreich

. Hebert

Heddleson
Herndon:
. Hudson

Jung

R.

Jds-

dJds

R,
W,
. T,
.S,

J.

" T.

W,

,A.

B.

I.
N.
G.
“E.
H.
D.
E.

E‘

C.V

F.

'Ko

E.

Kasten
Kedl
Kelly
Kennedy
Kerlin
Kerr
Kirslis
Knowles
Krakoviak
Krewson
Lane
Iindauer
Iundin- - -
Lyon -

‘MacPherson
‘MacPherson
"Marable
‘Martin
Mathews
McCoy

McCurdy
McDuffie

‘McGlothlan

McNeese

 
 

 

%.
91.
,92.
930
9k.

95.

»

8.

99 .
100.
101.
102 .
103.
10k,
105.

106,
107.
108. )

109.
-110.
111.
112,

113.

- 11k,
. - 115.

54

Internal Distribution

 

(continued)
A. S. Meyer | 116. R. C.'Steffy
A. J. Miller 117. H. H. Stone -
‘R. L. Moore 118. J. R. Tallackson
"E. A. Nephew 119. R. E, Thoma - _
P. Patriarca 120, W. E, Thomas
H. R. Payne 121, “ M. L. Tobias
A. M. Perry , 122, G. M. Tolson
~ H. B. Piper 123, D. B. Trauger -
- P. H, Pitksnen . 124k, W. C. Ulrich
C. M. Podeweltz 125, D. R. Vondy
B. E. Prince 126. A. M, Weinberg
J. L. Redford 127, J. R. Weir, Jr.
M. Richardson 128, K. W. West
R. C. Robertson 129. M. E. Whatley
H. C. Roller 130. J. C. White
H. C. Savage 131. G. D. Whitman
A, V. Savolainen 132, H. D. Wills
D. Scott © 133. F. G. Welfare
H. E. Seagren 134. J. V. Wilson
J. H. Shaffer 135. L. V. Wilson :
M. J. Skinner 136-137. Central Research Li'brary
"A. N. Smith 138-139. Document Reference Section
0. L. Smith 140-142. TIaboratory Records
P, G. Smith 143, Isboratory Records - RC
W. F. Spencer 1L4h, Nuclear Safety Information Center
I. Spiewak oo ' o | -
. ‘f :
External Distribution
145-146. Reactor Division, ORO
| 147, A. Giambusso, AEC-Washington
148, C. L. Matthews, AEC-ORO o
149, T, W. McIntosh, AEC, Washington
150. H. M. Roth, AEC-ORO '
151. W. L. Smalley, AEC-0RO

152-167.

Division of Technical Information Extension

L‘C) v