ORNL-TM-2997.txt

 

o

 

OAK RIDGE NATIONAL LABORATORY
operated by

UNION CARBIDE CORPORATION
NUCLEAR DIVISION ‘ L
for the

U.S. ATOMIC ENERGY COMMISSION
ORNL- TM- 2997

COPY NO. -
DATE - April, 1970

EXPERIMENTAL DYNAMIC ANALYSIS OF THE MSRE WITH 2337 FUEL

R. C. Steffy, Jr. N\S\“

ABSTRACT

Tests were performed on the Molten-Salt Reacter Experiment to deter-
mine the system time response to step changes in reactivity, the neutron-
flux-to-reactivity frequency response, and the outlet-temperature-to-
power frequency response. The results of each of these were found to agree
favorably with theoretical predictions. '

The time response tests were performed with the reactor operating
at 1, 5, and 8 MW and substantiated the theoretical predictions that fol-
lowing a reactivity perturbation the system would return to its original
power level more rapidly at higher power levels than at lower power levels
and was load-following at all significant power levels., A noisy flux sig-
nal (caused by circulating voids) hampered detailed comparison of the
experimental results and theoretical predictions.

Neutron flux-to-reactivity frequency-response measurements were per-
formed using periodic, pseudorandom binary and ternary sequences. This.
type of test effectively prevented much of the random noise contamination
of the neutron flux from entering the final analyses and gave results
which contained little scatter. The results were in good agreement with
the theoretical predictions and verified that for the MSRE, the degree of
stability increases with power level,

Outlet-temperature-to-power frequency-response measurements were com-
pared with similar measurements made during operation with the 235U fuel
and verified that the basic thermal properties of the reactor system were
essentially the same as expected.

Keywords: MSKRE, fused salts, reactors, operation, reactivity, testing,
time response, frequency response, stability, pseudorandom binary sequences,
Dseudorandom ternary sequences.

NOTICE This document contains information of a preliminary nature
and was prepared primarily for internal use at the Ogk Ridge National

Laboratory. It is subject to revision or correction and therefore does
not represent a final report.

DISTRIBUTION OF THIS DOCUMENT IS UNLIMITED
 

 

-—— LEGAL NOTICE R

 

This report was prepared as an account of Government sponscred 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 respect to the accuracy,
completensss, or usefulness of the information contained in this report, or that the use of
any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; or

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

As used in the above, '‘person acting on behalf of the Commission’ includes eny employee or

contractor of the Commission, or employese of such contractor, to the extent that such employee

or contracter 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 contracter.

 
CONTENTS

ABSTRACT . v v v v v v v v e e e e e e e e e e e e e e e e e e
INTRODUCTION . . v v v v v v e e e e e e e e e e e e e e e e e e
TRANSIENT RESPONSE. . . . v v+ v v v v v v e e e v e e e e e e e
FREQUENCY RESPONSE. . . v + v v+ v v v v v o 4 o o 0 o v o v« « .12
Neutron Flux to Reactivity . . . . . . . . . . .« . . . . . .12
Testing Procedure . . . . . « « v v v v 0 v e w e e e .. 12

Analysis Programs . . . . . . « « v 4 o+ 4 e e 4 4+ 4 o+ 13
Discussion. . . v v v v 4 e e e e e e e e e e e e e e .. L

Outlet Temperature to Power. . . . . . . « « « « « « « « « . . 22
CONCIUSION. . . v v v v v vt e v e e e e e e e e e e e e e e e 25
LIST OF REFERENCES . « v v v v v v o v e e e e e e e e e e e e e w26

LEGAL NOTICE —————

This report was prepared as an accourt of Government sponsored work. Netther the Uniied
States, nor the Commission, nor any person acting on behalf of the Commission:

A. Makes any warranty or representation, expressed or implied, with respect to the aceu-
racy, completeness, or usefulness of the {uformation contained in this report, or that the use
of any information, apparatus, method, or process disclosed in this report may not infringe
privately ovned vights; or

B. Assumes any llabliltles witk respect to the use of, or fur 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 em-
ployee or contracter of the Commissian, or employee of such contractor, to the extent that
such employee or contracinr of the Commissiun, or employee of such contractor prepares.
disseminates, or provides access to, any informativn puranant to his employment or contract
with the Commisgion, or his employment with such contractor,

 

 

 

g e DTG
e BUTLION OF ThbsD oo e :
EXPERIMENTAL DYNAMIC ANALYSIS OF THE MSRE WITH 23 FUEL

R. C. Steffy, Jr.

INTRODUCTION

Several reports and articles (References 1 - 6) relating either to
the theoretical or actual (or both) dynamic response of the Molten Salt
Reactor Experiment have been published. However, none of these has re-
ported in & concise form the dynamic response of the U-233 fueled MSEE.
Reference 4 contains much of the frequency-response information reported
herein, but it is presented in a lengthy context which is primarily con-
cerned with comparing testing signals and techniques. The purpose of
this report is to give a brief description of the observed dynamic re-
sponse of the U-233 fueled MSRE, compare it with the theoretical and sug-
gest possible reasons for differences when applicable, but to eschew any

lengthy description of the testing technigues.

TRANSTIENT RESPONSE

A common method of describing the dynamic response of a stable sys-
tem is to display the system response to a step change in an input vari-
able. For a nuclear reactor, reactivity is usually the perturbed para-
meter. This type description (i.e. description in the time domain) has
the advantage of an intuitive appeal to people since we deal directly
with time in day-to-day living. However, analysis of a system response
in the time domain does have some disadvantages. Notably, if the system
output of interest is contaminated with a large noise component, the part
of the output resulting from a step input may be undiscernible from the
part caused by the noise. The reason for making this point is the large
difference in the neutron noise level between the Z°°U fuel loading and
2330 fuel loading of the MSRE. (The increase in noise level was due to
a concomitant increase in circulating void fraction and was not an in-

trinsic function of the fissile isotope.) An example of the uncontrolled
neutron flux during high-power operation for each fuel is shown in Fig. 1,
and the relationship between the flux noise and void fraction is readily
observable., The void fraction estimates which are labeled on Fig. 1 were
achieved by varying the fuel pump speed; however, the fuel pump was
operated at full speed (~ 1180 rpm) for all of the dynamics tests reported
here.

During the initial approach to power with the #°7U fuel, time re-
sponses of the neutron flux to a step change in reactivity were recorded
and are shown in Figures 2,* 3, and 4 for the reactor at 1, 5, and 8 MWQ**
respectively. Also shown in these figures are the theoretical predictions
for step reactivity changes of the same magnitudes, The theoretical cal-
culations were performed using the mathematical model and method described
in Reference 2, The noisy flux signal hinders a comparison of the finer
detail of the theoretical and experimental curves, but the noise was low
enough that some features may be compared., In general, the theoretical
and the experimental curves are in good agreement,

For the 1-MW case (Figure 2), the initial flux peak was slightly
nigher thar the theory predicted, then it oscillated below the initial
level and later increased again with a second peak occurring after about
360 sec. The theoretical curves agree that the change in power should
nave returned to a positive indication at this time but indicate that it
should not have teen as large in magnitude as the observed behavior. The
extent to which noise contaminatiorn forced the positive indication is not
KNCwn,

The noise contamination in the 5-MW case (Fig. 3) makes it diffi-

cult to compare directly the experimental and theoretical results. They

 

3
The original plot of the response at 1 MW was made by a different
machine than the other two plots. This accounts for the difference in
general appearance of the plots.

*¥
Full power was taken as 8.0 MW during the data analysis and writing
o this repcrt.
Pig. 1.

ORNL-DWG 69-5374R

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

'—; PERCENT RCENT PLRCEN PLRCENT
o 0 50 60 10 0 40 50 40 50
™~
£ ]
£ 1
0 1[
~ {
L) J
= :
— | J;
1 ,
{
PERCENT RCENT PERCFN PERCENT
0 50 60 10 0 40 50 40 50
{180 rpm 160 rpm  1120rpm 1070 rpm
<0.1 vol % 0.6 vol% O0.3vol% OJvol%
i )
235U 233U

RR-8{00 CHART ({percent of 15 Mw)

Sections of Nuclear Power Recorder Chart Contrasting 275U Fuel,
Full Flow and Few Bubbles with 73 Fuel, Varying Flow and Bubble
Fraction. Conditions in each case: T MW, 1210°F, 5 psig,

52 - 56% Fuel Pump Level.
ORNL-DWG 70-2922

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.7 | | | 1
0.6 POWER LEVEL =1 Mw .
Ejl — — — THEORETICAL
' j \;L EXPERIMENTAL
0.4 ,. . REACTIVITY INSERTED=0.0139 % 34 —
3 \
= 0.3
< I
0.2
0.1 N“_’J-L__‘-
5 S~ —n _iilpfmjfl:“:_m.—
~0.4
0 60 120 180 240 300 360 420

TIME AFTER REACTIVITY INSERTION (sec)

Fig. 2. Response of the Neutron Flux to a Step Change in Reactivity
of 0.0139% Bk/k with the Reactor Initially at 1 MW.
AMw

ORNL-DWG 70-2921

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

08 | | | | |
POWER LEVEL =5 Mw
06 --— THEORETICAL ]
' —— EXPERIMENTAL
REACTIVITY INSERTED=0.0190 % 8k/k
0.4
> /\N"’\ \/\W\/\A /\
O ----- s sm e === v v ==
-0.2
0 60 120 180 240 300 360 420

TIME AFTER REACTIVITY INSERTION (sec)

Fig. 3. Response of the Neutron Flux to a Step Change in Reactivity
of 0.0190% &k/k with the Reactor Initially at 5 MW,
are 1in general agreement, but detailed comparison would be guess-work,
The swells and rolls that occur after about 150 sec are almost surely not
directly related tc the original reactivity input since the system set-
tling time at 5 MW is about 150 sec,

For the reactor operating at 8 MW, the flux response to a reactivity
step of 0.0248% 8k/k is shown in Figure 4. The maximum power level was
reached during the first second after the reactivity input. This rapid
incregse was accompanied by a rapid increase in fuel temperature in the
core, which, coupled with the negative temperature coefficient of reac-
tivity, more than counter-balanced the step reactivity input, so the power
level began to decrease. The temperature of the salt entering the core
was constant during this interval, and when the power had decreased enough
for the reactivity associated with the increased nuclear average tempera-
ture to just cancel the step reactivity input, the power leveled for a
brief time (from ~ 6 to ~ 17 sec after the reactivity input). About 17 sec
after the reactivity increase, the hot fluid generated 1n the initial
power increase completed its circuilt of the loop external to the core, and
the negative temperature coefficient of the salt again reduced the reac-
tivity so that the power level startec down again. At large times the
reactor power returned to its initial level, and the step reactivity in-
put was counter-balanced by an increase in the nuclear average temperature
in the core. For the 5-MW case, a short plateau was probably present
.50, but the noisy signal obscured its presence. At lower powers, how-
ever, the slower system response prevented the reactor from reaching the
peak of its first oscillaticn tefcre the fuel completed one circuit of
the external fuel loop. The plateau therefore did not appear in the 1-MW
case.

An important characteristic of the MSRE dynamic response was that as
the power decreased the reactor tecame both more sluggish (slower respond-
ing) and more oscillatory; that is, at low powers the time required for
oscillations to die out was much larger than at higher powers, and the

fractional amplitude of the oscillations (A power/power) was larger.
11

CRNL-DWG 70-2923

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.4 ] . | T T
POWER LEVEL = 8 Mw
1.2 -—- THEORETICAL
L0 —— EXPERIMENTAL
' REACTIVITY INSERTED = 0.0248% 8k/k
0.8
2 i
= 06 f‘\
<] I
\ W,
0.4 ¥
0.2 .
0 M‘l'
-0.2

0 20 40 60 80 100
TIME AFTER REACTIVITY INSERTION (sec)

Fig. 4. Response of the Neutron Flux to a Step Change in Reactivity
of 0.248% 8k/k with the Reactor Initially at 8 MW.
12

FREQUENCY RESPONSE
Neutron Flux to Reactivity

Most of the effort in experimentally determining the dynamic response
of the MSRE was expended in determining the neutron-flux-to-reactivity
frequency response. One advantage of working in the frequency domain 1is
that a periodic waveform may be continuously imposed on a system input
(e.g. reactivity, through control rod movement) until several periods of
data have been collected. All of the signal power of a periodic signal
is concentrated at harmonic frequencies, and subsequent analysis at a
harmonic frequency very efficiently eliminates most of the noise contami-
nation which is usually dispersed over a wide frequency band. There are
other advantages to wecrking in the frequency domain, but the more noisy
flux signal with the 33 fuel loading makes this a salient advantage.
Several step and pulse tests (aperiodic tests) were also attempted but
these do rot have the signal energy concentrated at particular frequencies
and tke system noise was large enough that the results contained too much

scatter to bhe useful,.

Testing Procedure

 

"he periodic signals used in the frequency-response tests were either
pseudorandonr binary or pseudorandom ternary sequences.? These are par-
ticular series of square wave pulses that were chosen because they evenly
distributed the signal power at the harmonic frequencies over a wide fre-
guency range, which permitted determination cf the freguency response
cver a wide spectrum with conly one test. The frequency range over which
we obtained freguency-response results was from about 0.005 to 0.8 rad/sec,
The lower limit was set by the length of one period of the test pattern
and the high-frequency limit was determined by the time width of the
square wave pulse of shortest duration which the standard equipment would
adequately reproduce. The shortest basic pulse width used in these tests
was 3.0 sec. The frequency range covered by these tests was essentially

the range cver which thermal feedback effecis are important.
13

The on-line computer, a Bunker-Ramo 340, was programmed to generate
the sequences by opening and closing a set of relays. Voltage was fed
through the relays from an analog computer (Electronic Associates, Inc.,
Model TR-10). This voltage was used to determine the mcvement of the con-
trol rods, which were forced either to follow the pseudorandom test pat-
tern themselves or to cause the flux to follow the test pattern.® The
control-rod position and the neutron flux were digitized and reccrded
every 0.25 sec on maghetic tape., The data were retrieved from the tape
and stored on punched cards which could then be processed with the anal-

ysls programs to yleld the frequency-response information.

Analysis Programs

Before discussing each of the programs used to analyze the data, it
is pertinent to note that in some instances the different analysis pro-
grams yielded markedly different results when applied to the same data,.
It is beyond the intent of this report to delve into the possible theo-
retical explanations, but the interested reader may consult Reference 4
for a more complete treatise on the subject.

FOURCO.® This code directly Fourier transformed the time
records., The transformed output (flux) was then divided by the trans-
formed input (rod position) to give the frequency response., This analysis
was usually performed on the full data record, which would contain several
periods of the same waveform, but occasionally was performed on individual
periods of data with the several resulting answers then ensemble gveraged.
This latter method is denoted FOURCO ENGEMBIE on the figures,

CPSD.”?»©® This analysis method utilized a digital simulaticn of
an analog filtering technigue for obtaining cross-power spectral density,
CPSD, functions. This code calculated the power spectrum of the input
signal and the cross-power spectrum of the input and output signals and
divided the cross-power gpectrum by the input power spectrum to obtain
the frequency response at each frequency of analysis. The key feature of
this code is an adjustable filter width about the analysis frequency.

CABS.7 The third calculational procedure was more involved,

The auto-correlation functions of the input and output signals were calcu-

lated and the cross-correlation function of the signals was calculated.
1h

These were then Fcurier transformed to obtain the input, output, and
cross-peower spectra. The input power-spectrum was then divided into the

crcoss-power spectrum to obtain the frequency response.

Discussion

With the fuel stationary, the frequency response of the zero-power
MSRE was essentially the same as that of any stationary-fuel, zero-power,
£33J-fueled reactor. The measured frequency response with the fuel not
circulating is shown in Figure 5. The magnitude ratio, Bn/NO-Sk, is
seen Tc be in general agreement with the theory, but the phase angle is
rot in particularly good agreement. At the higher frequencies for tests
at all power levels, the magnitude ratio and the phase angle were lower
than the theoretical. This is thought to have been caused by the control
rod not adequately following the test pattern yet giving the indication
that it was. The indicators, which are physically located with the drive
assembly, accurately display the action of the rod-drive motors; however,
the flexibility of the control rod makes it doubtful that the tip of the
rod, which 1s about 17 ft from the drive assembly, reprcduces the high
frequency component of the rod-drive movement.

The results of a typical zero-power test with the fuel circulating
are shown in Figure €. The shape of the magnitude ratic curve is in ex-
cellent agreement with the theoretical curve, but the results have been
normalized by multipiying each experimental value by 1.75. The phase
angle data was in better agreement with the theoretical predictions than
wag the case for the non-circulating data, but there is scatter in the
results.

The need to normalize some results and not to normalize others is
2ls0 considered to be caused by poor contrel rod indication.® The nor-
malization was not Power dependent since some data did and some did not
need normalization at each power level, and the normalization factors,
when they were required, were different for different tests.

As we mentioned in the intrecduction, several different testing tech-

niques were used in obtaining the experimental results. An example of
8n

 

15

ORNL -DWG 69-12050

FUEL STATIONARY
POWER LEVEL - ZERO

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g
[ s}
‘o
>

ANALYSIS METHODS

s CPSD

a FOURCO ENSEMBLE

— THEORY

/
o ,””.
® —30 -
© _‘/
o 7 /4’1" "2 fifi"-
7 1 441182 % ‘fi "B
<I / ? a B
T -60 ~
a ”‘/ 4|
AT
| P 4
—90 —
10° 2 5 102 2 5 40! 2 5 40°
FREQUENCY (rad/sec)

Fig. 5. Neutron Flux-to-Reactivity Frequency Response of the

£33j-Fueled MSRE at Zero-Power with Stationary Fuel.
PHASE (deg)

16

ORNL-DWG 69- 12044

10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2
10°
POWER LEVEL -ZERO
. ANALYSIS METHODS
o FOURCO
= CPSD
5 = THEORY
10°
0
-30 _
] g b /r"
o /-g -+ a-a ;?“BSQEB . Bo °
1. 0 o ® 2C
-60 1 {1? | ° 00y2¢
o ‘
z/’+
9

 

 

 

 

 

 

 

 

 

 

03 2 5 1072 2 5 10 2 5 10°
FREQUENCY (rad/sec)

Fig. 6. Neutron Flux-to-Reactivity Frequency Response of the
233j-Fueled MSEE at Zero-Fower with Circulating Fuel.
Ly

results® obtained using a technique* that was unsatisfactory on the MSRE
is shown in Figure 7. The results do not disprove the theoretical pre-
dictions, but they do little toward verifying them either. Certainly,
the results would have done little toward describing the reactor's re-
sponse if the theoretical response were unknown. These data are shown
primarily to display the system response at low, but significant, power.
A satisfactory testing technique** for this reactor was not found until
after the preliminary tesis were completed, and it was not convenient to
return to 1 MW to perform further tests. However, the good agreement be-
tween the experimental results and the theoretical predictions at both
higher and lower powers almost insures that the theoretical curve is
very close to the actual response, hence the 1-MW theoretical curve may
e taken as the actual response, In addition, this figure illustrates
the importance of the testing technique which accounts for the difference
in appearance of the results in Figure 7 and those in Figures 8 and 9.
The scatter in the results shown in Figure 7 is due to inaccuracies in
the indicated control-rod position which were accentuated by the testing
technique,

Typical results from tests which employed the most satisfactory
testing technique are shown in Figures 8 and 9 for the reactor at 5 and
8 MW, respectively. The results are in excellent agreement with the theo-
retical curves except for the slight discrepancy at the higher frequencies.
The dip in the magnitude-ratio curves at ~ 0,25 rad/sec (corresponding to
a loop transient time of ~ 25 sec) results from temperature feedback from
the external loop. During a periodic reactivity perturbation at a fre-

guency of about .25 rad/sec, the fuel in the core during one cycle returned

 

*This was the technique in which the neutron flux was forced to follow
to follow the test pattern. It was necessary for the control rod to move
almost continually during this type test and errors in the indicated control-
rod position caused the unsatisfactory results. The technique is basically
sound and could be well utilized on a system with favorable hardware.

**The technique that gave the most satisfactory results was one in
which the control-rod position was forced to follow the test pattern. The
rod moved to a new position and then remailned stationary for several seconds
until a different pulse was needed. This minimized control-rod movement
and the associated errors.
18

ORNL-DWG 69-12051

POWER LEVEL-1{ Mw
ANALYSIS METHODS

© FOURCO
® CABS
THEORY

 

 

90 j ' Y
-*k ! ‘
101 248
o

AT
o N

 

 

 

 

 

 

 

 

 

 

 

 

W
O
|

 

 

 

 

 

O

 

PHASE (deg)

 

 

 

 

 

!
(N
O

 

 

 

 

 

 

 

-60

 

i !
|

‘H) -120

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-90 - -
0> 2 5 0% 2 5 10 2

FREQUENCY (rad/sec)

Fig. 7. Neutron Flux-to-Reactivity Frequency Response of the
2335-Fueled MSRE at 1 MW. -
19

ORNL—-DWG 69—-12245

10%

 

PHASE (deg)

POWER LEVEL — 5Mw
ANALYSIS METHODS
v FOURCO, CABS
(EACH GAVE SAME RESULTS)
— THEORY

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30
7N
%
o N

) /‘\

%W?% "-n-.....‘\..
-30 Y o 7
-60 ~ W

10~3 102 10" 1
FREQUENCY (rad/sec)

Fig. 8. Neutron Flux-to-Reactivity Frequency Response of the

233j-Fueled MSRE at 5 MW.
20

4 ORNL -DWG 69-12246

POWER LEVEL =8Mw
ANALYSIS METHODS

o FOURCO, CABS, CPSD
(EACH GAVE SAME RESULTS)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- THEORY
x
|
mao
o
s3]
Z
i
o
<
a
0
G
N
ik%%: % %%5 \N\
n
—-30

 

1073 2 5 4072 2 5 107! 2 5 100
FREQUENCY (rad/sec)

Fig. 9. DNeutron Flux-to-Reactivity Frequency Response of the
233-Fueled MSRE at 8 MW.
21

to the core one pericd later and, because of the negative temperature coef-
ficient of reactivity, produced a reactivity feedback effect that partially
canceled the external perturbation, The dip i1s obviously present in the
experimental results as well as in the theoretical curves; however, the
dip in the experimental data 1s not as pronounced as the theory predicts,
Since the magnitude of the dip has been shown® to be a function of the
amount of salt mixing which occurs as the fuel circulates around the loop,
this difference between the experimental and theoretical implies that not
enough mixing was assumed in the theoretical model. Additional work with
the theoretical model has shown that 1f the salt transport in the piping
is represented by a series of 2-sec first-order lags (well-stirred tanks
with mean holdup times of 2 sec) rather than the pure delays that were
assumed in the earlier work, the dip in the experimental and theoretical
responses are in good agreement.

Below about 0.5 rad/sec, the magnitude ratic decreases as the power
is increased. This substantiates the observation drawn from the time
response plots; the degree of stability for the MSRE increases with power
level, The lower magnitude ratic at the higher power levels over the
frequency range in which thermal effects are important says, in effect,
that for the same change in reactivity the fractional power (A power/power)
change will be less at higher power.

The frequency-response curves shown in this document display the
MSRE 's frequency response at several power levels. Of course, several
tests were performed at several different power levels, but in order to
keep the presentation as straightforward as possible, we chose to show
the results from representative tests. Table 1 summarizes the freguency-=
response tests performed with the £33 fuel loading and indicates the
scope of the testing program which included 28 different tests of approxi-
mately one-hour duration each. Other experimental results for the &3]
fuel loading are given in References 4 and 5. Complete results of theo-
retical dynamic analyses are given in References 2, 5, and 6. Note that
some tests were performed shortly after the start of operation with 2°-U
fuel, and others were performed near the end of operation with 27U fuel.
There were no indications that the response of the reactor had changed with

operating time,
22

Table 1

Information Related to Frequency-Response

 

Testing of ©2°U-Fueled MSEE

 

 

 

Integrated No., of
Testing Power Power Tests
Dates (MW-hrs) Level Performed

10/15/68 0 100 W 1
11/7-8/68 0 50 W 6
1/16/69 86 1 MW 3
1/20/69 435 5 MW 1
2/3/69 2,390 8 MW 1
2/17/69 L, 080 5 MW 1
2/20/69 L, L90 8 MW 3
3/11/69 7,220 10 kW 1
L/2L /69 14,000 8 MW 2
5/26/69* 19,500 8 MW 9

 

.X.

These tests were performed for M. R. Buckner and
T. W. Kerlin of the University of Tennessee as part of a
gradugte studles program.

Outlet Temperature to Fower

During the neutron-flux-to-reactivity frequency-response tests which
were conducted at significant power levels, the response of a thermocouple
(TE-100-1A) on the outlet pipe was also recorded. The data records then
included power (or more specifically, neutron flux) and outlet temperature
during a time in which the power was varied in a periodic waveform. Hence,
the outlet-temperature-to-power freguency response could be determined at

the same harmonic frequencies as the neutron-flux-to-reactivity frequency
23

response. The results of this determination could then be compared with
the results of theoretical predictions.®

The outlet-temperature-to-power frequency-response results.from a
test conducted during operation with 25U fuel as well as two tests per-
formed during operation with 277U fuel are shown in Figure 10. The experi-
mental results of all three tests are essentially the same. This should
be expected since the temperature response to a given change in power is
a function of the thermal properties of the system, and these were changed
very little with the change in fissionable material.

Three theoretical magnitude ratio plots are alsc shown in Figure 10.
Curve 1 is the as-calculated curve and curves 2 and 3 are this same curve
multiplied by 0.5 and 0.1, respectively. Normalization of the theoretical
by multiplying by 0.5 forces agreement with the experimental results at
low frequencies and multiplying by 0.l forces agreement at high frequencies.
The reason for the discrepancies between the experimental and theoretical
have not been explained leaving this as an area open for more analysis. It
is of interest to note that in some experimental work!'® performed by
S. J. Ball and T. W. Kerlin in which they attempted to determine the re-
sponse of outlet-temperature-to-inlet-~temperature perturbations, they too
found a larger degree of attenuation than had been theoretically predicted.

The phase angle plots shown in Figure 10 are in good agreement if
the theoretical thermocouple response to a power perturbation is delayed
by 0.7 sec more than was assumed in the original calculation. (A pure
delay gives a phase shift that changes linearly with frequency.) The
theoretical response of the thermocouple was represented by a l-sec pure
delay plus a 5-sec first-order lag. This was based on calculations per-
formed by 8. J. Ball.'! This represents a good estimate, but could be
in error by 0.7 sec for this particular thermocouple depending on its
particular response characteristics and physical contact with the pipe.
Another possible source of error is the estimate of the location of the
thermocouple on the pipe.

The experimentally-measured outlet-temperature-to-power frequency
response verified that the basic thermal properties of the MSRE were es-

sentially unchanged by the change in fuel loading. The disagreement
(8T/3n) (°F/Mw)

PHASE ANGLE (deg)

Fig.

2k

ORNL -DWG 70-3423
100

50

o FROM TEST ON 23%U—FUELED SYSTEM
® FROM TEST ON 233U~FUELED SYSTEM
s FROM TEST ON 233U~FUELED SYSTEM

20

10

0.5 - (D THEORETICAL
@ THEORETICAL x 0.5
® THEORETICAL x 0.1
0.2

 

 

~60 S —

 

 

 

 

%" “ba ™
o \r\\~
'“’%a? N

 

-120

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-180 %q,:

— THEORETICAL \
—— THEORETICAL WITH o

—-240 ADDITIONAL O.70sec TIME LAG

L]

1073 2 5 4072 2 5 107! 2 5 100
FREQUENCY (rad/sec)

 

 

Sl
]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10. Outlet Temperature-to-Power Frequency Response of the
MSRE with the Reactor at 8 MW.
25

between the theoretical and experimental magnitude ratio determinations
makes 1t meaningless to draw any conclusions about the mixing effects in

the circulating system.
CONCIUS ION

The dynamic response of the £3Fj-fueled MSRE was analyzed by three
different methods, each of which had deficiencies but each of which added
information. The transient response of the neutron flux to a step change
in reactivity at various power levels verified that the general response
of the system was as anticipated, but the noisy flux signal made detailed
comparison of the theoretical and experimental results difficult. The
shape of the experimentally-determined neutron flux-to-reactivity fre-
quency-respense curves was in excellent agreement with the theoretical
curves over most of the frequency range which was realizable with the in=-
stalled hardware. There were problems assoclated with finding a test
method which would give good results, and erroneous control rod position
indications necessitated normalization of some experimental results. The
outlet temperature-to-power frequency-response determination did not agree
well with theory but did show that the basic thermal properties of the
MSRE were essentially unchanged by the change from 235 to 27 fuel.

At high powers, the MSRE is a highly damped system. It returns to
its original power level rapidly with no undershoot or wallowing. At
low power levels, the uncontrolled reactor tends to be sluggish and slow
in returning to its original power level. With the reactor at 1 MW, it
was observed that over LOO sec was required for the flux level to stabi-
lize after a step change in reactivity. In summary, the MSRE was stable

at all power levels and the stability increased with power as predicted.
1.

10.

11,

26

LIST OF REFERENCES

S. J. Ball and T, W. Kerlin, Stability Analysis of the Molten-Salt
Reactor Experiment, USAEC Report ORNL-TM-1070, Oak Ridge National
Laboratory, (December 1965).

R. C. Steffy, Jr., and P. J. Wood, Theoretical Dynamic Analysis of
the MSRE with U-233 Fuel, USAEC Report CRNL-TM-25T71, Oak Ridge
National Laboratory (July 1969).

T. W. Kerlin and S. J. Ball, Experimental Dynamic Analysis of the
Molten-Salt Reactor Experiment, USAEC Report ORNL-TM-1647, Oak Ridge
National Laboratory, (October 1966).

R. C. Steffy, Jr., Frequency-Response Testing of the Molten-Salt
Reactor Experiment, USAEC Report ORNL-TM-2823, Oak Ridge National
Laboratory (March 1970).

MSR Program Semiann. Progr. Rept., Feb. 28, 1969, USAEC Report
ORNL-4396, Osk Ridge National Laboratory.

MSR Program Semiann. Progr. Rept., Aug. 31, 1968, USAEC Report
ORNI-L3LL, Oak Ridge National Laboratory, pp. 46 - 52.

S. J. Ball, A Digital Filtering Technigue for Efficient Fourier
Transform Calculations, USAEC Report ORNL-TM-1662, Oak Ridge National
Laboratory, (July 1967).

S. J. Ball, Instrumentatlon and Control Systems Division Annual Pro-
gress Report, September 1, 1965, USAEC Report ORNL-3875, pp. 126-127,
Oak Ridge National Iaboratory (September 1965).

T. W. Kerlin and J. L. Lucius, CABS — A Fortran Computer Program for
Calculating Correlation Functions, Power Spectra, and the Frequency
Response from Experimental Data, USAEC Report ORNL-TM-1663, Oak
Ridge National Laboratory, (September 1966).

MSR Program Semiann. Progr. Rept., Feb. 28, 1966, USAEC Report
ORNL-3936, Osk Ridge National Laboratory.

S. J. Ball, Personal Communication to R. C. Steffy, Jr., July 2L, 1968.
O QO] Ot i N

27

ORNL~TM- 2997

INTERNAL DISTRIBUTION

. Ackermann
Adams
Affel

. Anderson
Baes

Ball

. Bauman
Beall
Bender

S. Bettis
Blumberg
Bohlmann
Borkowski
. Boyd
Briggs

. Buhl
Burke
Cardwell
Clark
Cope, AEC-CRO
Cottrell
Crowley
Culler
Ditto
Eatherly
Engel
Epler
Ferguson
Ferris
Franzreb
Fraas

Fry
Furlong
Gabbard
Glambusso, AEC-Washington
R. Grimes — g, M. Watson
Grindell
Guymon

. Harley
Harms
Haubenreich
Houtzeel

T. L. Hudson

W. H., Jordan

@

B b O 5] 2R

s

o

o

S

Eiplet+2 OO0 QEHYEH=Z o e

o

&

m?i'ZZ'UONZWFUSUFUC—lt‘*HtU'fiEZSWtflFJC—iQ

ZO':E!'.IE}Q

a

m:wjgiwjm =EPEQID OGO EG

o

81.
82.
83.
8L,
85.
86.
87.
88.
89.
g0-9k.
95.

Kasten

. Kedl

Kerlin

Kerr

. Krakoviak

. Kress

. Kryter

1t Laughon, AEC-OSR
Lundin
Lyon
MacPherson
Martin
Matthews,
McCoy
McCurdy
MeGlothlan
McIntosh, AEC-Washington
. Mclain

McNeese

McWherter

Miller

Moore

. Nicholson

Oakes

Perry

Piper

Prince

Ragan

Redford

Rlchardson

. R. Riley, AEC

W. Rosenthal

M. Roth, AEC-ORO

A W, Savolalnen

Dunlap Scott

R. M. Scroggins, AEC

M. Shaw, AEC-Washington

H. Sides

J. 8Bkinner

L. Smalley, AEC-ORO

N. Smith

Spiewak

C. Steffy

D A. Sundberg

T H DTy

>
D -«
=

b*p*?jyflfiif)t*t*cfltflifi S E R OQEHAY Elmtkia QUHAZXN®

AEC-0SR

EZUZQthHbt*HDJbQHmI—:IOtE!mOO':U':UE

Iikir:iig =
28
ORNL-TM-2997

INTERNAL DISTRIBUTION

(continued)

96. J. R. Tallackson

97. R. E. Thoma

93. D. B. Trauger

99. F. N. Watson, AEC-Washington
100. A. M. Weinberg
101. J. R. Weir

1062. M, E. Whatley

103. J. C. White — A, 8. Meyer

10k, G. D. Whitman
105. Gale Young
106-107. Central Research Library (CRL)
108-109. Y-12 Document Reference Section (DRS)
110-112. Laboratory Records Department (IRD)
113. TIaboratory Records Department, Record Copy (LRD-RC)
114 . DNuclear Safety Information Center
115, ORNL Patent Cffice

EXTERNAL DISTRIBUTION

116-130. Division of Technical Information Extension (DIIE)
131. ILazboratory and University Division (ORO)