i

el

X-822

 

DATE:

SUBJECT:

TO:

FROM:

OAK RIDGE NATIONAL LABORATORY | - |
Operated by - - Internal Use ~Only

 

UNION CARBIDE NUCLEAR COMPANY S » )
~Division of Union Carbide Corporation | 0 R N L

= | CENTRAL FILES NUMBER

Post Office Box X | |
Oak Ridge, Tennessee o ' - | 68_5-—11

 

 

 

May 1, 1968 cory No. 4

»Performance of MSRE Nuclear Power COHbTOl System@

(MSRE Test Report 5.2.1)

Distribution

C. H;”@abbard»
ABSTRACT

The nuclear power control systems of the MSRE were evaguated
by observ1ng the steady -state opefatlon of the reactor and by
conducting a series of transient tests, The temperature servo
was found capable of controlling all the transients that were
1ntmoduced However, because of the relatively slow response
and inherent stability of the reactor system, the temperature
servo was found to be relatively inactive during many of the
load change transients, The automatic load control operated as
expected except that the minimum power &vallable to the auto-
matic control was about 2 Mw. 1nstead of 1 Mw as had been planned.
This has not caused a problem in the reactor operation because
the load control has normally been operated in "manual",

NOTICE

This document contains information of a preliminary nature
and was prepared primarily for internal use at the Oak Ridge
National Laboratory. It is subject to revision or correction
and therefore does not represent a final report. The information
is not to be abstracted, reprinted or otherwise given public
dissemination without the approval of the ORNL patent branch,
Legal and Information Control Department.
INTRODUCTION .

CONTENTS

TEMPERATURE SERVO CONTROL OF NUCLEAR_POWER .
oteady-State Performance . .

Normal Setpoint Change . . . v v ¢ « « .« .

- Rapid Load Changes

Temperature Setpoint Error.

Control by Negative Temperature Coefficient

At Reactivity
AUTOMATTC T.OAD CONTROL .
CONCLUSIONS. . . . . .
APPENDIX . . . . . .

O N U = = LQ

13
15
18
21
INTRODUCTTON

Two independent control systems are used to set and control the
‘nuclear power of the MSRE at power levels above 1 Mw. The heat removal
rate is set by the air flow conditions at the radiator. The temperature
servo con'trollerl then adjusts the control rods to match the nuclear
power with the heat removal rate while maintaining the reactor outlet
temperature constant at a preselected value. An automatic load control,Z
which increases or decreases the radiaﬁor heat removal rate, is provided
for the convenience of the operators.

A test program for evaluating the performance of the‘temperature
servo and the automatic load contrcl was outlined in MSRE Test Memo 5.2.1,
The program for evaluating the temperature servo consisted of steady-state
performance at each power level as the reactor was initially brought to
full power and of transient tests at several»pOWér levels'after'steaéy-
state full-power operation had been satisfactorily demonstrated. |
Additional transient tests were performed with the rod control in
"Manual' to demonstrate the inherent self-regulating characteristics of
the reactor. The autdmatic load control was tested during both a load

increase and a load decrease.

TEMPERATURE SERVO CONTROL OF NUCLEAR POWER

Steady-State Performance

 

- The steédy~state evaluation of the temperature servo was based on
the degree of hunting by the regulating rod, cycling of the reactor power

or outlet temperature, and aﬁy long term drifts or cycling.

 

1J. R. Tallackson, MSRE Design and Operations Report, Part II-A,
Nuclear and Process Instrumentation, ORNL-TM-729, Part IT-A, Oak Ridge
National Laboratory, February 1968, p. 229.

- 2Tpbid., p. 30L.
With two exceptions the steady-state performance'of the temperature
servo has been excellent. Farly in the power operation of the reactor, a
daily cycle of 3 - 4°F occurred in the reactor outlet temperature as
measured by the computer and other instrumentation. ' The temperature
servo, however, indicated that the temperature was steady. This problem
was found to be the lack of room temperature compensation of.the thermo-
couple signals into the temperature servo controller, and the fluctuation
in the reactor temperature was the result of the day-to-night variation
in control room temperature. There has been no long term cycling or
drifting since the thermocouple compernsation was corrected, There was
cne period of erratic operation when there were oscillaticns in both the
power and the outlet temperature. This was corrected by the replacement
of a faulty operational amplifier in the controller.

- In all other respects the temperature servo has been completely
satisfactory. Adjustments in rod position are made rélatively infre—
quently and there are no cyclic fluctuations in the power or temperature
and no long term drifts in temperature. The steady-state performance
was originally to be documented by 15 minutes of fast scan data on the
computer each day during the approach to power, but this was discontinued

after the uneventful operation of the controller was observed.

Normal Setpoint Change

 

The first set of transient tests was changes in the reactor outlet
temperature by changing the temperature demand setpoint in the normal
manner, The temperature setpoint can be changed by moving a seléctor
switch to the "increase”’or "decrease" position until the desired tempera-
ture demand is feached, The setpoint is motor-driven at 5°F/min so that
the reactor»can follow the setpoint with about a 1-Mw power change. The
test program consisted of increaéing the temperature demand from 1210°F
to 1225°F, decreasing the demand to 1200°F, and then increasing the demand
to 1210°F. This test program was run at poWer levels of 1, 5, and T Mw.
The data was recorded on magnetic tape by the computer fast scan during
the transient, Appen&ix A lists the date, time period, and tape number
for each of the transient tests that was run. The results of a typical

setpoint transient are shown in Figure 1 which is a‘bomputér.plot of data
  
 

FHSTSCHN

 

 
 
  
 
 
 
  

   
  
 
  
  
  
    

    

 
  

      

  

   
  
  

     
 

 

PIGURE - 1 H e z==“!§"-355n PR ;iz: L !a*iifgks‘fa’-i i
i NORMAL SETPOINT CHANGE 1210 to 1225°F =’E=§“‘§si'§§3 HiH iiii!!ﬁ ii E?&iiiii %iiﬁiiiﬁﬁ;!’#iﬁ i Lﬁiﬁxéﬁﬁﬂ.ﬁiﬁa?#ﬁi giﬁiixﬁﬁ i ‘Eﬁi"d;-!-ii“ E!:’ i‘ﬁg‘fgﬁi'! ;
H sl ot TNTTIAL POVER of 7 Mw HH ““ii!i,;?ﬁﬁii ""“EEE i BREREHE i "“'E!i!"=’i"i"“5”“"s§.‘f' i
R i - R nEtH sﬁ! ﬁ% R §ii§!.§==5.s==E’Ei‘i‘=i=155!§u'iﬁs=s§‘=5§iﬁaiﬁi! Hii
i ik Eiil T T —— ...a,g...:;;;iii..xiﬂ;‘mﬂﬁ ’i-Enmiﬁ?ﬁE“ﬁ?:é‘ﬁéﬁi’i“ﬁiﬁﬁ‘iﬁ e ii‘u‘ii‘éi !ain i
T : i i HHH ’3’::i5!=li"f_i5! L "?:‘=§__== := LR ii: =ii=-=i§i HHE i:l 3%‘:5 ::5! e 5::!:355 !ii!;I!.IiHEEEiE.-ii‘!iiiii iiii;_’ i §§§i=.§=!i'i§=ii‘=§ii" ‘3. -==!==.'i!!'5=! =ill‘ii'i'lgilggi!§§i‘=! i=§§!3=' Eii‘gl.iiggg'gié §§5§§=:= o
--------- o i i é;i.ii’..#i Iggf.auﬂﬁ.!?' i =.=z==5§'§555=§_= != T .ﬁiiﬁisizs"’!‘?ﬁﬁi-ﬁ‘ T ‘i‘:i.ainsiiaisgiigg.ai==s==s;= ;;sim
e ﬁi‘“‘ﬁlﬁ‘aﬁiﬁﬁﬁ" ' “'iiiiiﬁsi’ ‘iﬁﬁiﬁﬁ?? ﬁgﬁi: S 5,5 i
T .

mﬁﬁ!&sﬂ!éﬁﬁsﬁim nﬁéi s ,E’Eii .3 ’!’ﬁiiii'ﬁ!i!i!i"i
i i E'M- i
ﬁ“iﬁﬁiﬁ’!ﬁ ﬁ;’!ﬁ ﬁg’ﬁ. ';lgi [ :.. =.===. ====== :i::!!i!‘!

 
 
 
   

      
  
 

HH !i;li.... ,i!‘?ii"

i. |=.. '§§.: L '-=i§.,=§‘;!_s_=|!’§

’i‘aqa;
== ] “ EE ii m:E
=i§§g e ’==EEE"E‘:,:5’. | “ ;;i - ﬂiagi:ﬁ;, ;i% §I§ gég gnii‘i it

t e R

  
    
 
 
 

   
  
 

==...

     
  
  

  
   

  
 
 
  
 

 
 
   

  

  

   
  
 
 
  
   
 

     
 
 

   
           

         
 

    
 
   

     
 
 
      
   
   
     
    

    

  
    

      
 
 
 
  
  
  
  
  
 

   
  

  

=;;='= EiEEE T T T T i a.? s !=§i=sisaai?.ﬂ"5=5 i _¥§= S i.!=5=! L HEmE it

i ”5’5‘4’“ §-§s§5’=§ i i B '35"‘3?5" i “?iiﬁ’ii!i%ﬁ' f’iﬂi!?!“?i i ’?hin‘ Eﬁ‘iﬁ?“&%ﬁﬁiﬁﬁ L HEEIER R -:"‘Ei,!ﬁ"ﬁ:" *’u i :’?u’iﬁ’i i =‘5H” i E._E'.’... E'Eiﬂiﬂi’ i i

i L BB e 'aﬁ"iﬁ §§5§Ei!sa==§!i i ii’!is ﬁiﬁﬁ;ﬁiﬁ" i ;iifiiiii. ;s ’Hﬁg’ﬁ!ﬁﬁﬁiﬁ L ifi‘iﬁii"'i iii.“%!iiiﬁ‘i;sﬁﬁ‘ﬁ'ﬁ! e "ii.i’i o aii
i ﬁ””“ el %iﬁiiéﬁiiﬁfiﬁif aiizi T i L E'hi'!!ﬂ' e L ES‘ iy
I i -i.iii’;Egg, T %"" ’ “'ﬁi ‘“‘Fﬁi Eiﬁ“’ i !=isii§i¥:§ ==5E: isﬁi‘; 35” i ] ?Eiiiig? .ﬁ‘ =§5§'i'EE’ig'gliiﬁg"'“’-i:gi,g:fgé,ﬁ, s i ] il ;gEE!!'-
| i gisﬂg&' R isiiii‘i}sﬁ;:ﬁgﬁ"iiﬁ-ﬁil Hi EH i i Hi -i i

| :Z "ﬁﬂqﬁ_ﬁ’ﬁﬁ , ’ﬁ% ii-i"ﬁi’ﬁai“ﬂ.i-”gs hﬁiﬁiai%gﬁ;g E ’% ﬁ”‘ ﬂiﬁ’ii iiﬁéi! ‘ﬁ!ﬂia?iﬁ!ii :ﬁiﬁiﬁs B

ig r?ﬁf:fi;?%ﬁjgiﬁfﬁiii i ;i‘: I sa= il ia Inggiﬁfnia‘ﬁ;%:“ %?15335 ;;;i '; iﬁﬁ g::igiiii‘i i;ii%giii ;’i{;ﬁ; il %Eﬁ'ﬁ?ﬁﬁgi E..ii 5 gﬁ:ﬁ% i ='E " --‘ '5 '==5- -

i e =-=a§-:s!s;i§ﬂ;a=sssaaaai,;:;-ﬁ;hiﬁ=-a: -%1‘5“ Eﬁ‘i§= i ig L L aségfjiﬁéiﬁi L iﬁi%igﬁ’&m e .

| o ii:i!giagiﬁ?!‘“,ﬂ:iiiii i ﬁ’:: i ﬁjﬁ iii%' @i’ iiii!ﬁ !E!liiiliiiiiliiﬁiiiﬁuﬂilﬁﬁi hﬂgiiiisai"as: il ﬁgiiigh%iﬁii iﬁﬁ;ﬁﬁ"‘ﬁéﬁ‘"’f‘"’ﬁ!ﬁaﬁ:ﬂ i =E=§:' E;: =§-3g :E:ii::' s gé ﬁg i
.‘ :’r i §.i=i§5' i,i‘iiﬁgi'iihgiﬁ!ﬁi 5! iﬁﬁ g ; Efii.ig’ifﬁ! iﬁﬁ i !ﬁfs H ’ :==§ CTR ' :li! @ﬁﬁ“ HHH iﬁij!ii 5;%’ :i Elﬁ ﬂ!’ﬂi ::i% ﬂgiﬁiﬁi i‘!i-iéii!i iH e Hii! =E5’=§‘ !iiiﬁ E; ii.'i =‘§lﬂiﬁ!fil§§§
e e = ..

i m??wnamu I e ] !!; e R “*"”“’ L ﬁﬁiiﬁﬁéﬁiﬁﬁﬁiﬁ L i §= "Ei!ig'i ';E_'r:j-i' 'iii’m:'::;néi ;géii;:_nj-wﬁmw;”:
H Gl -E.:‘“' e ’::*:’!f.;!r.ni. HH

 

=!=a§=ﬁiiii§rr=5‘§§c§=-iis’i=a= ilsi;;. L§i=s§=§i§i“i e

  
   
 
 

       
   
 

 

  
  

  
  

   

E:igg niil;g:"" ":’Eilii i;?““-m‘?... !Jﬁ?ii EE“ isl E’i: CTOR nna aa H I’gl ;i EE;;.F’ ﬁ:;”:;i.ﬁ Eﬁt?lgl QE
ﬁ%@ﬁ@g n%d%%ﬁ?ﬁi.ﬁ*%myﬁ ﬁﬁ%ﬁ%@% m?ﬂ %% %ﬁﬁ

  
 

.:E ! HH H
R =====:'§ "§= ag =§ i: HE ] ".) i !! o ; §= f' :r l!!l = r! E“,ﬁ!&:ﬂ B '5“35 E‘mnimnigﬁiilﬁglw g iﬁ ST Ei‘:::::a:
..... mﬁﬁﬁfﬁﬁﬁﬁﬁwﬁﬁa' i h:%ﬁ% i %ﬁﬁgiw%ﬁ%ﬁﬁ%s%wﬁfwﬁ%fiﬁiﬁs
i H H “":“ T T : 1» .. i H »_r S - HH 1 e Eiainﬁlg':; ’E'“ e T HHRh llﬂ' R
ﬂfﬁﬁﬁ%f%ﬂﬁ%ﬁ%ﬁﬁﬁﬁLﬁa_ a_“ ﬁ’s@ . gﬁﬁﬁ %%ﬁﬁ%%ﬁﬁﬁ %hﬁrmééﬁﬁﬁﬁﬁﬁﬁaﬁﬁﬁwﬂ
]

   
  

.

E%g%i i .giii“! =E“ﬁi!:i a'mi"i": Y ijﬂ i
L E“i'niis. “=§'=5= i ‘iﬁ" i
==§§'55‘i =¥’!=5=, Eﬁi'ﬁiiia i

e T

i She ""s; et e siain Gl
IR %iigka i i ‘faﬁﬁﬁﬁg'ﬁﬁi’d@"ﬁ% e é; iﬁl!g.tga sii i
4 L i AN R ﬁiiéiissg ;3_;! ’!; ;Eég ‘:i:..iﬁ%iii’"i‘m,,_ o 553 gf;;*: i ,sﬁiiﬁiz’:ii.iéﬁjgi
] @gwme. o mhﬂm@mm“'sﬁ#ﬁﬁﬁWWW$mﬁﬁmmwﬁ%ﬁﬁ#ﬁwﬂw_ﬁﬁ%ﬁﬁmi
%' Siai,;..'ididﬂi..i::iiaq i=“*=§.!& !iii i =’§§i=5'i’='E§!5§E=!§55'§i il iﬁ; i R i L

STARTING DATESs 1-29-67

%ﬁﬁ_ﬁ: ﬁ""ﬁ! *“*az-s'éﬁ?e =""-~

T
'f?*ﬁmm%ﬁﬁﬁmﬁﬁﬁﬁ%

si=i.=§aa.as§“§§5 e
4;@%g%ﬁﬁ%§%,mmﬁd%mﬂmmﬁgﬁ
S i i:!“’:ﬂ,: A i ‘:::"!iin’ NI T
:Elﬂiiigiﬂ::l!:i _gé,:é;u;;;ggzs! g;';;“ !'5!;5: E;,::éi aﬁ';i rﬁliggiglii!“f;é?i ;Eg :Ei . ﬂ.E:::i _4.}
. ;-’E,i'i"’ . . -
i 5=§==§§-’ iL, e e L ‘“iii“ii“iiii‘ﬁﬂﬂ'ﬁ ﬁﬂ

2HH T =§§
i =5’ "’ =si'iii§§!§ :::E'q's"i‘-i‘:iiiﬁmiau*ﬁ:‘ i iiii ‘Ei’m M EESH!

B
o

H
H
!

===G'-€"'

 

  
  
       

-

Eﬁ”iﬁ'

  

H oo e =t
o 1 T
A% HH
R mgﬁ ]
Tt %EEE
..‘ E LT I..
-+ I
3
T
7 1

£ ,E_ELE;E;""

  
  

  
 
 

 
    

 
stored on magnetic tape by the MSRE on-line computer. Data was recorded
every 1/4 second during the tests in this report. The temperature set-
point was increased from 1210 to 1225°F at an initial power level of
about 7 Mw. The scattered points on some of the traces are computer
errors in reading the signal and are not actually changes in the par-
ticular signal. If the temperature demand setpoint were also plotted, it
would fall essentially on top of the reactor outlet temperature (TE-lOO~lA)
trace. The results of the other tests in this series were similar except
“for the temperature decrease from 1225 to 1200°F at a power level of 1 Mw.
In this test the;reactor outlet temperature could not track the setpoint
because a sufficient decrease in power was not available to pre&uce a
5°F/min temperature decrease., The control system cannot request a power
level below 500 kw so that a power decrease of only about 0.5 Mw was
available as compared to the 1-1/2-Mw increase which occurred during the
test plotted in Figure 1. This is not a deficiency in ﬁhe control” system
since the only effect is a longer transient time in reaching the new

temperature.

Rapid Load Changes

 

The second sefies of transient tests consisted of making sudden
changes in the heat-removal‘rate at the radiator and observing the
reéponse.of the system as the temperature servo adjusted the reactor
power to matech the heat load. The following load changes were made:

1l Mw to 3 Mw
3 Mw to 5 Mw
5 Mw to 3 Mw
3 Mw to 7 Mw
T Mw to 5 Mw
| 5 Mw to Full Power (T7.22 Mw)
The results of the 3 - 7 Mw load change are shown on Figure 2., This

load change was made by changing the radiator settings as outlined below:

 

 

 

 

Nominal Power Inlet Door - Outlet Door Damper ‘Blowers On
3 30" ‘Upper Limit 100% open MB-1

T U.L.  U.L. 50% MB-1 and 3
FASTSCAN

FIGURE - 2 ‘
STEP IOAD CHANGE 3 Mw to 7 Mw

in TEMPERATURE SERVO

 

STARTING DATEs 2 1 B7
 

STARTING DRTEs 2 1 b7
9

The load change was intended to be completed in about one minute, but in
practice, three minutes were taken to complete the change in radistor
settings. Although this load change was much slower than intended, it
was representative of the normal rate of increasing the power. The re-
actor power increased emdothly without oscillations or overshoot and there
was essentially no change in the reactor outlet temperature. The regu-
lating rod also showed little or no adjustment during the power transient.
The sine wave fluctuation on the regulating rod trace, and'also ch some
of the other traces during this and some of»the other tests, was some
type cf noise signal in the computer. The rod was actually more or less
stationary. The temperature servo could have made some,small corrections
in rod position that were masked by the superimposed sine wave. The
lack of regulating rod adjustment and the constant reactor outlet tempera-
ture indicated that the réactor is inherently stable and self regulating.
A similar, but more severe,'test was conducted when the autematic
"Load Control" was tested. The results of this test are shown on Fig. 6.
The temperature servo made only slight rod adjustments during this test
10 maintain a constant outlet temperatureo The power transient was

smooth without oscillations or overshocot.

Temperature Setpoint Error

 

The most severe transients were introduced into the reactor System
by switching from manual rod control to temperature servo control with a
mismatch between the actual reactor outlet temperature and the controller
setpoint., The servo controller immediately requests a power change in
‘proportion to the temperature mismatch to adjust the reactor outlet to
the setpoint temperature. The controller was restricted and could not
request power levels above 11 Mw or below 500 kw during these tests,
The upper limit has now been changed to 8.625 Mw to be consistent with
full-power operation at 7.5 rather than 10 Mw. The actual'response of
the reactor power is slowed somewhat by the rate at which the regulating
rod can be withdrawn and by the "rod withdraw inhibit" whieh limits the
 reactor period to 25 seconds or greater. | | |
The following test program was used to evaluate the performance of

the controller in handling the relatively fast transients as the controller
10

corrected the temperature errors. This program was designed to test the
controller under a variety of conditions and also to test the limits of
500 kw and 11 Mw to see if these limits would be seriously exceeded. The
transient performance was recorded for temperature errors_ef + 5°F, + 15°F,
and + 25°F. These temperature transients were repeated at starting power
levels of 1, 5, and T Mw.

The results of the tests showed that the controller was more than
adequate to handle the transients that resulted. Figure 3-shows»the
response of the system during the most severe transient when.a +25°F
setpoint error was introduced at a power level of ~ 7 Mw., The initial
withdrawal of the_regulatingrrod was stopped several times by the 25-sec.
period "rod withdraw inhibit", The.power overshetﬁthe~ll-Mw limit
slightly (0.43 Mw) as the coumtrol rod was beingaiﬂserted to stop the
power increase, As the reactor temperature approgched the setpoint, the
reactor power was reduced relatively smoothly to the steady-state value,
The outlet temperature overshot the setpoint about 5°F. Figure L4 shows
the system response when a -25°F temperature error is introduced. The
reactor power was first reduced from ~ 7.1l Mw to ~ 1.5 Mw and then re-
turned to about the original value. The regulating rod was limited in
this case by the lower limit of its normal 6-inch range. If this physical
1imit on the rod motion had not stopped the rod insertion, the rod would
probably have been stopped by the "power less than 500 kw" limit. There
was 8 small overshoot as-the power was being increaSed_to the normal
operating point and there was a small undershoot in the reactor outlet
temperature at about the same time. The small smounts of overshocot and
undershoot shown in these tests were of no practical significance.

The tests conducted at ether power levels were similar to the ones
shown and the only difference was an increased tendency to oscillate at
the 1-Mw power level, However the'oscillations.in power and temperature
were small and were satisfactorily damped in all cases. The overshoot
or undershoot in the»temperature and,powerwere-damped out during the
first cycle for the tests at 5 and T Mw. There was no overshoot or

undershoot for any of the tests with a setpoint error of + 5°F.
FASTSCAN

FIGURE - 3
SETPOINT ERROR OF +25°F at INITIAL POWER of T Mw

STARTINC DBATESs

2

1

 

67

T
FASTSCAN

FIGURE ~ b
SETPOINT ERROR OF -25°F at INITIAL POWER of T Mw

 

STARTING DATEs 2- 1 67

cl
13

 

Control by Negative Temperature Coefficient at Reactivitz

After the above testing of the temperature servo controller was com-
‘pleted, a series of.tests was conducted to demonstrate the inherent con-
trol characteristics of the MSRE. The stability analySiS3 and the dynamics
testing4 that were completed previously showed that the system was stable
under all operating conditions and that the system would’be'self-regulating.
However, no previous tests had been run to demonstrate the system response
to sudden load changes. The following test program was conducted with the
load changes being made as quickly as was practical.
| Fullﬂpower to 4k Mw
L Mw to 2 Mw
2 Mw to 4 Mw
L Mw to 6 Mw
6 Mw to Full power
4 Mw to Full power
2 Mw to Full power
Figure 5 shows the results of the final test when the radiator heat
load was increased from.~ 2 Mw to full power in about 50 sec. with no
control rod movement., As can be seen in the plot there was a smooth
increase in power with no tendency to overshoot or oscillate. The re-
actor outlet temperature remained essentially constant except that there
was about a 2°F loss in temperature at about 150 seconds which recovered
‘about one minute later. The chenge in the rate of power increase occurred
at about the same time as the minimum reactor outlet temperature. The
other tests in the series were similar in having a smooth power trace
without overshoot or oscillation and having a constant reactor outlet
temperature. The 2°F dip in the reactor outlet temperature mentioned
above was not present in the other tests. Some of the early ana.log

studies on the MSRELhad indicated oscillations in the nuclear power and

 

°S. J. Ball and T. W. Kerlin, Stability Analysis of the Molten-Salt
Reactor Experiment, ORNL-TM-10T70, Oak Ridge National Iaboratory, Dec. 1965.

4T, W. Kerlin and S. J. Ball, Experimental Dynamic Analysis of the
‘Molten-Salt Reactor Experiment, ORNL T™- l6h7, Oak Ridge National
Leboratory, October 13, 1966.
TEMPERATURE (°F)

1k

ORNL—DWG 68—964

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

15
=z
3 10
& | REACTOR POWER
= - -
2 5 //
//

0

1250
REACTOR OUTLET
-
1200 —] |
\\ REACTOR INLET
M50 ‘\\\\
| \ RADIATOR INLET
. ‘\\\\\ — ]
1050 RADIATOR OUTLET
1000 \ —
0 40 80 120 160 200 240 280 320
TIME (sec)

Figure 5. Step Load Change 2 - 7 Mw with No Control Rod Motion.
15

in the various system temperatures especlally at low-power levels,
However, there were no indications of oscillations in any of the tests
listed above. |

This series of tests demonstrated the self-regulating characteristics
and the stability of the system under transient ioad conditions. The
main difference between the step load changes with the servo and without
the servo was‘that the initial change in power was faster when the servo
was used. The.servo made only a few rod adjustments after the first
portion of the transient. The nearly constant reactor outlet temperature
at the various power levels indicates that the nuclear average temperature
of the reactor is about equal to the fuel out let temperature and confirms
the very low power coefficient of reactivity (based on reactor outlet

temperature) that had been reported previously.>

AUTOMATIC LOAD CONTROL

The Automatic T.oad Control on the MSRE provides a method of in-
creasing or decreasing the radiator heat removal rate. The load control
either increases or decreases the load in a preselected sequence and it
does not actually control the load at a given level. The automatic load
control was designed for use at power levels greater than 1 Mw, and a
radiator door setting was to be specified with one main blower in opera-
tion to provide a starting point at 1 Mw of heat remOval. However, the
heat removal was about 1.9 Mw with the radiator doors at their minimum
open setting of 12 inches. The test during a power increase was started
from this point, but the power could not be lowered below ~ 3 Mw when
the load was decreased because the radiator doors were stopped at 50 in.
by the interlock switches.

The results of the load control test are shown on Figure 6. The

regulating rod position was not included on this plot because of a

 

S0ak Ridge National Laboratory, MSRP Semiann Progr. Rept. August 31,
1966, USAEC Report ORNL-L4037, p. 11.
FIGURE - 6
PERFORMANCE OF AUTOMATIC LOAD CONTROL

 

STARTING DATEs  9-21 67
 
17

relatively large low-frequenoy noise signal that existed in the computer.
No rod motion could be detected from the computer fast scan data during
this test. However in an earlier test, the rod was withdrawn about
0.9 inches at the beginning of the power'transient and was reinserted
to about the original position after about 30 - LO seconds. As can be
seen in the plot the power and temperatures are smooth without overshoot
or oscillation. -

The time sequence of the various control actions taken by the auto-

matic load control are given in Table T.

Table T

. | | | | | Computer
Action Time
1., ILoad Switech to "Increase'" — Radiator Doors start :
| | to raise S 1159:00
2. TRadiator Doors at upper limit — Bypass Damper started —
to close - 1159:28
3. Bypass Damper closed — Automatic Main Blower energized -1200:02
k. Bypass Damper partially open —-starts to reclose 1200:30
5. Bypass Damper fully closed | 1200:58
6. Reactor at full power | 1204 :00
T. Load Switch to "Decrease'" — Bypass Damper starts
to open - 1208 :24
Bypass Damper open — Automatic Blower deenergized 11209:00
9. Bypass Damper starts to close — Radiator Doors close
| - slightly 1209:10
10, Bypass Damper partially-closed'é-begins to open 1209:18
11. Bypass Damper open — Doors start to close 1209:35
12. Doors at intermediate limit | 1209:52
13, Reactor at minimum power_by automatic load decrease | 1213:00

‘The closure of the radiator doors at Step 9 had not been planned at
this point; However, the interlock conditions were.satisfied at this
time when the damper was open and the automatic blower was off. This

~problem was not serious and could be easily corrected by adding the
18

appropriate interlocks. As soon as the damper élosed'slightly,,the
lowering of the doors was stopped until the damper again reached the full
| open position. At this time the doors were lowered to their intermediate
limit as had been intended. The only other fault noted was that the
minimum power available to the sutomatic load control is abdve the transfer
point from Start to Run Mode. The bypass damper was also too slow to
follow the acceleration of the main blower, but this has no practical
significance. |

| Test Memo 5.2.1 outlined a series of tests in which the automatic
load control was to be used to increase and decrease the reactor power
in 1/2-MM'steps to determine any discontinulties in loading curve, These
tests were not conducted because the heat load at various radiator
\éettings had been previously determined. Figure 7 shows the performance
of the radiator at various settings. The heat load using the automatic
load control would follow the solid curves. There is a small gap,inhthe
heat-load curve at about 6 Mw when the second blower is energized,

However, this gap can be covered by manually adjusting the doors,

CONCLUSIONS

- The performance of the temperature servo has been excellent except
for the two cases mentioned earlier in this report. The servo adequately
controlled all the transients that were introduced with a minimum of
overshoot and oscillation,. In normal operation of the reactor, the
temperature servo is not called upon to control rapid transients because
of the relatively slow response and.fhe inherent stability of the system.
When operating at steady conditions, rod adjustments are made infrequently
and a steady power level is maintained by the inherent self-regulating
characteristics of the reactor. | o

The automatiec load céntrol performed generally as expectéd except
that the door settings for 1-Mw operation were below the drop point when
one blower was operated. If the intenmediate 1imit switches were set

at the minimum position, the reactor power would be about 2 Mw. This
DAMPER POSITION (% CLOSED)
DOORS OPEN

LOAD SETTING

DOOR POSITION (%% OPEN)
DAMPER OPEN

19

ORNL-DWG 67-4763
.

»

100

 

 

@
o
Sy

\~\‘

 

 

[ / /

1 | | A/ /'/

/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TWO MAIN
/' » BLOWERS
- RUNNING
0 ® A -
/ BOTH MAIN BLOWERS OFF
80 /
| /
/ A
/ |
60 _
/ ONE MAIN BLOWER RUNNING
40 ,/ _ 7
/ A
20 / ' - 7 INLET AIR TEMPERATURE =40°F
/ Atk
0
o - 1 2 3 4 5 6 7

HEAT BALANCE POWER (Mw)

Figure 7. MSRE Radiator Performance.
20

is not an acceptable power léVel in regard to switching between start
and run mode, and therefore the'switch.settings have not been changed.
Since manuvally adjusting the heat load is not'a serious disadvantage,
the automatic load control has not been routinely used in the operation

of the reactor.
MAGNETTC DATA

TAPE STORAGE

 

 

Test Computer
Memo Test Description Date Time Tape
5,2,1.3 Normal Setpoint Change 1 Mw 1210 to 1225 1/29/67 0300 - 0315 M-103 and 104
. 1225 to 1200 1/29/67 0332 - 0336 M-104
1200 to 1210 1/29/6T7 1005 - 1021 M-104
5 Mw 1210 to 1225 1/29/67 1154 - 1206 M-10kL
1225 to 1200 1/29/67 1222 - 123k M-105
1200 to 1210 1/29/67 1321 - 1332 M-105
T Mw 1210 to 1225 1/29/67 1415 - 1426 M-105
1225 to 1200 1/29/67  1h54 - 1508 M-105
1200 to 1210 1/29/67 1524 - 1537 M-105
5.2.1.4 Step Load Change 1 Mw to 3 Mw 2/1/6T7 0045 - 0130 M-106
3 Mw to 5 Mw 2/1/67 0130 - Ol4l M-106
5 Mw to 3 Mw 2/1/67 0156 - 0210 M-106
3 Mw to 7 Mw 2/1/67 0331 - 0351 M-106 and 107
T Mw to 5 Mw 2/1/67 0355 - 0kLOT M-10T7
5 Mw to T7.22 Mw 2/1/67 ok10 - 0423 M-107
5,2,1.5 Temperature Setpoint Error 1 Mw + 5°F error 2/1/67 0920 - 0932 M-107
‘ - 5°F error 2/1/67 0932 - 09hLT M-107
+15°F error 2/1/67 0950 - 1005 M-10T7 and 108
-25°F error 2/1/67 1005 - 1022 M-108
+25°F error 2/1/67 1038 - 1048 M-108
5 Mw + 5°F error 2/1/67 1103 - 1108 M-108
| - 5°F error 2/1/67 1108 - 1120 M-108
+15°F error 2/1/67 1120 - 1129 M-108
-25°F error 2/1/67 1131 - 1143 M-108
+25°F error 2/1/67 1143 - 1152 M-108 and 109
T Mw + 5°F error 2/1/67 1205 - 1219 M-109
- 5°F error 2/1/67 1220 - 1225 M-109
+15°F error - 2/1/67 1227 - 1234 M-109
+25°F error 2/1/67 - 1238 - 1301 M-109
-25°F error 2/1/67 1303 - 1315 M~-109

 

XIANHIAY

1<
MAGNETIC DATA TAPE STORAGE

 

 

(continued)
Test Computer
Memo | Test Description Date Time Tape
5.2.1.7 Control by Negative |
Temperature Coefficient 7 Mw to L Mw Cqa )
| b M to 2 My 5/L/67 0945 - 1010 M-122

L Mw to 7 Mw 5/4/67 1058 - 1111 M-122

2 Mw to T Mw 5/L/67 112k - 11L0 M-122 and 123
5.2.1.6 Automatic Load Control Load Increase (2 to 7 Mw) 9/21/67 1159 - 1208 M-1L42

Load Increase (7 to ~3 Mw) 9/21/67 1208 - 1215 M-1L2

 

cC
17,
18.

20.
21.

22,

23,
2L,

25,
26,
27
28,

29,

30.
31.
32,
33.

35

=0 > HuynOOHgOOQOEHQRE OE n W

23

~ DISTRIBUTION

MSRP Directorfts Office
Bldg. 9204-1, Rm. 325

. K. Adams

G. Affel

J. Ball

. Bauman

F. Beall

Bender

k., Bettis

Bettis

Borkowski

Cope, AEC-ORO

. Crowley

. Culler

Deering, AEC-ORO

. Ditto

. Ingel

Epler

. 'erguson

. Fraas

. Gabbard

Giambusso

AEC-WaShlngton

. Guymon

. Harley

. Harrill

. Haubenreich

. Herndon

. Hudson

. Jordan

. Kasten

Kedl

Kerlin

. Kerr

. Krakoviak

. Krewson

®

®

) O R O G b e G

. @

C-IP>SISI—II':U"U.2I-EI"U"UO"U':U
Z:HI—EI‘.Z!C-I':U';E!L—'QZU)m';E{

¢

T1l=T72.
73-Th.
75-TT.

L,
L8,
k9.
50,
51,
52,
53.
5k,
55.
56,
57.
58.

59.

60.

61.

62.
63.
6k,
65.
66.
67.
68.
69.

0.

Central Research ILibrary
‘Document Reference Section
Laboratory Records

HoQmowmY9 =2 = g

SRrER O

. lLane
Larkin, AEC-ORO
Tundin
Lyon
MacPherson
MacPherson
Martln
cClaln
. McHargue
. McIntosh
~Washington
McWherter
. Moore
Nelms
. Nicholson
Oakes
Perry
Pickel
Prince
. Ragan
Redford
Rlchardson
R C. Robertson
H. M. Roth, AEC-ORO
Dunlap Scott
M, Shaw
AEC-Washington
W. L. Smalley, AEC-ORO
I. Spiewak
oteffy, Jr.
Tallackson
. Thoma
Watts
Weinberg
West
. Whatley

®

E

® -
o, ®

IHEESROHPHTAOSGR0EQEHG

ZQQWH!DL"MDJ'JUQ'

HERPEP0

T78. Laboratory Records (LRD-RC)