ORNL-CF-60-11-20.txt

N

X-822

 

DATE:

SUBJECT:

TO:

FROM:

OAK RIDGE NATIONAL LABORATORY
Operated by

UNION CARBIDE NUCLEAR COMPANY

Division of Union Carbide Corporation

ucc

Post Office Box X
Oak Ridge, Tennessee

November 4, 1960

S8

;
i

|
INTERNAL USE ONLY

 

 

ORNL
CENTRAL FILES NUMBER

GO0-1[1-2o

 

COPY NO. 9[:

MSRE - Analog Computer Simulation of a Loss of Flow Accident in the
Secondary System and a Simulation of a Controller Used to Hold the

Reactor Power Constant at Low Power Levels
G. A. Cristy
O. W. Burke

- ABSTRACT

These analog computer studies had two major aims.

The

first aim was to give the MSRE design personnel an indica-
tion of the time-temperature relationship of the secondary
salt as a result of loss of flow in the secondary system.
The second aim was to check the feasibility of a closed
loop primary pump speed controller to hold the reactor

- power constant at a low power level.

The secondary system flow loss was simulated, and the
secondary salt temperature at the radiator outlet reached
the freezing point in thirty (30) seconds. The radiator

fan remained in operation.

'The above mentioned flow controller was simulated. The
gain, damping, etc., were varied over a wide range, seek-
ing a stable controller. The controller could not be made

to operate in a stable manner.

NOTICE

This document contains information of a preliminary nature
and was prepared primarily for internal use at the Oadk 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 Deportment.

 
J

II.

INTRODUCTION

From an earlier investigation reported in ORNL C.F. 60-6-110, the
reactor system designers reached the conclusion that the primary salt
temperatures attained with loss of primary flow would cause no
difficulty.

The designers were concerned about the very definite possibility that
the secondary salt would freeze subsequent to a loss of flow accident
in either the primary system or the secondary system. The freezing
would be caused by the continued operation of the radiator fan after
a8 loss of flow accident. |

During a preliminary investigation using the analog computer, 1t
was determined that the secondary salt temperature would reach the
freezing point much sooner for loss of flow in the secondary system
than it would for loss of flow in the primary system. In view of
this fact, only results obtained as a result of loss of flow in the
secondary system will be included in this report.

The designers of the reactor system proposed using flow control in
the primary system as a means of holding the reactor power constant
at a low power level. This controller was simulated on the computer.

DESCRIPTION OF THE SYSTEM SIMULATED

A. Thermal System

 

The thermal system was based on the latest design information
and it differed appreciably from that used in the preliminary
simulation that was reported upon in ORNL C.F. 60-6-110. The
design information used in this simulation is included in this
report on pages 5 and 6.

In order to preserve design point temperatures around the primary
loop, the "after heat" was added as if it were all released in-
side the reactor. Of course, this is not true; however, the
error appears to be very small. Pipe losses, which were ignored,
would tend to offset the "after heat" generated outside the re-
actor. ‘

The loss of flow occurred exponentially on a three (3) second
period. At this time there was no available pump "run-down"
information.

The heat transfer coefficients between the secondary salt and
the primary heat exchanger wall and between the secondary salt
and the radiator wall were made to vary with flow rate in
accordance with the curve shown in Figure 1.

The temperature differential between the hot and cold legs in

the secondary system produces a force due to the density differences
in these legs. This force induces flow in the system. This
phenomenon was incorporated into the simulation.
III.

B. Nuclear System

The delayed neutrons were lumped into one weighted group.
Using the values from document LA-2118, dated 1957, A for

the one group was found to be 0.0769 sec.~l and @ was found
to be 6.4 x 103, Using the given fuel transit times and the
curves in ORNL-LR-Dwg. 8919, A& was found to be 1.800 x 10~3,

where ) @
/43 = z Gfié/é&
<=/

i ig the ratio of the population of the ith group of de-
layed neutrons inside a circulating fuel reactor to the popula-
tion of this group in an equivalent stagnant reactor.

In the simulation, the delayed neutron contribution varied
with primary flow rate. The simulation method was admittedly
not precisely correct, but it was considered a good approxima-
tion. The delayed neutron contribution is correct at design
point steady state and at zero primary flow. The variation of
the delayed neutron contribution between these two points is
only an approximation.

ANALOG COMPUTER PROGRAM

The analog computer programs of this simulation are filed as ORNIL
drawings, numbers E-40327 and E-40328.

The program for loss of flow in the primary is drawing number
E-40327. The power level controller program is also on this
drawing. The program for loss of flow in the secondary is drawing
number E-40328.

Copies of the above drawings may be obtained from the Print Files
in the E & M Division.

PROCEDURE
These studies were conducted in three phases. These phases were:
1. Loss of flow accident in the primary system.

2 Investigation of the characteristics of proposed power level
controller.

3. Loss of flow accident in the secondary system.
The following procedures were employed:
1. Loss of flow accident in the primary System.
With the simulator in steady state operation at design point,

the primary flow rate was decreased to zero,exponentially on
a three second period. The secondary flow was not disturbed.
-

Pertinent temperatures and nuclear power were recorded versus
time. Runs were made for three different graphite temperature
coefficients of reactivity. |

2. Investigation of power level controller characteristics.

The controller was simulated so that its gain, damping, and
response time were variables which could be changed by changing
a potentiometer setting. With these pots set at given values,
and the reactor simulator in steady state operation with no
temperature coefficients of reactivity, a small perturbation
was introduced into the system. The effect of this perturba-
tion on the nuclear power level was observed. This procedure
was repeated for a number of potentiometer settings in an
attempt at determining optimum settings.

3. Loss of flow accident in the secondary system.

With the simulator in steady state operation at design point,

the flow rate in the secondary system was decreased exponentially
on a three second period. The primary flow rate was not disturbed.
Pertinent temperatures and nuclear power were recorded versus

time.

SUMMARY OF RESULTS

With loss of primary flow, the temperatures attained were of the
same order of magnitude as those recorded in ORNL C.F. 60-6~llO.-

Three runs were made, 351ng graphlte temperature COjgficients of
reactivity of -2 x 10~%, -1 x 10-%, and -4 x 10~ KOF. The
maximum primary temperatures attained in these three runs had a
spread of only 20°F. Loss of primary flow was simulated in these
three runs. Due to the large heat capacity of graphite, the
temperature of the graphite changed very little and the rate of
change of temperature was very low.

The power level controller showed a tendency to be oscillatory and
it could not be stabilized (see memo from E. R. Mann to J. R.
Tallackson, dated September 12, 1960, entitled "Closed Loop Pump
Speed Controller to Hold the MSRE Power Constant at Low Level")..

In the simulation of secondary flow loss, the radiator fan continued
in operation. The temperature of major concern was that of the
secondary salt at the radiator outlet, since this was the coldest
point in the system. This temperature reached the freezing point
(8600F) thirty (30) seconds after initiation of the flow loss

(see Figure 2). This gives an indication of permissible time for

corrective action. /ffi
W el
% .

O. W. Burke
. Design Department

OWB:njn Engineering & Mechanical Division
..5..

Updated Computer Information to Ray Mann, 3rd Issue

Reactor inlet temperature: 1175CF

Reactor outlet temperature: 1225°F

Mean graphite temperature: | 1230°F

Residence time in reactor: 2.02 sec

'F1Im drop from graphite to fuel: linear with power

Heat capacity of graphite: 40Lo §§% (cp = 425 Igzgp“

Prompt ¥ and neutron heating in graphite: 6% of 10 Mw power

Residence time in piping from reactor outlet to H.E. inlet 3.56 sec
- Residence time in H.E. 1.62 sec

Heat capacity of metal in H.E. . 200 B2

Avg. film drop between primary coolant and metal at D. P. 55.20F

Avg. drop in metal at D. P. 564 TOF

Avg., film drbp between metal and secondary coolant at D. P. 26.10F

Film drop between primary coolant and metal as function of flow:
See graph, displace curve if necessary so that at 7.5 fps velocity t = 55.2CF

Mean secondary coolant temp. at D. P. 1062CF
Residence time in piping between H.E. outlet and reactor inlet
(including coolant annulus) | 5.76 -sec
Total circulation time 13.0 sec
Temp. coeff. of reactivity of graphite: -2.X lth . Bé%ffir

- Temp. coeff. of reactivity of fuel: 4.5 x 1077 ~%F
Melting point of primary coolant: 8L42°F
Melting point of secondary coolant: 860°F

Check points:

Thermal resistances: ( EEB%%—EE) \
in primary coolant film: 3.28 x 10~
in metal: 3.32 x :LO"'21L
in secondary coolant film: 1.56 x 10

The rate of heat removal in thermal convection follows the correlation (for
the primary circuit)

448 x 10° At () -9.15 x 1073 x ( &) 0.8 _ )11

where &t is the temp. differential between the hot and fi%fid leg (OF) and
q is the rate of heat removal from the primary system.(—E;)

""loo
 

Simulator data_for secondary loop

Air temperature rise in radiator
Air suction temp.
Air flow

Heat capacity of radiator

Heat capacity of secondary salt

.Density of secondary salt

Residence times of secondary salt:
in primary heat exchanger
in piping to radiator |
in radiator
in piping from radiator
Total

Residence time of air in radiator
Temperature differences in radiator:
in salt film
in tube wall

in air film

200°F
o
100°F
166,000 cfm (7.11 x 10° 32)
Btu

2h2 —'b"E';

Btu
0.57 %507

120 %%3—

0.92 sec
2.56 sec
4.85 sec

L.72 sec
13.65 sec

0.01 sec
2o°F

78°F
762°F
)

 

Joo

-

ESIGY o/ T CoLLF
™

G\
Qo

QD

HEaT TraxsFEL CoEFFICIENT (4 oF O
O |

/o

UNCLASSIFIED
7 "ORNL-LR-DWG. 53662

F/

   

40 o o /oo
SECONDARY Flow RATE (4 oF DESIGH PoswT)

-

   
Distribution

 

1. S. E. Beall
2. E. 5. Bettis
3+« F. F. Blankenship
h-6. A. L. Boch
T R. B. Briggs
8. F. R. Bruce
9. O. W. Burke
10. R. A. Charpie
11, W. K. Ergen
12. C. H. Gabbard
13. W. R. Gall
14. W. R. Grimes
15. H. W. Hoffman
. L. N. Howell
17. W. H. Jordan
18. P. R. Kasten
19. .B. W. Kinyon
20. He G. MacPherson
21 . W. D. Manly
22 E. R. Mann
23. W. B« McDonsald
2k, R. L. Moore
25.. C. W. Nestor
26, L. F. Parsly
27 « H. R. Payne
28. D. Scott
29. M. J. Skinner
30. I. Spiewak
31, J. A. Swartout
32. A. Taboada
33. J. R. Tallackson
34. A. M. Weinberg
35. J. He Westsik
36. C. E. Winters
37. C. H. Wodtke
38=39. Library, 9204-1
LO-U1, Central Research Library
L2-43, Document Reference Library
Lh-53,  Laboratory Records
5k, ORNL~-RC
55 C. E. Bettis
56. G. A. Cristy
57« H. E. Seagren
58. G. Morris
59. D. E. Ferguson
60. MSRP - Director's Office (9204-1, room 259)