I OFL
ORNL P
1679
TEEFE EFE
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
AL 1679

.
"1-65/00%
MASTER
WIN 8
A CONTROL SYSTEM FOR A GAS -COOLED NUCLEAR POWER PLANT*
-
H. G. O'Brien
8. J. Ditto
Oak Ridge National Lazoratory
Oak Ridge, Teanessee
LEGAL NOTICE ...
The report mo propared u an account of Government sponsored work, Noither the United
Statas, nor the Commission, nor any person acting on behalf of the Commission:
A. Makor any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information containod in this report, or that the wa
of my information, apparatus, method, or process disclosed in the report may not infringe
privately owned righto; or
B. Assumes any Habilities with rospeot to the use of, or for danages resulting from the
um of any Information, apparatus, method, or process disclared in this report.
As used in the above, "psrron acting on babalf of the Commission" includes pay om-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such omployee or contractor of the Commission, or employee of such contractor preparse,
disseminatos, or provides accou to, any information pursuant to his employment or contract
with the Commission, or his employment with such contractor.
.
vnit
For Presentation at the
12th Nuclear Science Symposium
Institute of Electrical and Electronics Engineers
San Francisco, California
October 18-20, 1965.

-
RENTABILID TOK MATBOUNCEMAAKT
II MICILORAR SCLEKCE' ABSTRACTS
r
.
.
*Research sponsored by the V.8. Atomic Energy Commission under contrast :
with the Union Carbide Corporation.
4
.
L.
.
- PE
VI
RA
ANA

2
A CONTROL SYSTEM FOR A GAS -COOLED NUCLEAR POWER PLANT
A. Q. O'Brien : : 8. J. Ditto
ABSTRACT
vinmarkennaratinhthuatandaan
A major consideration in the control of a nuclear power plant
Buch as the Experimental Gas-Cooled Reactor is the regulation of
both the temperature and pressure of the steam entering the turbine.
This paper describes a control system which varies the primary coolant
flow and reactor power to make the plant produce steam at the required
conditions of temperature and pressure over a wide range of steam
flows. Analog computer studies showed that the control system was able
to make the simulated plant smoothly follow both large changes of
turbine load (changing at a rate of 10% per minute) and small rapid
changes of load (1% per second) while limiting the transient variations
In steam conditions to acceptable values. .
nako mahimatan tanah adai un
maalattades kan
man
k
-
-
-
..
.--
-.
.
..
---
.
a n sida tanaman
LIB
.
..",
...
.
..
... .com,
.
-.
-.
...
..
.
.
.
. ....
::
::1497 **
1
!: 0
A CONTROL SYSTEM FOR A GAS-COOLED MUCLEAR POWER PLANT
1. Q. O'Brien
8. J. Ditto
INTRODUCTION
The automatic control of power generation in a muclear reactor has,
for many years, been achieved by using closed loop control of neutron
flux as measured by an ionization chamber. Where neutron flux at the
location of the chamber is the variable of interest, the problem 18
quite simple. Control-rod pouition 18 changed in response to the dif-
ference between the measured flux level and a set point as selected by
the operator. The close coupling between rod motion and neutron flux
makes it possible to achieve good control characteristics.
The use of a simple flux controller 18, however, not a very satis-
factory solution for & high-performance research or experimental reactor
or for a power reactor, because changing flux patterns and the incentive
to control variables other than flux (thermal power, temperature,
coolant flow, etc.) make it necessary to make frequent adjustments in
neutron flux level. This paper describes an automatic control system
that regulates the temperature and pressure of steam delivered to the
turbine of a nuclear power plant under, steady-state or transient electri-
cal load conditions. The system was tested on a simulated model of the
Experimental Gas-Cooled Reactor (EC-CR) to determine its ability to
control such a plant.
Although the BOCR characteristics were used in the detailed design
of the control system, we believe that the basic principles used are
..'.
"
.
.
i.men det er e
n
travelino Donna
..
..
.....
......men score
thi
applicable to the control of other nuclear power plants where
temperatures, flow, pressures, etc., are of greater Interest than
neutron flux, per se.
DESCRIPTION OF THE CONTROL SYSTEM
The EGCR 18 an 85-Mw (thermal) reactor being constructed for
the United States Atomic Energy Commission. The fuel 18 uranium
dioxide pellets, enriched to 2.5% uraniuz-235, enclosed in stainless
steel cladding, and surrounded with a graphite moderator. The reac-
tor core is cooled with helium at an inlet temperature of 495°F and
an outlet temperature of 2020°F. As illustrated in Fig. 1, the
helium is circulated from the reactor to two steam generators by a
variable speed blower in each of the coolant loops. The two steam
generators deliver steam at 903°F and 1250 p818 to a single turbine-
..
SA
generator set to produce 29.5 Mw of electrical power.
. If the plant load is to be an independent variable, the control
system must make the plant produce steam at the required conditions
of temperature and pressure over a wide range of steam flows. The
parameters available for control are the reactor power and the coolant
flow. Previous studies of the EOCR dynamic response have shown that
.
S
variations in reactor coolant flow are tightly coupled to steam pres.
n
sure and temperature; however, large heat capacities and poor heat-
transfer characteristics create long lags between a change in reactor
power and the corresponding change in steam conditions. This problem
of loose coupling between reactor power and stean temperature has
one or more och har
en
el ch
5
set
been solved by including a tight flux-control loop in a rod controller
used to regulate the steam temperature Tg. The flux set point set
for this controller 18 computed from the product of the measured reac-
tor-coolant flow W., and the difference between a reactor-outlet-
temperature set point Tro and the measured reactor-inlet temperature.
Top• This reactor-outlet-temperature set point 18 slowly and auto-
matically adjusted until the desired steam temperature T 18 achieved.
A blower controller maintains the steam pressure P, at the desired
value by regulating the reactor coolant flow. Since the steam pressure
is immediately affected by a change in steam flow, and the coolant flow
· 18 a factor in the computed flux set point for the rod coatroller, the
·blower controller couples the flux to the steam flow and provides true.
load-following capabilities in which the reactor 18 a slave to the
electrical load imposed on the turbine.
Conventional feedwater controllers maintain the water level in each
steam drum by regulating the feedwater flow to each steam generator. .
1
.III
A conventional speed governor is used to maintain the turbine speed by
regulating the steam flow.
Rod Controller
The rod controller has features of an automatic rod controller
first proposed to control the reactor-coolant-outlet temperature for
the Pebble Bed Reactor Experiment? and later modified for the Molten-
Salt Reactor Experiment.* The rod controller will be described with
the aid of the diagram shown in the "Rod Controller" portion of 718. 2
and with the following idealized relationships :
set
set
Poet + + * We Pro
+ KW
-Trg,
Bet
.
.
V(g) = + B [[®cet - ] + cent Corpet - 6) + DS (set - ) at] + V,($).
.
.
The 21 control-red drives are actuated by 021-off relays whenever the
output v(a) of the flux controller exceeds the limits of a dead bando
The gain adjustments A, B, C, D, and K are selected to give the best
overall system response.
A feature of the rod controller 18 that a change in the reactor
coolant flow or the reactor inlet temperature produces an immediate
change of the flux set point. This gives a "prediction," or
"anticipation," of the approximate new value of flux needed to produce
the reactor outlet temperature that is required to maintain the desired
stean temperature. The rod controller holds the flux approximately
proportional to the reactor coolant flow during changes in flow, thereby
minimizing the probability of exceeding safe operating limits in the
relationship between flux and flow.
The use of a comprated reactor-outlet-temperature set point
(T ) in this system is an artifice to enable the rod controller
to regulate be steam temperature. This reactor-outlet-temperature set
point, 18 slowly readjusted until the steam temperature reaches the
desired value. The actual numerical value of the reactor outlet
.
set
:
:
temperature is not required to be equal to the value of the reactor
outlet-temperature set point, and in fact, the measured reactor-outlet
temperature is not even used in the control system.
slow drifts or steady-state errors in the sensors associated
with flux, coolant flow, or reactor inlet temperature, or in the
computed reactor-outlet-temperature set point signal will not prevent
the rod controller from maintaining the desired steam temperature.
Also, the actual inequality that always exists between a measured flux
and the product of coolant flow and the differential temperature can-
not prevent the achievement of the desired steam tamperature. However,
these errors, or inequalities, can latroduce small transient dis-
crepancies during maneuvering. Since the maneuvering rate of a large
plant 16 necessarily limited by thermal considerations, these discre-
pancies do not appear to be serious.
.
Blower Controller
The blower controller will be described with the aid of the dia-
gram shown in the "Blower Controllor" portion of 719. 2 and with the
following idealized relationship:
V(P) --> [bout - 7,3 +7 de Post Pg] *5 port • Py] at] + v.(P2)
Bet
The output (P) of the blower controller 18 combined in a signal suming
device for each coolant loop with the output of a speed synchronizer for
that loop. The outputs from the signal summer, adjust the speeds of
the main coolant blovers by changing the variable spoed couplings between
the blowers and the drive motors. The speed synchronisers are wed
.
to equalize the speeds of the two blowers. The gain adjustments ,
T, and G are selected to give the best overall system response,
Control Actions following a load Change
When the turbine load is increased, the speed governor causes the
turbine control valves to open, which ircreases the stean flow and
decreases the steam pressure. The blower controller increases the
réactor coolant flow to return the steam pressure to its set poiat.
The increase in coolant flow produces an immediate proportional increase
in the flux set polat.. The flux control loop in the rod controller then
raises the control rods to increase the flux and, thereby, maintains the
reactor outlet temperature approximately constant. This immediate action
alone does not prevent a slight decrease in stoan temperature or small
changes in reactor outlet and inlet temperatures. These small variations
are corrected on a slower time scale as the rod controller returns the
stean temperatıưre to its set point. By suitable choice of the various
controller characteristics, one can optinise the response of the entire
system.
PERFORMANCE
:
The ability of the control system to make the BOCR plant operate
satisfactorily in the load-following mode described above was demon-
strated on an analog computer model of the BOR. I The study bracketed
the types of changes in stean flow, or lond, that the plant night be
expected to follow and determined the optimum values of the gaino 10
various control loops.
After being optimized, the control system gave excellent performance
in that it made the simulated plant smoothly follow large changes in
load (of about 70% tull stean flow) at the rate of 10$/min. It also .
made the plant follow small, rapid changes in load (of about 10% of
full stean flow) at the rate of 1%/sec. In neither case did any variable
exceed acceptable limits. In both cases there were transient deviations
in stean temperature and steam pressure from their rated values, but
these were acceptably small and were corrected in a reasonably short
time after the turbine load reached its final value.
Large Changes in Turbine. Load
The first maneuvers Lovestigated vere large changes in steam flow
which represent deliberate, or planned, loading or wloading of the
turbine. The maximum cate cf such a change in steam flow, determined
by limitations in the reactor, steam generator, and turbine, 18
about 10%/min. Since the operator will make these large changes in
load by using the load limit mechanism or the speed changer mechanism,
the load variations were simulated by ramp changes in the positions of
the turbine control valves.
Figure 3 shows the response of the plant to a ramp change in the
positions of the turbine control valves from 35 to 100% power conditions
1n 7.2 min. The resulting increase in steam flow, which closely
followed the movement of the control valves, caused a reduction of 20 poi
-
in the steam pressure. The blower controller increased the reactor
.
John
coolant flow until the stean pressure returned to the 1250 poig set
point, and the increased coolant flow produced an immediate proportional
Increase in the flux set point. The flwr control loop in the rod con-
troller then positioned the control rods to maintain the flux within
.
."
.
.
1% of the coolant flow and, thereby, to maintain the reactor outlet
temperature approximately constant. This immediate action alone did
not prevent a decrease of 27°F in the steam temperature; however, this
small variation was corrected on a somewhat slower time scale as the
rod controller returned the steam temperature to the 903°F set point.
During this maneuver, the feedwater temperature was increased as a
function of steam flow with a first-order time delay having a time
constant of 20 min. About 16 min. after the beginning of the maneuver,
the feedwater temperature was manually increased to the value corre-
sponding to 100% steam flow to determine the response of the system to
a change in feedwater temperature. This step increase in feedwater
temperature produced a rapid increase in reactor inlet temperature.
The rod controller reduced the reactor power slightly to compensate for
this increase in reactor inlet temperature and to maintain the steam
temperature at its set point. In most of the rod movements shown in
the plo% of "Rate of Reactivity Addition by Rods," the duration of the
"on" signal from the rod controller was quite short, and the delay
caused by the simulated rod-drive starting characteristics prevented
the achievement of the maximum reactivity addition rate by the bank of
21 rods.
This maneuver represents the maximum rate of change in steam flow
(10%/min) for large changes in load. The magnitudes of the deviations
in steam pressure and temperature would be less in maneuvers made at
lower rates of 3 to 5%/min--rates normally expected to be used in
loading and unloading the turtine.
111
.
Bmall Rapid Changes in Turbine Load
The second series of maneuvers investigated were small rapid
changes in steam flow, such as might be encountered if the plant were
supplying a variable electrical load without the benefit of being
connected in parallel with a large electrical grid. Since the maxi-
mum amplitude of rapid changes in steam flow would depend on the
particular load, we examined several changes from 5 to 20% of full
flow, and simulated them by moving the turbine control valves at a rate
of 1%/sec.
As Illustrated in Fig. 4, the control system made the plant smoothly
follow changes of 5% and 10% of full steam flow and held the steam
pressure and temperature close to their set points. To do this, the.
control system changed the coolant flow and the reactor power rapidly
during the initial 30 sec of the transients however, they did not
deviate from each other by more than a few percent. If these dis- .
turbances had been initiated by changes in turbine load instead of by
simple ramps in the positions of the turbine control valves, the speed
governor would have readjusted the turbine control valves slightly at
the conclusion of the load changes to maintain the value of the steam
flow necessary to regulate the turbine speed.
REASFERENCES
1. 8. J. Ball, Reactor simulator Study, "Gas -Cooled Reactor Program
Semiannual Progress Report, September 30, 1962," UBAEC Report
ORAL-3372, Oak Ridge National Laboratory, pp. 146-159.
13
VL
"
..
.:
-
ri
.
.
.
.
-
-
.
:
.
7
.-
.
: 2. E. N. Fray, s. J. Ditto, and D. N. Fry, Design of MSRE Automatio
Rod Controller, "Instrumentation and Control Division Annual
Progress Report, September 1, 193," USABC Report ORNL-3578,
Oak Ridge National Laboratory, pp. 113-114.
3. H. G. O'Brien and C. 8. Walker, Operator Training and Control
Systems Analyses, "Gas-Cooled Reactor Program Semiamual Progress
Report, March 31, 1965, "USAEC Report ORNL-3807, Oak Ridge
National Laboratory, pp. 277-284.
.
...is
:
:
.
.
.
. .
. .
1. Q. O'Brien
A Control System for a Gas-Cooled Nuclear Power Plant ..
FIGURE LIST
Dwg. or
Photo No.
65-4012R
Caption
Fig. 1. Simplified Flow Diagram of the BGCR.
Fig. 2. Block Diagrams of the Rod and Blower Control Systems.
65-402
65-40LOR
Fig. 3. Response of the plant to a Ramp Change in the Positions
of the Turbine Control Valves from 35 to 100% Power Conditions in 7.2 Min.
654006R
Fig. 4. Response of the plant to Rapid Changes in the Positions
of the Turbine Control Valves About Full Power Conditions.
in
mantenimiento entre connections litt omination
intensiteeti
. .
m
..commencem
i m inare
.
.
:
,
ntenbos
ORNL - DWG 65-4012R

TURBINE VALVE CONT
BY SPEED GOVERNOR
Ws, Ts P
CONTROL
ROD DRIVE
MOTORS
TURBINE
GENERATOR
STEAM
GENERATOR
STEAM
GENERATOR
CONDENSER
REACTOR
CORE
,,Wew
..
.
FEEDWATER
HEATERS
.
DRIVE
MOTOR
VARIABLE
SPEED
COUPLING
MAIN
BLOWER
A
MAIN
BLOWER
VARIABLE
SPEED
COUPLING
DRIVE
MOTOR
FEEDWATER
PUMPS
Simplified Flow Diagram of the EGCR.
DWG 65-4014R2
MULTIPLIER
MULTIPLIER - From Trid
3 Tro
INTEGRATOR
OR
CONTROLLER
WITH RESET
ACTION ONLY
(Tsser-Ts]
*set=* WG [Troms Tri]
RAISE
do
CONTROLLER
WITH
PROPORTIONAL
RATE, AND
RESET ACTION
ON-OFF
RELAYS
TO CONTROL
ROD DRIVE
MOTORS
LOWER
ROD CONTROLLER
LOOP A
CONTROLLER
WITH
PROPORTIONAL,
RATE, AND
RESET ACTION
v (Ps)
Dob0.d.
TO SIGNAL
SUMMERS
CONTROLLING
BLOWER
SPEED


LOOP B
BLOWER CONTROLLER
ES
EN
+ 5x10-3
ORNL-DWG 65-4010R
RATE OF REACTIVITY
ADDITION BY RODS

1 PERCENT PER SECOND
01
-5x103

TOTAL REACTOR POWER
COOLANT
FLOW
PERCENT OF FULL
POWER VALUE
STEAM FLOW
1050
-REACTOR OUTLET TEMPERATURE
IIIIIW

950
STEAM TEMPERATURE
TEMPERATURE
850
550
REACTOR INLET. TEMPERATURE
450
INCREASE FEEDWATER
TEMPERATURE
350
1350

STEAM PRESSURE
PRESSURE (psig)
.
1250
12
4450
0.
200
400
600 800
TIME (sec) ..
9000
1200
rri.
.
..
+ 3x10-3
ONWL-DWG 65-4006R
RATE OF REACTIVITY ADDITION BY
ROOS

PERCENT PER SECOND
-5*103

INC. 5%
DEC. 10%
TOTAL REACTOR POWER
100
PERCENT OF FULL
POWER VALUE
STEAM FLOW
90
-COOLANT
FLOW
• 80

1050
REACTOR OUTLET TEMPERATURE
950
STEAM TEMPERATURE
..
850
TEMPERATURE (°F)
REACTOR INLET TEMPERATURE
450
350
1350

STEAM PRESSURE
PRESSURE (psie)
1250
1150
200
600
400
TIME (sec).
.
.-.
-
-
---
-
-
-
- -
-
-
RE
12/ 1 /65
DATE FILMED
END
.
.
.
.:.:..
.