ORNL-4191 - the Molten Salt Energy Technologies Web Site
ORNL-4191 - the Molten Salt Energy Technologies Web Site
ORNL-4191 - the Molten Salt Energy Technologies Web Site
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
trouble proved to be a failure in a glass seal that<br />
allowed water to enter <strong>the</strong> magnesia insulation in<br />
<strong>the</strong> cable, which is an integral part of <strong>the</strong> chamber.<br />
Safety System<br />
A period safety amplifier failed when lightning<br />
struck <strong>the</strong> power line to <strong>the</strong> reactor site, and a<br />
replacement amplifier failed as it was being in-<br />
stalled. The field-effect transistor in this type of<br />
amplifier is susceptible to damage by transient<br />
voltages, and it was found that under some condi-<br />
tions, damaging transients could be produced when<br />
<strong>the</strong> amplifier is removed from or inserted into <strong>the</strong><br />
system. A protective circuit was designed, tested,<br />
and installed 011 <strong>the</strong> spare unit; in <strong>the</strong> interim, a<br />
different and more stable field-effect transistor was<br />
installed. The module replacement procedure has<br />
been modified to reduce <strong>the</strong> possibility of damage<br />
incurred on installation of <strong>the</strong> module.<br />
Experience at o<strong>the</strong>r sites with <strong>the</strong> type of flux<br />
safety amplifiers used in <strong>the</strong> MSKE had shown that<br />
<strong>the</strong> input transistor could be damaged, causing<br />
erroneous readings, if <strong>the</strong> input signal became too<br />
large too rapidly. Therefore, a protective network<br />
of a resistor and two diodes was added to <strong>the</strong> input<br />
circuit nf each of <strong>the</strong>se amplifiers.<br />
Two relays in <strong>the</strong> safety relay matrices failed,<br />
both with open coil circuits. A chattering contact<br />
on <strong>the</strong> fuel pump motor current relay caused safety<br />
channel 2 to trip several times before <strong>the</strong> problem<br />
was overcome by paralleling two contacts on <strong>the</strong><br />
same relay. A defective switch on <strong>the</strong> core outlet<br />
temperature also caused several channel trips and<br />
one reactor scram before <strong>the</strong> trouble was identified<br />
and <strong>the</strong> switch was replaced<br />
Four rod scrams, all spurious, occurred during<br />
operation in <strong>the</strong> report period. Two were caused<br />
by general power failures during electrical storms.<br />
Anc<strong>the</strong>r occurred during a routine test of <strong>the</strong> safety<br />
system when an operator accidentally failed to reset<br />
a tripped channel before tripping a second channel.<br />
:?<br />
1 he o<strong>the</strong>r scram, also during a routine test, came<br />
when a spurious signal from <strong>the</strong> switch on <strong>the</strong> core<br />
outlet temperature tripped a channel while ano<strong>the</strong>r<br />
channel was tripped by <strong>the</strong> test.<br />
TWBL SYSTEM DES3GN<br />
P. G. Herndon<br />
As experience showed <strong>the</strong> need or desirability of<br />
more information for <strong>the</strong> operators, improved per-<br />
48<br />
formance, or increased protection, <strong>the</strong> instrumentation<br />
and controls systems were modified or added to.<br />
During <strong>the</strong> report period <strong>the</strong>re were 25 design<br />
change requests directly involving instruments or<br />
controls. Six of <strong>the</strong>se required only changes in<br />
process switch operating set points. Fourteeri<br />
requests resulted in changes in instruments or<br />
controls, one was canceled, and <strong>the</strong> remaining four<br />
were not completed. The more important changes<br />
are described below.<br />
The single T32-v dc power supply formerly sewing<br />
<strong>the</strong> nuclear safety and controls instrumentation<br />
was replaced with three independent power supplies,<br />
one for each safety channel. Each obtains +32-v dc<br />
from 110-v ac, but <strong>the</strong> ac power sources arc different.<br />
One is <strong>the</strong> normal building ac system.<br />
Ano<strong>the</strong>r is 110-v ac from a 50-kw static inverter<br />
powered by <strong>the</strong> 250-v dc system. The third comes<br />
from a 1-kw inverter operating on <strong>the</strong> 48-v dc system.<br />
Thus continuity of control circuit operation is<br />
ensured in <strong>the</strong> event of a single power supply or<br />
power source failure. It also increased <strong>the</strong> reliability<br />
of <strong>the</strong> protection afforded by <strong>the</strong> safety<br />
system by ruling out <strong>the</strong> possibility that malfunction<br />
of a single voltage-regulating circuit could<br />
compromise more than one channel. (Before <strong>the</strong><br />
change, on one occasion, a failure in part of <strong>the</strong><br />
regulating circuit caused its voltage output to increase<br />
from 32 to SO v, and only a second regulator<br />
in series prevented this increase from being imposed<br />
on all <strong>the</strong> safety circuits.)<br />
To prevent <strong>the</strong> reactor from dropping out of <strong>the</strong><br />
“Run” control mode when a single nuclear safety<br />
channel is de-energized, <strong>the</strong> “nuclear sag bypass”<br />
interlocks were changed from three series-connected<br />
contacts to a two-of-three matrix.<br />
Circuits were installed to annunciate loss of<br />
power to <strong>the</strong> control rod drive circuits. This reminds<br />
<strong>the</strong> operator that he cannot imrnediately return<br />
to power simply by manually withdrawing <strong>the</strong><br />
rods after <strong>the</strong> controls have dropped out of <strong>the</strong><br />
“Run” inode.<br />
A wiring error in a safety circuit was discovered<br />
and corrected. Interlocks had recently been added<br />
in <strong>the</strong> “load scram” channels to drop <strong>the</strong> load wheri<br />
<strong>the</strong> control rods scram. A wiring design error resulted<br />
in <strong>the</strong>se interlocks being bypassed by a<br />
safety jumper. Although <strong>the</strong> circuits were wired<br />
this way for a <strong>the</strong> before being discovered, <strong>the</strong><br />
scram interlocks were always operative during<br />
power operation, since <strong>the</strong> reactor cannot go into<br />
<strong>the</strong> “Operate” mode when any safety jumper is<br />
inserted.