ML19345G183
| ML19345G183 | |
| Person / Time | |
|---|---|
| Site: | Humboldt Bay |
| Issue date: | 09/23/1975 |
| From: | PACIFIC GAS & ELECTRIC CO. |
| To: | Engelken R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| References | |
| NUDOCS 8103110366 | |
| Download: ML19345G183 (3) | |
Text
..
O I.
e PACIFIC GAS AND E I_,E C T RI C C O M PANY FGWE l
77 B E ALE S T R E ET. S A N F R A N CISCo. C ALIFo R NI A 94106. (415) 781 4211,
.sonn c. uonnieenv
,t.:'. ;*;* *
. m.u., o. u.o ca.us
- n-::_ ; ;-.
- g :;::,n :..
u u.c o t.u n. r u n a u a n September 23,.1975
= ===a"acaa"'
e.
6..
,.v.~
o u..~
- .h':;;r::.
- ~ :::::.
u.6eo6=..u...e66oa att:-! 07:17 **
su:!.'*..rt::
.6....c.....o.
- P.. :::1.
6..>.
v.
- .v.::.*v. ::*-a c :..! c..
..a..e6....
- .::w. :::'-"
=:n ':=.....
g n;,.g....
- =t.:=:... 3.,,,,,,, tr :. =---
Mr. R. H. Engelken, Director Office of Inspection and Enforcement Region V b'(
U. S. Nuclear Regulatory Comission 97
-@Ypcq M.
1990 N. California Boulevard
/4
.h N-i,\\
Walnut Creek Plsza, Suite 202 Q
-ww Walnut Creek, California 94596
@ SE.o 25 gg
.i Ecew i-(EMQ.[
Re: Docket No. 50-133 License No. DPR-7 g
Dear Mr. Engelken:
Q On September 9, 1975, during a hot startup of Humboldt Bay Power Plant Unit No. 3, two spurious reactor trips occurred on the approach to critical. On the second trip it was observed that the control rods took 5-10 seconds to fully insert.
It was also observed that on reactor safety bus No. I the master reacter trip solenoid valves and reactor trip dump tank solenoid valves de-energized immediately, but that the individual reactor trip solenoid valves did not de-energize until after the control rods had fully inserted. On September 10 this incident was reported to your office by members of the plant as an abnormal occurrence per Section IX.H.2.a. (2) of the Technical Specifications. The incident was classified as an abnormal occurrence at that time since the control rod insertion times were greater than the 2.5 second maximum specified in Section V.B.2.c. (4) of the Technical Specifications, and the mechanism causing the slow insertion times was not fully understood. The investigation program described below was undertaken, and the results of this program show that the incident was not an abnormal occurrence.
l An investigation was conducted on September 10 and 11 to determince the cause of these spurious reactor trips and to understand why the reactor control rod trip system and clie control rod drives behaved as they did on the second spurious trip. The cause of the spurious trips was found to be noisy power range neutron monitors (picoammeters). Tests showed that when all three monitors are on their most sensitive scale, called the minimum operating range, a transient in one monitor l
(from a picoammeter alarm, downscale or upscale trip, etc.) would sometimes induce l
a trip signal into one or both of the other picoammeters. Occasionally, as a result of these spurious trips, the 16 volt d-c output voltage of one or both of the reactor safety system logics would drop by 4-8 volts and recover over a period of from 10 to 20 milliseconds. The 16 volt logic output voltage energizes vacuum relays in the power switch units and contacts of these relays control the voltage
[
across the master reactor trip, scram dump tank and individual reactor trip solenoid l
=
r/o3// asa P00R ORIGINAL l
(;
3 Mr. R. H. Engelken 2
September 23, 1975 valves.
It was observed that on these very fast transients, the first and usually the only relays to de-energize in either half of the reactor safety system were those controlling the power to the master reactor trip and reactor trip dump tank solenoid valves. This is exactly the condition observed during the second spurious reactor trip on September 9 and is the only logical explanation for the slower than normal control rod insertion tices. (Four master reactor trip solenoid valves are provided as backup for the 32 individual reactor trip solenoid valves. The master trip valves back up the 32 individual control rod valves and are not designed to
- vent, i.e.,
cause control rod insertion, as rapidly as the individual reactor trip solenoid valves.)
Part of the test program conducted consisted of scram testing each scram accumulator and its associated control rod drives to demonstrate that the drives would insert within 2.5 seconds when the individual reactor trip solenoids were de-energized. All drives performed properly sn these tests. Also, the reactor safety bus No. I logic was bench tested and all co=ponents were found to be satie-factory. As a final check, all relevant operational tests normally conducted on the reactor trip system were performed to functionally test the system.
The conclusions reached as a result of the tests and investigations were:
1.
The spurious reactor trips occurred as a result of picom=neter noise, and this is only a' problem when the range switches are on thei: most sensitive range.
2.
The slow scram time, during the second spurious trip, resulted from the 32 individual reactor trip valves having to vent through the four master trip solenoid valves. This condition existed because the relays controlling power to the individual reactor trip solenoid valves did not drop out before the logic output voltage transient was over and the logic voltage returned to normal.
3.
Neither 1. nor 2. above represent a safety problem. Every test performed proved that a normal reactor trip would have occurred whenever any reactor safety system sensor operated.
4.
Neither 1. nor 2. represent a reliability problem. Once the reactor is made critical and reaches heating power level picoa= meter noise ceases to be a problem. This, was demonstrated by 19 separate trip tests on each picoam=cter when heating power was reached during the September 11 startup.
5.
Neither 1. nor 2. constitutes an abnormal occurrence. Picoammeter noise is an operational nuisance only when the reactor is suberitical.
The reactor was satisfactorily started up on September 11.
On September 12 the problems and their apparent causes were discussed with the General Electric Company instrument personnel who were in agreement with the above analysis of the P00R ORIGINAL
~
1 Mr. R. H. Engelken 3
September 23, 1975 e
reactor trip and slow rod insertion times. They stated that this condition has been observed by at least one other BWR plant with instrumentation similar to Humboldt. They further indicated that they will make recoc:mendations for reducing the picoa= meter noise.
Very truly yours, CC:
Mr. Karl R. Goller
' Assistant Direc.or for e
Operating Reactors Division of Reactor Licensing U. S. Nuclear Regulatory Cocnission Washington, D. C. 20555 l
l P00RBRIBlHR
- ~. -....
w
,--