ML19208C052
| ML19208C052 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/17/1979 |
| From: | James Anderson, Mihalczo J, Schultz M Metropolitan Edison Co, Oak Ridge |
| To: | |
| Shared Package | |
| ML19208C045 | List: |
| References | |
| NUDOCS 7909240465 | |
| Download: ML19208C052 (3) | |
Text
..., '
s t
Three !!ile Island - 2 Technical Support t.
High Count Rate on BF. Startup Detectors a
F i
t J. L. Anderson, et. al.
I I,
Prebica:
Ex-vessel startup detect >rs (BF Proportional Counters) are indicating a
6 tiraes the normal shutdown counting rate. There is concern that if ao: :t L
this counting rate is real, the core may be considerably more reactive b.
than it should be in the present highly berated, fully shut lown condition.
(
W. nced to determine the reasons for the high readings.
L P
Des cription of '..'ork Perf 0:.cf.:
A considerable effort was required to establish the validity of l
1 the information received from TMI; to examine the geometry and characteristics of the installation as designed and as it portains to reactivity; to postulate and evaluate mechanism w'tich could explain the anomalous indications. We became convinced that 1) The detectors were undamaged, i
functional, and probably accurate; 2) 1][ the high readings represented P
I core cultiplication, then the reactor was dangerously near critical and further cooldown would result in criticality and higher core power.
The detector indications were followed carefully to determine trends and to I
attempt to identify the source of the high readings. This turned out to
{
be difficult, because there was considerable confusion about how the data j
from these detectors was recorded by the reactor operators.
This confusion persists, because of an apparent lack of understanding of the detector and amplifier characteristics by the operating crew, and their procedures are of questionable validity.
In spite of this uncertainty, it was determined i
7909240 gf e-a
~ *
- 2 i
j; that the counting rate displayed by the startup channels was following l
anticipated gamma decay curves. This implies that the neutrons detected i
were generated by gamma-neutron reactions in water and other materials and were not necessarily related to core reactivity.
Further, changes which 5
[
should have affected core reactivity such as boron concentration and i
[
temperature did not have a corresponding effect on the counting rate This
(
adds further weight to the idea that the neutrons seen by the detectors are e
t not core-couplud.
P k
t Conclusions and Results:
{
k One candidate to explain higher core reactivity is the redistribution g
of burnabic poisons from the reactor core.
These burnable poisons are
- f; made of Boron Carbide clad in :ircalloy. Some of the Zircalloy is undoubtedly oxidir.ud and the Boron Carbide may be redistributed.
Ilowever, there is no l
positive indication that the core reactivity is higher than normal.
Two more likely explanat ons are; 1) The neutron source containers are oxidi:cd i
and the neutron source s are distributed throughout the reactor vessel, perhaps in closer pr;ximity to the startup detectors; 2) Because of fuel i
cladding failure, the distributed gamma activity level is considerably
(
[
higher than normal and may be located much closer to the detectors yielding i
t
[
more localized gamma-neutron reactions and higher counting rates.
Significant gamma or neutron sources external to the reactor vessel in the vicinity of the detectors are unlikely.
It is concluded that a significant core reactivity l
problem does not exist.
t Work Requested By:
NRC-Bethesda, NRC-Ut1 (Beltracchi; Stello, Ackerman).
, - ~. -
i 3
i Results Reported To:
I.
NRC-Bethesda (Buhl); NRC-D!I (Ackerman). April 20, 1979.
5 f
Work Performed By.
t t
J. T. Mihale:o, J. L. Anderson, M. A. Schult:
I (Th!I-I AG).
4
}
Estimated Costs:
3 >!andays; $1,000.
e t
i ii r,
l ll
,\\
I
,t
.I h
P00R DMINAL 4