ML19338D158
| ML19338D158 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 08/18/1980 |
| From: | Michelson C NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8009190753 | |
| Download: ML19338D158 (3) | |
Text
.
,.ey ya c ppe
.d.~T c
n UNITED STATES -
7 i
' NL' CLEAR REGULATORY COMMISSION Q';; g, ]5i/c j.
3 WASHINGTON. D. C. 20555 D /
. AUG 18.1980 ms 1
. MEMORANDUM'FOR:
Harold R'. Denton, Director Office of Nuclear Reactor Regulation FROM:
Carlyle.Michelson, Director Office for Analysis and Evaluation of Operational Data u
~
SUBJECT:
POTENTIAL FOR UNACCEPTABLE INTERACTION BETWEEN THE j
CONTROL ROD-CRIVE SYSTEM AND NON-ESSENTIAL CONTROL AIR SYSTEM AT THE BROWNS FERRY NUCLEAR PLANT lSince completing its analysis of the Browns Ferry 3 partial failure to scram event,LAE00 has be.2n taking a closer lock at the added (temporary) scram dis-charge volume instrument arrangement in terms of its acceptability for con-tinued operation pending completion of the recommended system modifications.
Our evaluation is still ongoing, but-one established conclusion is that an imediate, in-depth' evaluation is needed of the potential for unacceptable interaction between the control rod drive system and the non-essential (non-
)
safety)- control air system.
l At the recent meeting with General Electric relating to.the Browns Ferry event, we questioned GE concerning the scram inlet and_ outlet valves and how they micht
-i respond-to a-slow loss of control air pressure. Their answer indic.ited that these valves would drift open' slowly, but without early indication, as the ai-i cressure decreased. ~In other words, the loss of air pressure would lead to a
'significant' scram discharge volume inleakage, but the control rods might not move until the air pressure-decreased substantially (with even greater inleakace).
The immediate concern is associated with this degraded air pressure situation and is focused on the scram discharge volume fill rate, the time for operatcr acticn, and the alarms and indications to guide his actions.- We are already aware that the scram discharge volume drain rate is less than the scram instrument volume drain rate so.any substantial inleakage may go undetecteo by the existing (permanent) alarm or scram instruments. Further, there is no positien indication for tne scran. valves, other than full-open, nor is there direct <.; cram protection on partial loss of control air.
In fact, the alarm signaling low air pressure would undoubt-
'edly cause the operator's attention to be distracted from the control rod drive
_ system.
l800919o 3 ]-
f fa
_ *. i-
e y
Harold R.1Denton 2-Thus, f ailures ini the non-essential control air system which result in a -
degraded air pressure could. result in a significant and undetected increase in the'. scram discharge volume inleakage. :In fact, an inleakage rate of 3 gpm per drive may exist before control rod movement is experienced. With
.this flow rate, the operator may have only a few minutes before the scram
~
volume.has filled to such an-extent that the scram function could be adversely affected,-and eventually even prevented.
We believe the one alarm which
- might ~ actuate. (control rod drive high temperature alarm) would not be a sufficiently timely;or. good indication-of what is happening to assure proper
-operator action.
An event ~ related to this concern occurred at Browns Ferry on August 18, 1978
. hen a massive loss of control air pressure to the entire plant was experienced w
(PNO-78-147). Units 1 and 2 were operating at full power. Their-control rods
=
drifted inward as a result of low air pressure. Both units were manually (and suc.cessfully) scramed in accordance with emergency procedures. This might have been a close call since the scram discharge volume was certainly
'j filling e 'he rods drifted in before the manual scram.
The actual' rate of control air dep.essurization and the timely manual scram might have been the i
saving factors.
-1 It appears essential that we obtain an adequate understanding of the control rod drive unit response to a loss'of' control air pressure.
It is an imediate -safety concern if the degradation of the nonsafety-related control-air system can lead -
to a loss of scram capability before the situation is diagnosed by the operator and manual scram achieved.
It should be apparent that the situation would be precarious if a loss of control air pressure lead _to an inability to scram from full. power and, at the same time, any significant portion of the full power heat J
-l removal capability.(turbogenerator, main steam or feedwater systems) was lost j
for any reason such as~the pressure reduction.cr its consequences. Recirculation pump trip, where available, could be timely and helpful, but not adequate. Boron i
injection would probably be too slow.
In any case, this is an unacceptable i
interaction between.a safety and nonsafety system.
j The licensed design. basis for Unit 3 was redundant level sensing capability, and appropriate provision for scram initiation, to protect against excessive water in the scram discharge volume. The AE00 recomended modifications contained in our report on.the BF-3 partial failure to scram would restore this design. basis and would protect against degraded air system pressures (provided the control rod drives are not adversely affected by any heatup prior to scram actuation).
HoEver, Unit 3isbeingallowedtooperateona'temporarybasiswithasingle ultrasonic detector to indicate accumulated inleakage to the scram discharge volume; This single detector does not indicate or alarm in the control room and provides no scram initiation.- It is read locally by a roving operator every en-a L
~
3 Harold R.-Denton ~
30 minutes. We believe that'these temporary scram discharge volume level monitoring measures do not adequately protect against a loss of scram function within a few minutes due to degraded air pressure.
a In summary, we recommend.that the margin of safety inherent in the licensed design basis should be promptly restored, although improved temporary measures and arrangements may be acceptable until permanent modifications can be completed. For example, improved temporary measures might be redundant
,j level sensors on the scram discharge volume with control room alarms and E
readout, or a dedicated operator might be used locally with adequate control room communication and plant procedures covering the manual scram function.
The adequacy of any modifications, temporary or permanent, should be confirmed by appropriate-analysis, testing, and onsite evaluation.
1 Althcugh our interim assessment and recommendation applies only to Browns Ferry 3, it is likely that it also applies to the other units at Browns Ferry
-l and to many other BWRs. We will su~cmit a full report for your consideration when our evaluation work is completed.
'b n k
t
(
Carlyle Michelson, Director Office for Analysis and Evaluation of Operatienal Data cc:
C. Berlinger R. Bernero i
E. Jordan W. Dircks J. Shea V.'Stello D. Eisenhut N. Moseley l
E e
--