ML20236K324

Forwards Draft Encl 3 Writeup on Loss of DHR at Facility as Discussed in 860425 Memo from J Martins to J Heltemes for Inclusion in Rept for First Quarter CY86 AO Rept

ML20236K324
Person / Time
Site:	San Onofre, 05000000
Issue date:	04/29/1986
From:	Johnson P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:	Bobe P NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
Shared Package
ML20236K265	List: ML20236K262 ML20236K277 ML20236K280 ML20236K289 ML20236K297 ML20236K324 ML20236K333 ML20236K337 ML20236K357 ML20236K380 ML20236K419 ML20236K427 ML20236K532 ML20236K546 ML20236K557 ML20236K576 ML20236K594 ML20236K601 ML20236K616 ML20236K623 ML20236K629 ML20236K641 ML20236K653 ML20236K661 ML20236K664 ML20236K677 ML20236K682 ML20236K695 ML20236K716 ML20236K723 ML20236K727
References
FOIA-87-377 NUDOCS 8708070079
Download: ML20236K324 (5)

Text

flo uq#o, UNITED STATES

,. .' ' ~,

g NUCLEAR REGULATORY COMMISSION i 3 zE WASHINGTON, D. C. 20S55 -

b i 4*** +

I i

5520 TRANSMITTAL APRIL 29, 1986 MEMORANDUM FOR: Paul Bobe, AE0D J

l FROM: P. H. Johnson, Chief l

l Reactor Projects Section 3, RV @a l

SUBJECT:

POSSIBLE ENCLOSURE (3) REPORT FOR THE FIRST QUARTER CY 1986 ABNORMAL OCCURRENCE REPORT l

Enclosed is a draft Enclosure (3) writeup on the loss of decay heat removal at San Onofre 2, as discussed in J. Martin's memorandum to I i

J. Heltemes, dated April 25, 1986. )

i I

P. H. Johnson, Chief )

Reactor Projects Section 3, RV s

Enclosure:

(a/s) cc: F. R. Huey P. Narbut I

i

\

8708070079 870804 PDR FOIA PDR ,

h GCRDON87-377

4 POSSIBLE ENCLOSURE (3) ITEM LOSS OF SHUTDOWN COOLING l

l Date and Place:

On March 26, 1986, Southern California Edison Company experienced a loss l of shutdown cooling at San Onofre Nuclear Generating Station, Unit 2.  ;

The unit had been shutdown for a refueling outage since March 15, 1986, I and therefore still had substantial fission product decay heat.

While attempting to lower the reactor vessel level by a few inches, suction to the shutdown cooling pump was lost due to tae  ;

entrapment of air in the pump. The loss of shutdown cooling allowed the l reactor coolant water to heat up to the point where local boiling was experienced. Shutdown cooling was restored in about one hour and no fuel  !

damage was experienced. The significance of this event is that, had '

shutdown cooling not been restored within a few additional hours, fuel damage might have been experienced.

San Onofre Unit 2 utilizes a pressurized water reactor designed by Combustion Engineering. The plant is located south of San Clemente, California.

I Nature and Probable Consequences On March 26, 1986, at 9:50pm(PST), the plant was in a shutdown condition for a refueling outage, which had commenced eleven days earlier, when the shutdown cooling system, which removes decay heat from the reactor core, became inoperable due to the entrapment of air in the shutdown cooling pump. The entrapped air had been pulled into the pump from the shutdown i cooling suction line which connects to the bottom of the reactor coolant system hot leg discharge piping a few feet from the reactor vessel. At the time of the event, the reactor coolant system was depressurized and vented to atmosphere and the reactor vessel water level had been '

purposely drained down to an intended level of the middle of the hot leg.

This condition provides a situation in which the horizontal reactor coolant system piping hot leg is half full of water and half full of air.

This condition provides adequate flow path and suction head for the shutdown cooling pumps, with water levels sufficiently low to permit removal of steam generator primary side access covers for tube inspections. Slightly lower levels are known, from previous experience, to be susceptible to creating an air / water vortex which can pull air into the shutdown cooling suction Ifne at the bottom of the reactor coolant hot leg, possibly binding the pump with air. Because of this susceptibility, the reactor water level and the shutdown cooling flow rate need to be carefully controlled when in this condition.

l The primary cause of this event was faulty indication of the reactor vessel water level. The level was, in fact, ten inches lower than believed. '

The operators had just lowered the water level a few more inches in response to a maintenance request and had reduced shutdown cooling flow to avoid vortexing, when the motor amperage on the shutdown cooling pump began to take large swings indicating air entrapment. The reactor  !

coolant temperature was ll5*F at this time. The operators waited three l minutes and restarted the pump. In a few minutes, the same amperage l swings recurred and the pump was stopped. The standby pump was started with similar results. l The operators then followed their abnormal operating instruction and dispatched equipment operators to vent the shutdown cooling piping outside the containment penetration area.

One hour and five minutes after the pump had originally been stopped, the I venting was completed and the pump was restarted. The reactor coolant system hot leg temperature increased momentarily, during this period, to 210 F, and then dropped tc below 200'F. However, the temperature indicators are located above the middle of the hot leg and therefore were not immersed in water and may not have responded in a timely way.

At this time, the ventilation system in the fuel handling building j automatically isolated due to high count rate on radiation monitors. '

This was due to the gas previously vented from the shutdown cooling pump, having been drawn from the pump area into the fuel handling building. This was a minor occurrence and no significant amounts of airborne activity were involved in this event, l Cause or Causes The primary causes of this event were inadequate level indication i equipment, in installation, calibration, and operation; inadequate shift turnover; insufficier,t operator sensitivity to the importance of reactor water level control; inadequate procedures; and inadequate training. Additionally, a contr'buting factor was the relative ease of inducing vortexing and subsequent false level indications inherent in the unit's design.

The licensee had installed a new reactor vessel water level indication.

system for this refueling. The system consisted of two electrical level indicators, a wide range and a narrow range indicator. Apparent problems with the system were noted on March 19, 1986, during initial draining of the pressurizer. The problems required recalibration, apparently due to design data errors. Other level accuracy problems were observed during the following few days and on March 22, 1986, the licensee installed the level indication system used during the previous refueling, a tygon tube standpipe.

(

l l

t NOTE:

Subsequent to the event, it was determined that the tygon tube had been mounted on an old structural stanchion which had reactor water levels marked with a felt tip pen rather than on an engineered scale installed for that purpose. Secondly, the tygon tube was installed with an air bubble. The scale used was in error by two inches. The air bubble's effect on accuracy varied dependent upon the bubble's location.

The reactor vessel level indicating system used during refueling was not designated a safety-related system by the licensee and consequently did not receive the level of attention and control warranted.

On the day of the event, the day shift had in fact commenced drain-down, noted that the tygon and electrical level indications were not in agreement, and had noticed the shutdown cooling pump amperage start to oscillate (indicating vortexing). The day-shift immediately added water until indicatior s of air entrapment ceased. These occurrences were not communicated to the oncoming shift, the shift that later experienced the loss of shutdown cooling.

Actions taken to prevent recurrence Licensee -

The licensee has taken certain specific actions to prevent recurrence which include establishing a more reliable indication of water level.

This was to be done by classifying the systems as safety-related 1 and applying the more stringent associated administrative controls; by controlling the configuration of the level indicating systems with more detailed design documentation; by improving procedures for installation; )

and by establishing criteria for verification of operability. l Additionally, the licensee is reconsidering a design change to make the l shutdown cooling pump self-venting. Consideration is being given to j l developing a correlation of shutdown cooling flow and reactor water level to identify vortexing regions. Further, consideration is being given to l

I extensive operator training concerning shutdown cooling system operation i and to operational procedure improvements. Implementation of several of '

, these improvements and a significant increase in management and operator l control were observed during a subsequent drain-down to reestablish  !

mid-loop conditions on April 22, 1986.

( l l

l

l

. l NRC Upon being notified of the event, the resident inspectors commenced an examination of the event including root causes. Their examination and documentation of findings are expected to be completed in Nay 1986.

It should be noted that NRC efforts to address shutdown decay heat j removal requirements are addressed in unresolved safety issue A-45 and ,

that AE00 Case Study Report C503, issued in December 1985, provides some pertinent analysis and recommendations regarding decay heat removal problems at U.S. pressurized water reactors.

i l

l l

l 1

{

l l

\

1

)

\

--