Information Notice No. 84-55: Seal Table Leaks at PWRs

                                                                SSINS: 6835 
                                                                IN 84-55   

                                UNITED STATES
                        NUCLEAR REGULATORY COMMISSION
                    OFFICE OF INSPECTION AND ENFORCEMENT
                            WASHINGTON, DC 20555

                                July 6, 1984

Information Notice No. 84-55:   SEAL TABLE LEAKS AT PWRs 

Addressees: 

All power reactor facilities holding an operating license (OL) or 
construction permit (CP). 

Purpose: 

This information notice is provided as notification of a potentially generic
problem involving reactor coolant leaks from incore probe seal tables. The 
leaks have occurred during maintenance on these systems while the reactor 
coolant system was at an elevated pressure and temperature. The events 
described have resulted in personnel hazards and also degraded plant safety.
It is expected that recipients will review the information for applicability
to their facilities and consider actions, if appropriate, to preclude a 
similar problem occurring at their facilities. However, suggestions 
contained in this information notice do not constitute requirements and, 
therefore, no specific action or written response is required. 

Description of Circumstances: 

On January 20, 1984 (reported by LER No. 50-295/1984-005), a reactor coolant
leak was observed in the seal table room at Zion Generating Station Unit 1. 
The unit was in hot shutdown with a plant heatup in progress. The reactor 
coolant system temperature was 445F and pressure was 2,235 psi. 
Inspection of the seal table by plant personnel revealed that a leak was 
located at a point where the high-pressure seal mates to the conduit for 
incore thimble E-11. An attempt to repair the leak was made when the system 
pressure was reduced to 1,000 psi. These efforts reduced but did not stop 
the leak. The system pres sure and temperature were reduced to 400 psi and 
370F and another attempt to repair the leak was made. The repairmen 
noticed a slight bowing between the high-pressure seal and the thimble 
isolation valve. It was believed that this bowing caused the Swagelok 
fitting to be improperly seated, thus causing the leak. To correct the 
problem, two bolts holding the isolation valve to the valve bracket were 
removed to allow straightening of the thimble tube. However, the two bolts 
and bracket were the primary support devices holding the fitting in place. 
When they were removed, the fitting broke loose causing an unisolatable 
reactor coolant leak of approximately 18 gpm to the containment. The area 
was immediately evacuated. Later, upon examination of the fittings, it was 
found that the ferrules on all but seven of the thimbles had moved 1/32" to 
3/8" up from their original position toward the edge of the conduits. 



8407050274 
.


                                                               IN 84-55    
                                                               July 6, 1984 
                                                               Page 2 of 3 

Review of the procedure for assembly of the high-pressure and low-pressure 
seals within the Swagelok fittings revealed that the low-pressure fittings 
could pull up the ferrules causing improper fitting of the high-pressure 
seals. This is believed to have caused the initial leak. Overtorquing of the
fittings during the initial attempt to correct the leak probably 
overstressed the ferrule and allowed it to break loose when the restraint 
was removed. 

Another event occurred at the Sequoyah Nuclear Plant Unit 1 (reported in LER
SQRD-50-327/84030). On April 19, 1984 workmen were in the seal table room 
brush cleaning the incore probe thimble guide tubes. The reactor was 
critical at 30% power and normal operating temperature and pressure. 

The thimble guide cleaning procedure had made it necessary to disconnect the
thimble tubes leading from the seal table to the 10 path selection device 
such that the high-pressure Swagelok fitting at the Seal Table was the only 
device restraining the thimble. Because it was not known if the cleaning 
brush was being fully inserted during the cleaning process, an unblocked 
tube was used to obtain information on brush travel. The cleaning assembly 
was installed and inserted 15 ft. into this tube before a shift change 
occurred. Following the shift change personnel again began inserting the 
brush. On the 78th turn (10 inch insertion per turn) the tool handler noted 
that an increased pressure was needed to turn the crank. On the 79th turn 
(approximately 80.8 ft. into the tube) the water was noticed on the seal 
table. 

Water was noticed on the seal table around the Swagelok fitting holding the 
guide tube that was being cleaned. This prompted an immediate evacuation of 
the seal table room. Subsequently, the fitting broke loose ejecting the 
entire thimble tube and cleaning equipment from the core. An unisolatable 
reactor coolant leak ensued; the leak rate during this time was 
approximately 30 gpm. The leak continued for approximately 11 hours until 
the reactor was cooled down with the pressure reduced to nitrogen blanket 
pressure and the reactor water level was reduced to a level below the seal 
table. A total of approximately 16,000 gallons of reactor coolant leaked 
into the containment. 

Radiation surveys conducted the following day indicated 2-3 rem at the 
entrance to the seal table room, 200-300 rem at the end of the tube near the
seal table and greater than 1,000 rem in the center of the ejected tube. The
tip of the thimble tube was reading approximately 4,000 rem. 

A very difficult cleanup operation followed as well as an elaborate failure 
evaluation and analysis which included mock-up tests, with identical 
fittings, tubing, pressures and forces applied to the thimble tubing. The 
tests simulated as near as practical the events at the time of the incident.

Based on the results of these tests, TVA concluded that the separation of 
the failed assembly appears to be the result of extraordinary and 
unanticipated loads on the assembly, caused by the cleaning fixture that was 
being used when the separation occurred. The assembly mock-up testing 
demonstrated that strains of considerable magnitude resulted from applied 
forces on the manually operated fixture crank handle. 

As a result of the tests, TVA recommends that the use of this tube cleaning 
fixture in its present form be discontinued. If cleaning is required, 
.

                                                               IN 84-55 
                                                               July 6, 1984 
                                                               Page 3 of 3 

modifications to this fixture or other techniques should be employed that do
not transmit bending forces through the fittings. 

Discussion: 

Even though the above incidents appear to be caused by different 
circumstances, both events point out the need for adequate controls and 
precautions to ensure personnel and plant safety while performing 
maintenance on high-pressure systems, especially activities involving the 
seal table. Both of these events occurred with the reactor at elevated 
pressures and temperatures, and in the case of Sequoyah the reactor was at 
30% reactor power. In both cases maintenance was conducted on a 
high-pressure system with what was equivalent to single valve protection. 
For both personnel and plant safety considerations, maintenance is not 
normally conducted on high-pressure systems while at high-pressure and 
temperature and with only single valve protection. To preclude the type of 
events just described from occurring, every effort should be made to 
schedule seal table maintenance during cold shutdown conditions. Also, the 
need for maintenance of any system under hot, pressurized conditions should 
be thoroughly evaluated before allowing personnel to perform the work. 
Licensees are urged to review their maintenance procedures to ensure that 
maintenance under these conditions is minimized. 

No one was injured during these events and the operators brought the plants 
to cold shutdown without undue problems. However, both of these events have 
caused problems associated with the radiological cleanup efforts. In the 
case of Sequoyah, a highly radioactive piece of equipment was ejected from 
the core. This required that extraordinary measures be taken during the 
decontamination of the room and the removal of the thimble. In both events, 
decontamination of the room was required. Increased personnel exposure and 
down time of the plant due to the cleanup and repair efforts provide 
additional incentives for precautions against maintenance under similar 
conditions. 

No written response to this notice is required; however, licensees should 
review the information contained in this notice for applicability to their 
facilities, especially if plants have performed maintenance on the incore 
probe guide tubes under similar plant conditions. 

If you have any questions regarding this matter, please contact the Regional
Administrator of the appropriate NRC regional office or the Technical 
Contact listed in this notice. 


                                   Edward L. Jordan Director 
                                   Division of Emergency Preparedness 
                                     and Engineering Response 
                                   Office of Inspection and Enforcement 

Technical Contact:  D. R. Powell, IE
                    (301) 492-7155

Attachment:
List of Recently Issued IE Information Notices