Bulletin 88-08: Supplement 3, Thermal Stresses in Piping Connected to Reactor Coolant Systems

                                                  OMB No.: 3150-0011
                                                  NRCB 88-08, Supplement 3


                                  UNITED STATES
                          NUCLEAR REGULATORY COMMISSION
                      OFFICE OF NUCLEAR REACTOR REGULATION
                             WASHINGTON, D.C.  20555

                                 April 11, 1989


NRC BULLETIN NO. 88-08, SUPPLEMENT 3:  THERMAL STRESSES IN PIPING CONNECTED
                                       TO REACTOR COOLANT SYSTEMS


Addressees:

All holders of operating licenses or construction permits for light-water-
cooled nuclear power reactors.

Purpose:

The purpose of this supplement is to (1) provide information to addressees 
about an event at a foreign reactor relating to thermal stratification of 
unisolable piping connected to the reactor coolant system (RCS) similar to the 
December 9, 1987, Farley 2 event, (2) emphasize to the addressees the need for 
sufficient review of their RCSs to identify any connected, unisolable piping 
that could be subjected to unacceptable thermal stratification, and (3) again 
emphasize the importance of addressees taking action, where such piping is 
identified, to ensure that the piping will not be subjected to unacceptable 
thermal stresses.  No new requirements are included in this supplement.

Description of Circumstances:

On June 6, 1988, while a foreign reactor plant was operating at 100% power, an 
abnormally high flow rate to the containment sump was detected at about 0.2 
gallon per minute (gpm).  The source of the leakage was a circumferential 
crack extending through the wall of an unisolable section of the residual heat 
removal (RHR) piping that is connected to the hot leg of loop A in the RCS.  
The crack was 3.8 inches in length on the pipe inner surface and was located 
on the top of a pipe-to-elbow weld in the horizontal pipe section upstream of 
the first isolation valve of the RHR suction line as shown in Figure 1.  The 
8-inch-diameter (schedule 140, 0.8 inch thick) RHR line was fabricated from 
316 austenitic stainless steel and was insulated.  The crack initiated at the 
weld-metal-to-base-metal interface on the elbow side and propagated through 
the weld metal.  No material or welding defects were found.  Further 
examination revealed another circumferential crack, which was 4.5 inches in 
length on the inner surface of the pipe and 60% through-wall, located on the 
top of the pipe-to-valve weld in the weld metal on the pipe side.





8904050515
.                                                  NRCB 88-08, Supplement 3
                                                  April 11, 1989
                                                  Page 2 of 3


The motor-operated wedge-type RHR isolation valve was normally closed.  The 
fluid in the RHR piping upstream of the valve was stagnant and was cool com-
pared to the fluid in the RCS primary loop.  There was no leakage from 
upstream to downstream of the valve.  However, the packing gland had been 
leaking as evidenced by the wet and rusted valve leak-off piping.  Subsequent 
examination of the surface of the lantern rings in the valve packing revealed 
markings which confirmed leakage from the packing gland.  Although the 
temperature of the leak-off piping was monitored, elevated temperature was not 
detected because the temperature sensor and the leak-off line were not 
insulated and were cooled by the environment.  The crack resulted from thermal 
fatigue caused by hot water, which was drawn periodically from the RCS hot 
leg, leaking through the packing gland of the RHR valve.  The hot fluid flowed 
on top of the cool fluid in the pipe and produced a temperature difference 
between the top and bottom of the pipe resulting in thermal stresses on the 
pipe.  

The thermal stresses were cyclic because the subject RHR valve permitted the 
following sequence of events as shown in Figure 2.  The stagnant fluid 
upstream of the RHR isolation valve cooled due to heat loss to the 
environment.  The cool fluid resulted in the thermal contraction of the valve 
disk.  The upstream pressure pressed the disk leaving a small gap between the 
valve seat and the disk on the upstream side.  Fluid flowed through the gap 
and up to the leaking packing gland and out via the leak-off piping.  The 
leakage drew hot fluid from the hot leg.  The hot fluid layer rode on the cool 
fluid upstream of the valve resulting in thermal stratification.  
Subsequently, the hot fluid resulted in the thermal expansion of the valve 
disk closing the gap between the valve seat and the disk stopping the leakage.
Then, the fluid upstream of the valve cooled due to heat loss to the 
environment.  The process was repeated continuously, drawing hot fluid from 
the hot leg and resulting in fatigue due to thermal stratification.  

This information was received through discussions with the foreign government 
and industry representatives.

Discussion:

Although the Farley 2 and foreign reactor events are similar in that they 
involved thermal stratification in unisolable piping due to valve leakage, 
there are the following important differences:

(1)  Farley 2 involved a small amount of "cold" fluid flow into the RCS, 
     whereas the foreign reactor event involved a small amount of "hot" fluid 
     flow out of the RCS.

(2)  Farley 2 involved valve leakage through the valve from upstream to down-
     stream, whereas the foreign reactor event involved valve leakage out of 
     the packing gland.

.                                                  NRCB 88-08, Supplement 3
                                                  April 11, 1989
                                                  Page 3 of 3


(3)  Farley 2 involved a steady valve leakage which caused the cycling of a 
     check valve due to fluid pressure buildup, whereas the foreign reactor 
     event involved a periodic valve leakage through the packing gland.

The staff at the foreign reactor evaluated several approaches in addressing 
the concern raised by the event.  Attachment 1 describes three approaches that 
were considered.

Actions Requested:

Although the actions requested in NRC Bulletin 88-08 are unchanged, it should 
be noted that periodic valve seat leakage through packing glands could result 
in unacceptable thermal stresses.

Reporting Requirements:

The reporting requirements set forth in NRC Bulletin 88-08 remain unchanged.

If you have any questions about this matter, please contact one of the techni-
cal contacts listed below or the Regional Administrator of the appropriate 
regional office.




                              Charles E. Rossi, Director
                              Division of Operational Events Assessment
                              Office of Nuclear Reactor Regulation


Technical Contacts:  W. Jensen, NRR
                     (301) 492-1190

                     S. Lee, NRR
                     (301) 492-0943


Attachments:  
1.  Three Approaches Considered in the Foreign Reactor Event Evaluation
2.  Fig. 1, "Schematic of the RHR Suction Line at the Foreign Reactor"
3.  Fig. 2, "Formation and Fluctuation Mechanism of Thermal Stratification"
4.  List of Recently Issued NRC Bulletins
.                                                  Attachment 1
                                                  NRCB 88-08, Supplement 3
                                                  April 11, 1989
                                                  Page 1 of 1


       THREE APPROACHES CONSIDERED IN THE FOREIGN REACTOR EVENT EVALUATION


(1)  The piping layout may be made such as to minimize the effects of thermal 
     stratification.  If the valve is sufficiently far away from the elbow, 
     the high temperature associated with the small amount of leakage flow may 
     dissipate before the leakage flow reaches the valve.  Thus, the potential 
     for thermal expansion of the valve disk caused by the leakage flow is 
     minimized.  However, because it was difficult to elongate the horizontal 
     pipe due to the limitation of space, the plant staff did not adopt such a 
     change. 
     
(2)  The closure of the subject RHR isolation valve was controlled by limit 
     switches which moved the valve disk to a preselected position.  With the 
     valve closed, a small gap existed between the valve seat and the disk 
     permitting the cyclic fatigue phenomenon.  This small gap may be 
     eliminated by closing the valve to a preselected torque so that the disk 
     is set tightly in the valve seat.  However, in order to ensure the valve 
     opening on demand, the plant staff elected not to adopt the approach.

(3)  The action selected by the plant staff was to adjust the valve limit 
     switches to position the disk so as to increase the gap between the valve 
     seat and the disk at valve closure.  Although the packing gland was 
     replaced a continuous leakage path would exist through the increased gap 
     should the packing gland leakage develop again.  By permitting continuous 
     leakage, the cyclic fatigue phenomenon is eliminated.  Furthermore, the 
     plant staff insulated the leak-off piping and the associated temperature 
     sensor to provide leakage detection and instrumented the RHR piping with 
     temperature sensors to monitor for thermal stratification.