Cracking of welds posssibly

[cougarfan]

We have SS tank ( 316 L ) that we think leaked where we welded it - the material going into the tank was at 200 F - could we have thermal shocked it - is this possible - is 200 F too hot - what do we do to prevent this in future? The tank is vertical hold capacity of 5700 gallons - leak is at bottom....

 

[CWIC]

Have you considered the possibility of any weld discontinuities before the assumption was made of the tanks service medium?

What type of medium is the tank holding?

 

[GRoberts]

Tanks usually have a relatively thin wall, so thermal expansion does not normally create high residual stresses. It prefers to buckle instead as a means to relieve the stress. And with the temperature being only 200F, it probably wouldn't do much of that either. If the thermal stress did cause the tank to crack, it is highly likely that it did so at a pre-existing defect. Without any details, it is impossible to say, but I would bet that you had a problem before you added any fluid.

 

[cougarfan]

The product in the tank added was napthalene if that makes any difference.....A 316 SS plate was welded over the cracks in the middle of the tank....

 

[stanweld]

Since leakage occurred after your repair of previously cracked and leaking bottom plate, you should first determine that leakage was at your repair welds and not in non-repaired base metal. It is quite possible that the original cracks originated from the outside surface and cracks not yet through thickness in adjacent base plate simply propogated during your welding repair and re-introduction to service.

Weld repairs involving pathches can be tricky and cracks at corner tie-ins have not been uncomon, especially with fillet welds. Proper weld sequencing is often required to prevent high induced weld strain at the corner tie-ins. Often these cracks are immediately evident through liquid penetrant testing but by no means always. Subsequent loading during operation will induce additional strain that can cause cracking.

How did you examine the completed patch-type weld repair? Vacuum box? Liquid penetrant?

 

[artzelda]

First of all you have not provided sufficient information to point to a solution. For example, when was the leak first observed, what is the fluid within the tank, what contaminants are in the fluid, is the material L grade or other grade????
If the leak was observed within a few days of filling then the problem could be due to hot cracking of the weld deposit. If this is the case then the weld procedure has to be reevaluated. This can be determined via a replica of the surface.

If the leak occurred much later than stress corrosion cracking could be the problem. Much additional information is required. Analyses with out this additional information is irrresponsible and may lead to an unsatisfactory action.

 

[stanweld]

In addition to my previous comments, was leakage found at existing cracks in the original base plate that you patched?

 

[PVRV]

1. Where were the crack located at bottom top side or
underside? what is the tank resting on?

2. How were the cracks detected? someone ramed the tank?
what are the cracks characteristics (lenght,
location,orientation)

3. Patching cracks, this scares me; anyways was the
welding procedure qualified?

4. Please revisit the 200F when considering welding
SS.

5. I dont think your unorthodox repair had anything to
do with the leak; rather propagation of the cracks.

Regards,

 

[unclesyd]

Having inspected and recommended repair of numerous S/S tanks I would almost bet you have Chloride induced Stress Corrosion Cracking originating on the outside of the tank floor. Where I worked this was such problem that all S/S tanks are now supported by Grillage Bars. SCC is essentialy non-repairable in this situation and extreme caution should be used when any welding is attempted on any tank floor. I think a PT test is advisable prior to any welding. A quick test in your case would be to lightly grind small area around the leaking area and immediatly spray PT Developer on the area, The napthlene should wet the developer. This is one way to overcome the tendency of materials to plug cracks and prevent the dye from penetrating.
Most the times best solution is to replace the bottom.
Also watch your codes.