From August 2012, edited March 2013, added 10 October 2016

[Bruce Bullough] I'd like to get your comments, clarifications, experiences, etc., on carbon steel storage tanks at low temps. We have a carbon steel tank that we use as a reservoir tank in our chiller system - the general assumption is that it is a low grade CS, by the way, it's old and we don't have the shop drawing anymore. The chilled liquid (50:50 blend of water and methanol) is kept at -25°F (it might get as cold as -28°F). The tank is well guarded by bollards and the walls of a containment dike, and in a low-traffic area. I'm well aware that CS is not a cryogenic material, and it loses most of it's "toughness" by time it gets to ~0°F. Ideally a tank in this service should be austenitic stainless steel, a "cryo" grade material. I have a crude and general understanding of what toughness is, and that it is not brittleness.

[Christopher Wright] Material Toughness alone is only part of the story. A brittle fracture requires 3 things: a crack of critical length, enough stress to propagate the crack and a material which isn't 'tough' enough to arrest the running crack.

What often happens in pressure vessels is that a locally high stress like the stress at a weld undercut or a machinery attachment or support or maybe a nozzle connection produces a tiny crack under cyclic loading over time. The crack extends itself to the socalled 'critical length' and the vessel ruptures suddenly with the crack suddenly extending over the length of the vessel. In this context, a 'tough' material is one that can sustain significant yielding and prevent failure by fracture or fragmentation.

The ASME Code specifies Charpy impact tests to determine the toughness of a particular lot of material, but you can make some generalizations about material toughness are possible about a material based on its chemistry and processing, and there are lots of exemptions such as those in UCS-65.

Only carbon steel undergoes a change in toughness with temperature. Austenitic steels and certain nickel alloys have the same toughness down to -300F and aluminum and copper which aren't very tough to begin with (compared to steel) have no change.

Carbon steel toughness varies with production practice, generally alloy content and heat treatment and grain refinement. The ASTM specs pertaining to lower temperature operation (SA-516, SA-537 SA-333 for example) typically specify a higher manganese content along with fine grain practice and normalization. There are also Charpy impact test requirements to insure that these practices are followed, and weld procedures need to insure that tough welds are also produced.

Design for added toughness. Every ASME Code shop has the tools to produce carbon or low alloy vessels suitable for temperatures down to -50F. Appropriate material choices need to be made and good welding practice is vital and you need good welders, but it's not rocket science.

It's possible to predict (roughly) the conditions for fracture, if materials data is available, but that's not always the case. Usually it's easier to avoid cracks and specify tough material for your service conditions at the time of construction. NDT is always a good idea. Production people who don't want to mess with fixing problems will tell you that 'you can't inspect quality into a vessel,' but in fact there's no better way to determine if someone's slipped up or cut a few corners than a mag particle or penetrant test on a weld.

[Bruce] I have seen the result of cold vapors (probably colder than -50° F) hitting a CS tank, with probable vibration suddenly nearby along with the effects of differential thermal expansion/contraction on the tank.

[Chris] I'd bet that there was a crack somewhere that'd already started maybe from previous thermal shock or other loading or perhaps a weld defect, although thermal shock is a possibility. As time goes on the crack grows, possibly aggravated by vibration. Then at some point the vessel gets hit with a sudden load and the rupture occurs, but the failure is cumulative extending over a long period.

[Bruce] But, I've also seen MANY carbon steel propane tanks sitting outdoors on farms and at factories in MN, ND, WI, WY, etc. where I know that temperatures can be -20°F and colder for extended periods.

[Chris] Low temperature alone won't cause a sudden fracture--as mentioned, brittle fracture needs a high stress and a sharp defect, like weld undercut of a fatigue crack and a susceptible material. Without all of those three things brittle fracture won’t occur. You can't do much about the temperature, but you can determine whether your material remains tough at the service temperature making certain to eliminate pre-existing cracks which are usually associated with welding. The LP tanks you're talking about also have some mitigating circumstances. The lower the ambient temperature the lower the pressure of the LP. That's because the saturation pressure of the liquid drops off with temperature. Hence the stress goes down. The construction of those vessels is simple and it's often automated so welding problems typically get worked out when the production cycle is set up. LP isn't corrosive so wastage isn't a problem. I'm sure abuse is an issue, so it's likely tank material is chosen for a degree of toughness to begin with, and ongoing testing likely helps weed out defective tanks. Chances are pretty good that the tank wall thickness is higher than needed for pressure alone to make sure the tank is rugged enough to withstand routine rough handling.