Relief scenario? Intrusion from 1/2 pipe jacket

[okidoki]

A steam jacketed vessel's contents will react violently with water. The jacket is a 1/2 pipe coil.

Would steam intrusion through a rupture of the vessel wall be considered a credible relief scenario? If so, what rate should be used for the steam flow into the vessel?

API RP521 3.18.3 focuses on tube failures to determine the flow rate. I could use the two orifice assumption, but what if the vessel had a dimple jacket?

The only solution I see is to calculate the flow rate of steam through the supply piping at critical flow.

Is this situation covered in other Codes?

 

[TD2K]

I don't think that is a credible scenario unless you have information it has happened or suspect problems like corrosion. Double pipe heat exchangers which are constructed out of pipe are typically not evaluated for the case of a rupture of the inside pipe in many company's specifications I've seen. The rationale is that unlike tubing, failure of a component constructed from pipe is much less likely to fail.

If you do decide this is a credible scenario, looking at the steam piping seems reasonable. You might also want to look at rupture disks. I remember them putting on a presentation wrt a problem similar to yours where they had to quickly isolate the supply to a vessel's heating coil (I think that was the arrangement) if it failed because of the reaction and their valves were essentially tripped on low pressure. They had done some testing and come up with some impressive reaction speeds.

 

[MJCronin]

okidoki,

I totally agree with TD2k above. Rupture of the vessel wall to allow steam into the vessel seems to be an unrealistic scenario for the following reasons:

- If the vessel is an ASME VIII constructed/certified and inspected vessel and is operated within the MAWP and maximum temperature, the vessel shell is the probably the lowest stresses portion of the vessel. The half pipe jacket should have been designed to the ASME Appendix addressing such jackets.

- The most typical failures of half-pipe jacketed vessels are at the outside surface of the "fillet" weld to the vessel shell and are due to the half-pipe undergoing thermal cycling. The weld typically begins to crack and leak. A "through the vessel wall" failure would be unusual. A "dimple jacket failure" would definitely not penetrate the vessel wall.

- Steam leakage, if it occurs would be into the atmosphere and not into even a slightly pressurized process container.

Well, these are my thoughts and opinions.

Anyone else ?

MJC

 

[TD2K]

I said rupture disks in my post, that is incorrect. I meant to say rupture pin device (actually, the pin buckles rather than ruptures). Their URL is

http://www.rupturepin.com/

 

[butelja]

Make ABSOLUTELY certain that the steam inlet to the 1/2 pipe coil has an impingement plate to protect the vessel shell. If you have moist steam flowing into the jacket and it has no impingement plate, the odds of developing a hole in the vessel shell are much higher than others have previously stated.