High Temperature Excursion on a Reactor Tube

[mabel12]

I have a reactor tube that (during an excursion) saw pressures of about 43.5ksi (300MPa) and reached 2372F (1300C) for approximately 45 seconds (very worst case scenario). The reactor tube is AISI4340 material and approximately 2" thick (2" ID). Reaction was instantly killed after that excursion and temperature and pressure came down to ambient and zero, respectively, in a question of minutes.

I have two questions:
- How deep/ thick into the wall of the tube could have gotten above 727C during that period? (there must be an equation for this?) Assume the reactor tube outside surface is chilled with water at 20C and there is no fouling on the inside of the tube.
- Could the internal pressure on the pipe have any effect on the melting point of the steel?

Thanks!

 

[metengr]

mabel12;
You can approximate a mean wall temperature profile IF what you describe is correct in that the OD surface of the reactor tube was exposed to cooling water, and the ID surface reached 2372 deg F. Do you monitor the coolinbg water temperature?

Could the internal pressure on the pipe have any effect on the melting point of the steel?

No.

After everyting settles down, I would perform the following inspections;

1. Measure the wall thickness and OD of the reactor tube to determine if the the high temperature exposure resulted in creep deformation.

2. Perform a visual inspection of the tube ID surface to check for fissures running along the length of the reactor tube.

 

[unclesyd]

I concur with metengr that a through internal inspection is required.
Just for curosity what are the design conditions for the reactor?

Where were the safety devices?

 

[mabel12]

Thanks for the feedback!

Design pressure for the reactor is 45ksi, all relief devices are set to this, and the reactor never got above that pressure. I was just concerned about the temperature effects. I am very limited regarding inspection and NDE options: no internal access (even for remote camera) and the tubes are permanently jacketed (no external inspection like UT in lieu of internal).

Cooling water temperature floats with ambient temperature (cooling tower), but thermocouples are on the inside of the tube (no local cooling water thermocouples on the reactor). I'd say that the water temp probably didn't change much.

 

[metengr]

mabel12;

I am going to ask this question because I find it odd that no internal inspection can be performed. This is obviously a pressure vessel, and it probably is insured, and it probably is regulated. How in the heck do you get around not being able to inspect this vessel and still operate it with no internal inspection capability?

 

[EdStainless]

If you can't inspect and your instruments tell you that you had a high temp excursion, then this one is trash.
I would assume that a significant amount of the material has been heated well above acceptable temperatures. Problems may include cracking or tearing, excessive grain growth, possible corrosion attack.
Or if no one can possibly get killed or hurt when it fails go ahead and run it. Your gamble.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[blacksmith37]

An excellent thesis subject; Relate the heat flux to the thermal conductivity + mass of the steel. A bit like a cannon barrel with very slow burning explosive; Maybe the military has caculation techniques.
My hunch is the steel is not significanly damaged (this hunch is worth what you paid for it).

 

[EdStainless]

They are call Heisler charts.
I don't know if the Fourier modulus is in the proper range for this application, but if not then there are other solution approximations available to handle estimation of short term non-steady state heat transfer.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[mabel12]

Thanks for the responses folks.

Just to add. Yes, this is registered equipment, but it's a tubular reactor, so it's actually considered a really long segment of pipe (not a vessel). Inspection requirements are a little different for pipe versus vessel around here. I can get away with X-ray, etc. I can actually disassemble sections of tubing to look inside with a camera (but this is not exactly your average flange + gasket).

I've dug up some old FEA modeling work for this system that shows that it would take few minutes for the material to get that hot (tube acts like a big heat sink) and my process data says it was actually less than 10 seconds (as opposed to 45sec. as I originally posted) at the high temperature range.

Again, thanks for the help.

 

[EdStainless]

Collect your data and dig out an old heat transfer book and look up Heisler. Even at 10 sec the inner layer of material got hot. You need to figure out how deep into the material you exceeded the original heat treat temperature.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[TVP]

Just to add to what the others have posted, which is EXPERT level knowledge, 4340 steel will undergo transformation to austenite when heated above ~ 845 C, so the microstructure will almost certainly be different in the affected areas, possibly even untempered martensite. Even if there are no cracks or deformation, the microstructure may be unsuitable for high pressure service.