H2 corrosion allowance, 304SS 2

[rocketscientist]

Hi,

I am a consultant asked to evaluate a hydrogen generating system. This system involves running an electric charge through a bed of KCl, separating the hydrogen from the oxygen and compressing the hydrogen to 500 psig. Of course, the temperature of compression is extremely high for hydrogen, about 1300 F. The compression step consists of two jacketed compressors and an intercooler. In other words, the gas goes through the first compressor, then an intercooler, and finally a second compressor. If the compressor load is balanced, the exit gas after the last compressor is at about 500 F, which oddly is about the reaction temperature in the fixed catalyst beds. These beds are used to hydrolyze allenes to saturated.

Here's the problem(s). The project engineer decided to go with 1-inch (25 mm) tubing: wall thickness is 0.049 inch. From what I can see there is no allowance for corrosion.

Just taking into account mechanical strength, the pipe supports need to be a whole lot closer given that the allowable strength of 304SS is about 15,300 psi at 500 F; dropping from 17,000 psi at 200 F.

I calculated a hoop stress that is in the 5,000 psi range.

My own opinion is that the tubing needs to be replaced by schedule 80 pipe that will allow a moderate corrosion allowance, say about 1/8 of an inch. I would add many more pipe supports to compensate for the loss of strength.

Although the pipe wall may be intact the H2 may eliminate structure in the interstitial areas. I saw this type of corrosion with Inconel 600; the wall was there but hanging there like chicken wire --- providing no actual support.

I am also concerned the internal tubing in the compressor itself since this is also tubing.

Any thoughts on how thick the allowance should be or how I should approach this problem?

 

[enmax]

Corrosion allowances are applied where it is believed that general corrosion will occur. I'm not aware that hydrogen causes general corrosion of austenitic stainless steel such as 304. However, there have been reports of embrittlement of 304 and 304L by hydrogen when the steel has been cold worked. Only use austenitic stainless steel in the solution annealed condition. It would also be an added advantage to use 316L stainless steel which is less prone to the formation of strain induced martensite and so is less prone to embrittlement.

 

[rocketscientist]

304S tubing was used. This implies that it was cold-worked. Swagelok fittings were used for ease of assembly.

It sounds like we have a problem.

Based on what I told you what do you think now?

 

[unclesyd]

We used Austenitic Stainless Steel piping and tubing all around our Hydrogen plants and compression system as well as our hydrogenation process. All the piping around the H2 plants was SS until the temperature dropped below 250°F. The hydrogenation system operated at 10,000 psig for years then was dropped to 6250 psig.
There were 4 H2 sample lines operating at 250 psig amb from the area to the lab that were connected with Swagelok fittings.
We never saw any problem with H2 embrittlement pipe or tubing or any problem in fact.

We have at least three laboratory H2 generators that use the same process as yours to generat H2, all with SS internals.

Having said all of the above here is paper on the embrittlement of SS by H2.

http://www.icf11.com/proceeding/EXTENDED/5638.pdf

 

[enmax]

I've had a look through my files and attach a summary I made some years ago (I've deleted certain words for confidentiality reasons). The company I worked for at the time had had no problems with the use of 304L carrying hydrogen but, as you can see, I ended up recommending 316L for new installations. The addendum refers to problems with vessels used at low temperature which doesn't apply in your case. My current company always recommends solution annealing of cold worked stainless steel when it is to be used for cryogenic purposes.
My view would be a pragmatic one - if the plant is already built then the risks are small and so leave well alone. But if I were to start from scratch then I'd use a higher alloyed stainless such as 316L and always solution anneal any that is cold worked.

 

[rocketscientist]

Thanks for you input.

 

[EdStainless]

Let's back up half a step. 316L will work fine, IF it is annealed fully to start with.
You need to demand less than 0.5% residual delta ferrite in the tubing. This would apply to both seamless and welded tube, I have seen it in both.
The easy way with welded is to spec A249 S7 with R<1.00.
For smls stick with magnetic measurements.
Actually well made welded tubing will be more reliable than smls.

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

 

[moltenmetal]

"Cold work" is a matter of degree.

We've used annealed 316/L tubing (both seamless and welded) with Swagelok fittings, bent it to common bend radii, swaged ferrules onto it etc. (i.e. doing a fair bit of "cold work" in the process). We've then used it with absurd partial pressures of hydrogen from ambient up to 1000 F. At lower pressures we've used 316SS tubing in services beyond 1000 F. We've never seen an embrittlement failure.

I concur that annealing is critical. Be aware that we and others here have seen some REALLY awful seamless tubing of late coming principally from China- tubing with serious internal metallurgical damage from bad practices in manufacture. We've had much more consistent success with welded seam tubing.

 

[EdStainless]

{Thanks MM, I'll pay you later}

We have been trying to source some hollows to draw into seamless tubing. We have rejected 8 of the 10 mills that we have looked at. This is stuff that wouldn't pass NDT or meet propertes after we did multiple cold draws and anneals. Imagine what it was like in the first place.

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

 

[TVP]

Ed,

Just out of curiosity, can you elaborate on the 8/10 statistic? Were all 8 mills outside of North America? Were the 2 accepted mills in North America?

 

[EdStainless]

There are none in NA. In theory there are a couple of people in NA that could, but there costs are 3-4 times what we need. All of these were in Asia.
We have one European supplier. He is only twice the price with a lead time that is 6 months too long.

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