Hydrogen service lines 9

[Mche]

Hi,
Are there any specific design criteria to size Hydrogen service pipe lines such as velocity, Pressure Drops etc.?
We attempted to optimize line sizes using indutry norms, but were questioned for basis.

Don't have any specifc references to answer back. One thought of is API RP 14E. However I am not sure if there exists any for Hydrogen service hence specifically looking for it.
Please suggest, if any one knows.
Thanks

 

[davefitz]

For normal sizing, I would guess you would size the piping for a velocity in the range 10-20% soundspeed, to limit pressure drop to 10% inlet pressure for typical piping systems.

You can also search for NASA design standards- those are the only public hydrogen standards I am aware of . There are private standards by GE for hydrogen cooled generators and associated piping.

 

[Montemayor]

Mche:

I'm going to assume you are a Chemical Engineer. As such, then, I also presume you are talking about Process design - as opposed to mechanical design. This is important to distinguish because H2 makes a big differnce in the mechanical design and very little difference in the process design.

For Process Design, you basically design a Hydrogen pipeline just as you would any other gas. The fundamental and basic guideline to all gas lines is the allowable (or design) pressure drop. Normally that is the key (& oftentimes) the optimum design level to base your calculations on.

If you were questioned for basis, state your scope of work - which you should have generated before any attempts to resolve the problem(s). Use the recognized velocities as they obviously will drop out when you design around the allowable pressure drop. You have to stay sub-sonic in velocity, so that gives you an automatic ceiling value to design from. Always resort to recognized, acceptable standards such as the NORSOK process standards. Norsok recommends that in lines where pressure drop is not crtical, gas velocity should not exceed limits that may create noise or vibration problems. A rule-of-thumb velocity is given as:

V = 175 (1/Den)^0.43
where,
V = Max velocity to avoid noise, m/s
Den = gas density, kg/m^3

Norsok also lists recommended velocities if you lack these. To download their complete process design standards, go to:

http://www.standard.no/standard/index.db2?id=2256

I hope this helps.

Art Montemayor
Spring, TX

 

[Mche]

Montemayor,
You guessed it correctly, I needed answer from Process Design point of view which I got from above.

However I am curious about impact of hydrogen on mechanical design also. Is there anything additional, that Process Engineer can do than mentioning presence of Hydrogen and its concentration in the service? (for material selection and fabrication procedure control?)
Could you please suggest some readings as to know why "H2 makes a big differnce in the mechanical design" and are there any guidelines, when it should be higlighted to Mechanical design? Preferably on internet or Codes.

Thanks

 

[Montemayor]

Mche:

As we both know, Hydrogen is the smallest molecule in the Cosmos. It will literally permeate through anything, if given enough time to do so. It goes through steel pipe walls and if you have ever operated and maintained a Hydrogen system, you will have experienced the fact that when you D&R (Demolish & Remove) old carbon steel Hydrogen piping, you cannot go in and burn out the pipe without first allowing the H2 soaked into the walls to evaporate out. If you don't, the pipe seems to "catch fire" when the cutting torch heats it up.

Hydrogen also has a negative effect on steel that is titled "embrittlement" - a process that weakens the parent metal to the point of failure. The higher the temperature and pressure of the H2, the more pronounced are the negative effects.

I don't have my materials library here handy at the moment but I recall that we were using a chrome-steel alloy on piping for H2 service back in 1970. There may be other aspects to H2 materials of construction that you should bear in mind, but those two particularly come to mind. I personally have always butt-welded all my H2 piping, reducing the quantity of flanges or possible leaks down to zero - if possible. If you have ever seen (it would have to be at night) a self-ignited H2 leak through a flange gasket, you will wind up doing the same. Hydrogen will always leak -to begin with- and what's worse is that the stuff has a negative Joule-Thomson value: it heats up when it expands. It also self-ignites with the expansion friction created and the leak catches fire. The way we found most H2 leaks in Steam Reformer units and Hydrogenation plants was to walk out at night and keep an eye out for the bright, whitish flame that H2 gives. You can't see it in the daytime and, in fact, we had an operator severly burned by walking on a catwalk besides a Hydrogen line where there was a leaking flange. He never saw the flame as he walked by and had his arm burned just as if it had been a cutting torch.

You are right on track in keeping your Mech. Engineers fed with as much process information as you can give them. They can only design the mechanical end to the same degree of information that they can gather on H2. Some of the M.E.'s should already know a lot of the H2 characteristics, but it's always wise to work together as a team to produce the safest, most flexible, and most cost-effective installation. There should be a lot of information out on the Web regarding Hydrogen piping, handling and materials of construction. I know books have been written on the subject. I can't come up with a list of them, but I know they have existed. I optimistically would expect some of my colleagues on this forum to pitch in additional or further information in accordance with their experience in this field.

I hope this has been of help.


Art Montemayor
Spring, TX

 

[25362]

I'd like to clarify one point on hydrogen's wall-penetrating characteristics, by saying it's not only the size of the molecule but -even of more importance- the mass of the molecule than enables it to effuse or diffuse through solid-like walls.

If my sources are correct, hydrogen is, in fact, the smallest diatomic molecule (diameter~1.5*10^-10m) and the lightest around. Helium, a monoatomic molecule, although heavier than the hydrogen molecule, seems to have even a smaller diameter of ~0.98 Angstrom.
1 Angstrom=10^-10m.

Effussion rates are inversely proportional to the square root of the molecular mass not volume. This is known as Graham's Law. Thus, hydrogen would penetrate (diffuse) more easily through the pores of a solid wall.

This is explained by the fact that lighter (i.e., of lower mass) molecules move faster and strike the wall more often increasing the chances to move across. Any comment ?

 

[davefitz]

I suggest you read some established design standards specific to hydrogen service, such as the NASA or GE standards. You do not want to learn by mistake with this gas.

For example:

a) NASA does not allow a buried H2 line- if it leaks, it will permeate into any unpressured pipe ( ie sewer line) and result in a fire hazard someone's building. If needed to be below grade , they recommend a trnech with grating. Although most outdoor combined cycle plants with H2 cooled generators ignore this provision, but be sure to avoid running the pipe near an unpressured line (drain or sewer)

b) Nasa recommends that if run overhead in a pipe rack, it be the highest pipe , so if it leaks, it will not burn any above laying component.

c) When H2 burns, it cannot be detected by visual cues or radiant heat, as it burns invisibly with near zero emmissivity. The only safe method of finding the leak is to proceed with a hay broomstick ahead of you or throw sawdust ahead of you ( this recommendation is from the people that flew to the moon!)

d) GE provides recomended types of piping material and joining techniques, and allowable sealants. Also, specific valve desigs that are proven safe with H2.

e) when passing into a building , use of a double wall pipe ( inner annulus vented to outside). Building roof vent fans to be explosion proof. cieling high points to have H2 detectors and vents.

 

[MJCronin]

Mche, and others,

What seems to be missing from the above discussion is the document that governs the mechanical design of hydrogen piping systems.....

"NFPA 50A: Standard for Gaseous Hydrogen Systems at Consumer Sites, 1999 Edition"

NFPA 50A references ASME B31.3 for piping system design and mechanical integrity

I suggest that any piping system smaller than 2 inch NPS be schedule 80

My opinion only.....

MJC

 

[unclesyd]

There are several material and design conditions that need to have someone with the knowledge base of particular design considerations of a system containing hydrogen.

As mentioned there is the H2 embrittlement concerns and along with this is very important consideration of H2 Induced Fatigue in cyclic services.

There is an often overlooked design consideration with H2 in severe cyclic service, as in a PSA system used for H2 purification, of H2 Assisted Fatigue.

Someone correct me if needed, I believe there is material selection chart based on H2 partial pressures chart from Nace and also one from Shell Oil.

Anecdotal, but an important consideration.

The H2 Assisted Fatigue problem got us really bad in PSA system for a H2 plant. We had the first failure that was actually examined metallurgically to determine the cause. Ultrasonic examination of the remaining vessels revealed fatigue cracking in all but 4 of the remaining vessels I was fortunate in the fact that I got to see all the small details and even had a sample of a 17" long H2 “Assisted Fatigue” crack, 7" long 1.125" wall penetration from an operating vessel that I was able to procure by convincing management that it was essential to cut in patch and not repair the crack. At the time is was the only one in the world available from an operating vessel and available for metallurgical examination. It went all over the world. It was a very enjoyable time.

All vessels were replaced on an expedited schedule with HIC resistant steel, hemispherical heads and very smooth weld roots.
There had been several failures at different locations but none had been metallurgically examined. One of my colleagues had appraised a large chemical company in England that had three similar systems. On Wednesday they had no failures in any of their vessels. The following Monday we got a call that the head to shell weld on a vessel had failed, within several weeks they had two additional failures. The commonality of the vessels other than the similar design conditions were they were all about 17 years old. On checking with two additional companies, both had failures in 17 year old vessels. The failure mechanism had taken quite a few years to progress about .125" while the last inch had taken only a few days.

In dealing with H2 and have any doubt get some help.

 

[jay165]

syd:

You are correct about H2 partial pressures. It's also called the "Nelson Chart"

 

[Montemayor]

All:

My intent here is not to lengthen this thread, but MJCronin's input, unclesyd's remarks, and jay165's timely reminder of the Nelson Charts point to the seriousness of this subject and what I failed to include in my first post (albeit, due to a lack of resources at the time). The above colleagues have delivered what I was unable to come up with originally and I thank them. An excellent discussion on some of the concerns with Hydrogen corrosion and effects on materials can be found at:

http://www.ndeassociates.com/hydrogen.htm

I hope that the original poster, Mche, has understood what I was trying to expound: the mechanical design of H2 systems is far more important to consider than the process design. This is an area where Chemical Engineers need a thorough "dousing" into mechanical design and skills. The difference between a ChemE and an MechE should be nil or nothing when considering the skills that both can impart to a safe and efficient H2 system. It has been my experience in the past that ChemE's have tended to separate themselves from their MechE brethren, preferring to remain in a "process area", when they should be doing the opposite and learning more about Mechanical expertise that is so essential to achieve a professional engineering level. An engineer can only enrich himself and his abilities when he actively learns from the other disciplines. This appliation is an excellent example where both skills need to be firmly integrated to ensure total success. My thanks to all for making up for my initial deficiencies.

Art Montemayor
Spring, TX

 

[MJCronin]

Mche,

FYI, the American Petroleum Institute has done research and published guidelines on hot hydrogen service. Have you investigated these ?

Try

http://www.ais-engineering-books.com/

and see :

API RP 941 $US74.00 31 pages - hard copy

"STEELS FOR HYDROGEN SERVICE AT ELEVATED TEMPERATURES AND PRESSURES IN PETROLEUM REFINERIES AND PETROCHEMICAL PLANTS"

Good luck

MJC

 

[moltenmetal]

Art: Amen brother! You're preaching my religion! As a chemical engineer, I've benefitted immensely from learning more about mechanical and electrical/instrumentation engineering. Likewise, mechs can benefit from learning a little chemistry too, especially if they're practicing in the chemical process industry. A combined team is best, but the team only works well if all parties understand where the others are coming from. Cross-disciplinary skills are very, very important to the engineering of a safe product.

 

[Mche]

Thanks all for your useful information. One more reference which I could locate is...

"Material Selection for Petroleum Refineries and Gathering Facilities" by Dr. Richard A. White

 

[Mche]

Following I note from various references I could read in last few days.

If a process engineer has to participate in design of a system where hydrogen is one of the components then he/she should, at least,
1.Specify hydrogen partial pressure in process data sheet
2.Based on operating temperature and hydrogen partial pressures verify the selected material. (Reference, API 941 be used.)
3.Verify for mechanical design that PWHT is specified.

There may be lot more as I am knowing more about the subject but I feel this is bare minimum.

 

[25362]

It should be noted that while molecular hydrogen can penetrate metal walls, atomic hydrogen formed by catalytic effects on the metal surface is even lighter, and can move faster within the metal matrix and may chemically combine with the alloy components, especially carbon.

 

[mpcfotia]

I am interested in the types of valves that are recommended for hydrogen service. Is a ball valve particlularly better suited for hydrogen than a diaphragm valve?

Where can I obtain a copy of the GE standard on hydrogen service?

 

[davefitz]

GE recommends a specifc style of globe valve for H2 service.

You can obtain the GE recommndations at <

http://www.gepower.com>

and registering.

 

[davefitz]

err, correction- GE recommends a specifc style of ball valve. Sorry about that.

 

[davefitz]

The latest (1997) NASA design standard for hydrogen piping systems, storage, etc is found at

<

http://www.hq.nasa.gov/office/codeq/doctree/871916.pdf>