Pilot Plant: Calculations for Radiant Heating of a Process Fluid to 1200 C 4

[plantprowler]

I need to heat a process gas (@ 10 bar) to approx 1200 C for a short residence time for a chemical conversion in a pilot plant. The geometry I am considering is a 6" to 8" dia coil bundle (approx 20 turns ) made of Inconel / Hastealloy tube approx. 0.5" dia with a gas flame at the center.

Questions I'd love comments on:

(1) Does the idea seem OK? Inconel 625 should be usable up to 1800 C it looks like? Also the flame temp. of a Natural Gas flame ought to be higher than 1200 it looks like (approx 1900 C) so long as I use the right air-gas blend.

(2) Any ideas about how I could estimate the max heat flux I could deliver into the pipe using a flame? Would the Stefan Boltzman law provide a good estimate? Any other tips?

 

[EdStainless]

What is the gas pressure? Is there a design code? This will establish the max temp.
Any risk of corrosion or decomposition?
You want to avoid direct flame impingement on the coils.
Typically this would look like a miniature superheater, a serpentine coil bank with the hot gas flowing up through it.
You need to control the gas temp wrt pressure rating and such.

A coil around a flame is inefficient since only the ID surface is getting any radiant heat.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[plantprowler]

Thanks Ed!

Gas pressure is 10 barg. I'm thinking of regulating this with a backpressure valve downstream. The gas is for most part superheated steam with some process gasses (10 parts of steam to 1 part of process gas) in the mix. The steam will be condensed downstream.

No risk of corrossion I see. The components of the feed are going to be H2O and Carbon (5% w/w). The products are foreseen to be CO / H2 / CO2.

Do you forsee any of these to be a problem to Inconel? Do you have any other alloy choices I ought to be considering?

You mentioned "gas flowing up". My current design has gas flowing down through the coil. Does it matter?

I am thinking of surrounding the coil with a firebrick casing (cylindrical). Would this solve the inefficiency problem of only one side of the coil getting exposed?

Can I use a coil material where even if the process gas was stopped or direct flame impingement happened coil meltdown wouldn't be a problem?

No specific design code is needed per se. This is for a very small pilot plant in an R&D setup outside the USA with no governing local code for such a setup.

 

[XL83NL]

Yes, Inconel is a probleml; it wont work. Even if there's no design code involved (from jurisdicitional point of view), you'd better give this a good mechanical design (review). Youre going to extreme conditions, and that's something that should not be taken lightheartedly.

10 bar(g) @ 1200 °C will (as good as) not work with any commerically known metal alloy like hastelloy, incoloy, etc. Im currently involved on a pilot plant reactor design for 25 barg @ 950 °C, and one for 1250 °C (with slight overpressure). A combination of both pressure and extreme temperature (far into the creep regime) is as good as impossible with any metal. You might be able to do it with ceramic. But beyond the reactor design, there are lot more issues to think of, which arent mentioned. Im sure other members here like moltenmetal can give you a good response with valuable input, as they have a lot more experience with this then me.

 

[MortenA]

XL83NL may have a point ;-) Just check this page for _melting point_:

http://www.engineeringtoolbox.com/melting-temperature-metals-d_860.html

Inconel will melt at 1400C so one would guess it would be a little "soft" at 1200C?

Best regards, Morten

 

[EdStainless]

The special metals literature has stress rupture and creep data.
For alloy 625 at 1100C you will see 100% creep (basically rupture) in 100hrs at a stress of 1ksi.

There are higher strength alloys, such as Haynes 214, which at 1200C will rupture in 1000hr at 0.5ksi stress


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[plantprowler]

@XL83NL

All great points. Yes, a detailed mechanical review will be done before the build stage.

Can you tell what MOCs you are using for your 950 C / 1250 C projects? One option for me may be to redo my conversion calculations at a lower T/P combination.

@EdStainless:

I could live with 1000 hrs replacement times easily. This is only a pilot. I will see what my stress levels are.

 

[XL83NL]

For the 950 C design were currently considering inconel 617, as it has the highest stress allowable per ASME IID. Inconel 800H/HT is also possible, although the HT version is not a code approved material. We make it from bar, ID 15 mm, wall thickness about 8 mm I think. There are other materials possible as well, like inconel 602CA and Haynes 230, but availability and allowable stress levels, together with machinability is what makes the choice. So it depends on your design pressure and ID, as they determine the required wall thickness.

For 1250 C design were considering a ceramic 'alloy', when I'm back at the office I'll see which one exactly it is. This is for a reactor with a design pressure of I think 5 barg. I'm not sure if a metal alloy would work if you have 0 barg design pressure, I'd have to check that Monday also.

But the design and material selection is just one thing. Safeguarding and inspection for creep deformation and ultimately replacement due to failure is another. You have to think of a program/procedure for that as well.

 

[plantprowler]

I came across this performance note about the alloy Hastealloy X, that mentions it retains its strength up to 1200 C. Any thoughts? Is that only a marketing hype pitch or would this be a viable candidate alloy? Or would it have the same, severe creep / strength limitations that would restrict its use to few tens of hours?

HASTELLOY X is a wrought nickel base alloy with excellent high temperature strength and oxidation resistance......Strong and Oxidation Resistant to 2200 Deg. F (1200 Deg. C) - HASTELLOY X is a solid solution strengthened grade has good strength and oxidation resistance up to 2200 Deg. F (1200 Deg. C).

http://www.hpalloy.com/Alloys/descriptions/HASTELLOY_X.aspx

 

[EdStainless]

Go to the Haynes Intl web site and retrieve the creep data yourself.
X is a 50 year old alloy that is very inferior to 230 or 214.
I don't see a reasonable alloy solution if you insist on 1200C at 10bar.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[plantprowler]

@EdStainless

Thanks! Then I guess I will abort the 1200 C trial and try to figure out a lower operating temp. perhaps around 900 C

I did check the Haynes data sheet. At 1100 C:

8 MPa produces 2% creep in 10 hrs.
6 MPa produces 1% creep in 10 hrs.

 

[XL83NL]

Ive reviewed some of the info for the 1250 °C design. I have little practical experience in actual built designs. Designs have always been reviewed by an experienced engineer.

The reactor at 1250 °C is based on Si3N4 ceramic alloy. Not sure if we can design it for pressure as well, but I seem to recall we have a quote for that.
At 1250 °C no metal alloy will work.

 

[moltenmetal]

My suggestions would be, in this order:

1) Give up, or
2) Do your heating electrically, or
3) Run a pressurized burner and use something like a FeCrAl alloy (fecralloy) such as Kanthal or perhaps a ceramic material.

You cannot transfer heat into a tube which is under internal pressure at 10 bar into a flowing gas leaving at 1200 C without having to replace the tube very, very frequently. Accordingly, you will have to either pressure balance the tube (i.e. pressure inside nearly equal to pressure outside) or find another source of heat. The latter is the most sensible, next to giving up and not doing this at all. Many people have the idea that heating with fuel is more efficient inherently than heating with work (electricity is basically thermodynamic work on tap), but they're wrong especially at high temperatures. Unless you have a use for the enormous amount of flue gas which will be leaving this thing hot, your efficiency is negligible and an electric heater will do the job better as well as being much easier to control.