Titanium Tube Evaporator (HEX) Design: Tube Tubesheet Joints; welding vs expanding. Clad vs. Solid 2

[plantprowler]

I'm trying to evaluate various options for an Evaporator, all of which are equivalent from the Process Point of View but I wanted some opinions on the fabricability, difficulty, chance of messing things up during the welding etc.

The unit is a fairly small evaporator calendria. 30 m2 area; 32 mm OD Titanium Grade-2 tubes, 1.2 mm thk. 6000 mm long. Single Pass. Boiling brine on tube side (NaCl) and low pressure steam on the shell side (1.5 barg)

The decisions I am confused about:

(A) Ti-Clad Tubesheet or solid Ti-tubesheet

(B) Expanded joints or Welded Joints. Or Expanded plus Seal Welded.

(C) What's the best option for the tubesheet-to-tube-shell joint? Shell handles only steam and doesn't need an exotic MOC. The only option I see is a TEMA N head. Any others?

(D) Is it an option to use Ti tubes but a non-Ti tubesheet (say SS316-Ti)? Since welding Ti to SS would be difficult or impossible is expanding an option?

The fluids aren't toxic or incompatible and the P/T isn't high. So for this sort of duty what'd be the things to consider while making the choice?

In terms of the difficulty in getting the fabrication right what are the safest options?

 

[gr2vessels]

A. Ti clad is cheaper
B. Expanded is fine, no seal weld
C. One expansion in the Ti clad, one in the middle (if required) and light expansion extended 1 mm beyond the shell side face of tubesheet.
D. Don't mix and mess the Ti, unless you have the expertise for it.
Safest option : ask a HEX fabricator to do the job and guarantee it (material and workmanship).
Cheers,
gr2vessels

 

[MJCronin]

Boiling brine is very corrosive. Titanium is usually the best choice.....

Is this heat exchanger for a frack water processing facility ? Frack water can have a wide variety of corrosive salts compounds and acids. Your choice of MOC may affect the types of brines allowed in the HX

MJCronin
Sr. Process Engineer

 

[racookpe1978]

Why Ti instead of inconel or monel? I saw very, very few Ti HX in ship installations.

(That's lower temperature salt water, and salt water - not "brine" but does the OP's higher salinty and temperature tip the requirement towards Ti? )

 

[plantprowler]

@MJCronin

Thank you.

Yes it is indeed very corrosive and hence the choice of Titanium for all contact materials. i.e. on tube side.

The many choices arise because the shell side holds steam and that is not corrosive and hence the desire to not use Ti on the shell side. That introduces all the possibilities of how to mate Ti with CS or SS.

No this isn't fracking related. It is a Chemical Process Waste that's essentially neutral and only contains NaCl.

 

[plantprowler]

I don't think Inconel or Monel survive saturated boiling NaCl brine. Correct me if I am wrong. (PS. Is there a cost advantage to using Monel or Inconel over Ti?)

Even on Ships I'm seeing a lot of Ti exchangers in the Alfa Laval plate type.

 

[SnTMan]

plantprowler, the chosen configuration of the exchange will dictate some of your choices.

A) Assuming a non-removable bundle, clad tubesheet is likely required to enable welding to shell, if other than Ti. In any case, clad will likely be less expensive.

B) Can you tolerate a tube leak? If no, weld (and roll) the joint.

C) See A), if fixed tubesheet, just a corner joint will be fine. Integral channel (N type) could, I suppose be welded to tubesheet cladding. In this case I think you'd want UT and shear testing of the clad to make sure it is well bonded to the base metal. Bit of a difficult detail. Might consider an A-L type, bolted channels. Gaskets would need to be Ti or non-metallic, else they are likely to corrode out of the joint.

D) I wouldn't. Tube yield higher than tubesheet yield might mean difficulties for an expanded joint, and you could not weld it.

Regards,

Mike



The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[plantprowler]

@SnTMan

Thanks very much!

So far I was thinking clad was just a way to save on the money since Ti is expensive.

But in hindsight, based on your comment cladding offers a second advantage too in that it makes a tubesheet to shell weld feasable when the shell is non-Ti.

That seems a huge plus. It removes the otherwise tricky problem of how to join a Ti tubesheet to a non Ti shell.

So the best option for me seems to go for a clad Tubesheet and then rolled + welded joints.

Follow up question: Would SS316 be a suitable base metal for Ti-clad tubesheets? Also, would you recommend only explosive clad material or any other feasible options.

Other than the usual TEMA guidelines about the minimum tubesheet thickness is there anything specific for Ti?

 

[SnTMan]

plantprowler, I suppose 316 could be clad with Ti, normally the tubesheet base metal would match the shell materials. I believe that Ti cladding can be applied only by either by explosion bonding or roll bonding. For small quantities explosion bonding is usually less expensive.

Nothing specific for Ti clad thickness, but tubeside details and manufacturing will usually dictate minimum thickness. For example, if no grooves of any kind in the clad you may be able to use as thin as 1/4". However if gasket grooves such as pass grooves (normaly 3/16" deep) are needed, you'd use 3/8". If flatness of the tubesheet is a concern (often is) you may need a machining allowance as well.

Configuration (TEMA type) of the exchanger will help clarify a number of things.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[plantprowler]

Possible a naive question: Why isn't a BEM-analogous design possible where we merely replace the welding with bolt force.

i.e. I drew a rough sketch:

Of course, once in place one could never take out the bundle but neither can one in a BEM design. This would have obviated the difficulties in welding Ti-tubesheets to non-Ti shells without the complexity of cladding, floating closures etc.

What gives?

Is the flaw the gasket between the tubesheet and the shell? I suppose once it wears there'd be no way of putting in a new gasket? Is that the flaw in this design?

Just curious.

 

[EdStainless]

Most common is explosively clad Ti gr1 on steel for tube sheet. This is considered to more robust than roll bonded.
Clad in generally 2-5mm depending on application.
If clad you must seal weld the tubes. They tubes are lightly rolled with a controlled projection and then there is an automatic autogenous seal weld.
If you go clad, the low carbon steel backer can be welded to the shell (otherwise you end up with flange and gasket).
Power main steam condensers are built this way all of the time.
In lower temperature service I have seen new condensers built with solid steel tubesheets, Ti tubes, and factory applied high-build epoxy coating.
You may want to consider acid cleaning after fabrication. In hot brine Fe smeared into the surface can lead to pitting attack.

How hot will this be? What concentration do you plan on reaching? Have you looked at the corrosion curves for gr2? If a tube get obstructed and begins to boil dry you will get corrosion attack.

http://www.timet.com/datasheets-and-literature

page 4


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

 

[SnTMan]

plantprowler, yeah that's the flaw. If the gasket were to begin leaking, there is about nothing you could do about it.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[plantprowler]

@Ed Stainless

Thanks for the tips!

This will be saturated NaCl so approx. 27% w/w. The temperature we still have some flexibility about. The possible range for the design Temp. is 75 C to 90 C but preferably as high as possible.

PS. Can you elaborate on what you meant by "otherwise you end up with flange and gasket"? If you don't go clad, so you cannot weld, is there a way to flange and gasket both tubesheets to the shell? How?

The ideal design choice would be Ti-Grade-7 (Timet 50A.15Pd) but the cost is way higher than Ti-Grade 2. By moderating the Temp. and perhaps adjusting the pH it is felt that Ti Gr. 2 can be made to work in this application.

But if you have any comments I'd love to hear.

 

[EdStainless]

gaskets, just like you sketch. Bad news in the long run. The fewer the better, especially in hot equipment.

Don't use gr7, look at gr12 or gr16. They should have much better corrosion resistance in hot brine than gr2, and almost as good as gr7 at a lower cost.
Do you want the brine or the water?
Are you going to pull a bit of vacuum on the brine to lower the boiling point?

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

 

[plantprowler]

@EdStainless

Thanks. I will look gr12 & gr16. Hadn't known about those.

Actually, we want neither the brine nor the water. The brine is actually a by-product produced in a chemical process & is a disposal problem. Evaporate it and then salt gets sent to a landfill & treat the condensate some more & then either discharge it or reuse it for cooling towers etc.

Regarding using a vacuum:

The final stage of the project will be a triple-effect evaporator. There vacuum is a must.

Stage-I is a single effect where ideally we don't need a vacuum. But we might still be forced to use it if the corrosion rates of the Ti grade chosen are too high without a vacuum. If it weren't for the corrosion constraint then 1 barg steam gives enough of a delta-T to use as a heating medium without pulling any vacuum on the effect.

 

[moltenmetal]

If pressure and temperature are low (in this case that would seem to be right) and there isn't much vacuum to worry about (i.e. this exchanger, not your triple effect evaporator), loose cladding the tubesheets is another option. More welding, but less material cost, so your fabricator (who should be somebody who specializes in exotic metal fabrication rather than merely dabbling in it) will know what is going to be the optimal construction.

Non-heat transfer surfaces (bonnets) can be Teflon-lined carbon steel rather than solid or clad Ti. No need to have titanium just for titanium's sake, especially if you need a grade other than Ti Gr 1/2/3 commercially pure. That's how we always do these sorts of exchangers, whether titanium or tantalum- they are done -EM style (i.e. bonnetless( with ASME B16.5 flanged tubesheets, then we make PTFE-lined bonnets. We have similar tricks with BEUs etc.. Harder to accomplish if you need multiple tube passes, but gets you away from the most expensive portions of the fabrication and puts you into a material which is immune to crevice corrosion rather than merely resistant to it.

 

[plantprowler]

@EdStainless

I noticed your Plymouth Tube signature: Are there any Stainless Alloys that might fit this duty? I've heard of something called SS316-Ti. Also the duplex varieties?

Would any of those fit my brine duty?

 

[Christine74]

I've seen fixed tubesheet exchangers like the one shown in the sketch above. I'm pretty sure Graham Corp. in Batavia New York used this design on some of their condensers. Not preferred since you no way to change the shell-side gaskets.

-Christine