Alternative material for Unalloyed Titanium in Heat Exchanger Tubes 2

[rifa92]

Dear All,

We have a vertical heat exchanger that comprises of seven boxes. These boxes are stacked over each other.

The tubes are made of Titanium ASTM B338 Grade 2 and boxes are cast iron. The life of tubes is approximately five years and we would like to choose material that will provide longer life.

Operation:
The function of this heat exchanger is to cool the gases and heat the liquid.

The gases enter from bottom and exit at top i.e. the seventh box. The gases temperature at entry is 100C plus and leave at 60C plus.

Composition of gas:
Ammonia NH3 = 0.402 kg/m3
Carbon Dioxide CO2: 0.255 kg/m3
Water H2O: 0.128 kg/m3
Hydrogen Sulfde H2S: 0.003

The fluid in boxes are different.
One fluid liqour enters the 2nd box and exits from 1st box.
Second fluid enters the 5th box and exists from 3rd box.
In 6th and 7th box cooling water is used.

Fluid composition of 1-5 boxes is similar.
NH3: 2.4-2.8%
NaCl: 6.8-8.1%
NH4Cl: 16.2-18.2%
CO2: 4.2-5.3%

The life of 1-3 boxes is the least, this is probably due to higher temperature of gases in this region. The tubes thin out and mostly at the u-bends. So, erosion has to play a part as well. Tube OD 1inch and thickness 20 SWG = 0.914mm.

 

[EdStainless]

So how long do the other boxes last? Are they good for 10 years?
You don't have a lot of options.
There are a few but there are issues with all of them.
Are you just looking to stay with the same design and change the tubes?
Is there a chance to redesign the tube bundles and move to 1.25" tubes to reduce the velocities?
You could also move to 20ga min wall and add 10% more metal.
Where are the tubes from? Do you have MTRs? What are the mechanical properties?
The first step would be to look at 2H or 3 and see if they are enough stronger to help.
These would be fairly low-cost options (15-25% more).
The easiest and most expensive is to move to alloyed Ti grades.
Such as 16H, 17, 26H, or 27.
These have PGMs added and they repassivate much better than CP grades.
This assumes that corrosion is playing a role in the metal loss.
These options will be 2-3x the cost of gr2.
There are number of high alloy SS grades and some Ni alloys that would work.
But these options involve a lot of thermal design to assure performance.
These alloys would have an order of magnitude higher erosion resistance.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[MJCronin]

rifa ... As always ..Not enough information

How about an outline drawing or flow sketch of your multi-stage HX?

- How do your tubes fail ?... Can you show a picture or sketch ?

- Is there failure at your tube/tubesheet joint ? What is the material of your tubesheet ???m

- Is the tube failure at the entrance/exit to the shellside ? (rho-v squared failure ?)

- Do all of your failures occur at one "box" of your HX and could this particular box be constructed of a thicker, higher grade titanium or other material ?

I agree with Ed .... IMHO, I would be talking to the Ti salesman about prices for all of the thicker and more superior gardes of titanium alloy. Perform a calculation to evaluate excessive gas entrance velocity

There is nothing to prevent you from getting budegetary costs for HX upgrades for various premium tube configurations

Oh, and whatever you do ....., STAY WITH YOUR TRADITION AND NEVER, EVER SAY PLEASE OR THANK YOU !!!

MJCronin
Sr. Process Engineer

 

[rifa92]

Thanks for your replies. The exact damage mode I haven’t not witnessed myself but from information gathered, the tubes leak at the bends. I have been studying and found that gr 2 is susceptible to under deposit crevice condition in the operating condition we are using and have recommended grade 12 instead. Price of which is 2x. This corrosion rate increases with temperature. We are facing issues in the boxes where temperature is higher.

 

[TugboatEng]

If you're failing in the bends I would suspect erosion. Your flow rate may be too high and you need to use a larger exchanger.

Your description of the water sounds as if it is silty. This will cause abrasive erosion.

A thought. Get rid of the bends. Size your exchanger so that it can be a single pass OR do the reversal in the header. You can apply a ceramic filled epoxy coating that is serviceable to the header to prolong exchanger life.

 

[MJCronin]

I agree with Tugboat .... A straight-thu STHX may be a good solution .. Perhaps TEMA BEM ?

Suggest that you pay the small price and have the eroded tubes evaluated by a professional materials lab

A question: .... Why do you need a "U-Tube" HX in this situation ?

Was your troublesome heat exchanger once used in another service and now you are paying the price of a long forgotten "savings" from the re-use of junk ?

MJCronin
Sr. Process Engineer

 

[EdStainless]

I second what Tug and MJ have said.
If this is erosion get rid of the ubends.
And what makes you think that gr12 is the right option?
You say that none of your temperatures are above 100C.
What are the deposits?
What is the actual pH at operating conditions?
Gr2 is very soft and will fail by both erosion and vibration easily.
What is the actual flow velocity?
If erosion is your real problem then you will not solve it by changing Ti grades.
You need to go from 5 yr to 20 yr, not just to 8-10.
This is a design issue.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[rifa92]

Apologies for such delayed response. Responding to EdStainless PH of fluid in tube is 7.5. Discussion on deposits below.
Now that one of the faulty exchanger boxes got replaced, I got a chance to look into the failed tubes. However, due to brittle nature and small thickness it is difficult to tell where the tube actually leaked. There are several cracked regions however. Following are the observations:
1. Deposits can be observed on the internal surface of tube of whitish and brownish nature.

2. No leak points at bends found. Instead cracking found at different region of tubes.
3. The thickness of tube has decreased significantly. I took thickness at different areas and it was around 0.3-0.5mm. Original thickness in around 1 mm.

Based on previous online review I did, I found that Grade 12 has better performance compared to grade 2 in Brine environment.

 

[sahsanb]

Have you considered bimetallic tubes such as Omegabond which have Titanium as substrate and Zirconium as top layer?

 

[rifa92]

No we haven't explored it. We don't to modify the material as it will affect performance of heat exchanger.

 

[blacksmith37]

A professional failure analysis is needed. You have a problem if Ti type 2 is corroding. Hydriding under certain conditions is one of the few failures we have seen in
ti- 2 . The other is accidental exposure to HF.

 

[EdStainless]

erosion and fatigue.
All Ti alloys are poor at resisting both of these.
Hire someone to analyze vibration in these units.
I would guess that you are running high flow velocities, and the deposits are causing the erosion.
Find out what the deposits are.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[rifa92]

Thank you EdStainless, if it were erosion shouldn't it have affected the bends more?

In our scenario, the bends are in better condition compared to the straight portion.

 

[EdStainless]

Not usually, the added turbulence at the entrance/exit is where erosion is most oftend an issue.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed