U-Bundle Heat Exchanger 9

[Inchtain]

Good morning,

We have one U-Tube Heat Exchanger where the tubes made of SS (SA 789) and others like baffles, tie rods shell and caps are made of CS (ASTM 516 Gr. 60).

First, what is the purpose of using the two dissimilar metals for the design?

Secondly, and because of this design the baffles and tie rods were heavily attacked by galvanic corrosion after years of operation, their replacement is deemed necessary now.

Saying that, and in order to insert the new baffles and tie rods to the good condition tubes, I propose to cut the tubes slightly before the tube sheet. So, is that practically acceptable? and what effects on the process and heat transfer efficiency after shortening the tubes by few centimetres?

Cooler details:
Design pressure: Shell 38 bar and Tubes 55.5 bar
Design temperature: Shell 70 degrees Celsius and tubes 170 degrees Celsius
Fluid: Shell Cooling Water and tubes HC+H2O
No. of passes: 1 for the shell and 2 for the tubes.
Capacity: Shell 430 liters and tubes 150 litres
Tubes: Diameter 19.05 X 1.65 - 6096 long.

Best regards,

 

[0707]

Inchtain

to mix SS with CS has been a disign error, or a cost price reduction, soon or later it was inevitable to have galvanic corrosion. What you intend to do to insert the new baffles and tie rods to the good condition tubes, proposing to cut the tubes slightly before the tube sheet, the process and heat transfer efficiency after shortening the tubes by few centimetres would be somewhat reduced. In my opinion to make a new tube bundle will cost almost the same as your proposal and the time consuming for the both options won´t be much different you have to balance the weight of the two options and decide accordingly.

regards

luis

 

[MJCronin]

Can you post the TEMA datasheet and perhaps some pictures of your massive corrosion ?

Some Observations:

1) Having your uncoated, ungalvanized carbon steel components attacked by plant cooling water system should be expected after years of service. Has the HX been "repurposed" since its original design ? Is t he rest of your cooling water system corroding ?..... Are other parts of your cooling water system made of raw carbon steel ?

2) Use of a U-tube TEMA-style heat exchanger for Cooling Water service is an incredibly bad choice. Using carbon steel internals compounds the nightmare. Was your HX originally designed for another service ?. Normally, a cooling water HX would be a TEMA BEM design which is easier to internally coat and periodically clean.

3) The better configuration for a S&T HX in this service is: Process fluid on shellside; Cooling water on tubeside. This enables easier coating and cleaning

4) These people have been making and selling HX internal coatings for decades:

https://www.curranintl.com/tubesheet-epoxy-cladding

https://www.heresite.com/

It sounds like you are simply replacing corroded carbon steel components with the same carbon steel. ......... Is that true ?

MJCronin
Sr. Process Engineer

 

[EdStainless]

With a u-tube HX you want the shell side fluid to be non-corrosive and lower pressure. When you don't have both of those criteria met then you should look for a different style.
Properly treated process cooling water should work with mixed CS/SS for 20 years of longer. But if you cut corners on water treatment then forget it.
People do shorten these all of the time, you only loose a couple percent of heat transfer.
But if you rebuild with SS baffles and tie rods then all of the galvanic issues will be pushed to the shell, not good. The inside of the shell would need a very good coating.

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

 

[Inchtain]

Greetings,

Thank you for the useful comments.

Kindly find some photos of the case.

It was designed according to TEMA R-API 660 and ASME VIII Div.1 90's editions, and the TEMA DS is missing.

The cooling treatment has experienced some failures and chemical injection shortages throughout previous years. That accelerated the corrosion inside the closed CS cooling water loop in the plant. The HX did not experience a change of service and it has been used as a compressor cooler.

Yes, our intention is to replace the corroded parts with similar material as the shell and piping are in CS.

Regards,

 

[EdStainless]

Those baffles look very thin, I am accustomed to ones 3 or 4 times that thick.
I have seen many units with SS tubes and tie rods with CS baffles. The SS tie rods keep the thermal expansion of the supports the same as the tubes.
The real long term risk to using CS baffles with SS tubes is that the holes corrode. This either causes the tubes to bind in the holes and greatly increase stress, or it causes the holes to enlarge and then the tubes are not properly supported and they can rattle and fatigue.

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

 

Cheaping out on the little bits, classic design failure.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

And treating cooling water as a 'non-essential' utility.
They wouldn't tun with a knowingly contaminated feedstock, but let water treatment fail and they just keep going.
Never mind that each hour is taking weeks off of equipment life.
Since it doesn't cause a problem 'right now' they think that it is fine.

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

 

[Inchtain]

Hello,

Thank you for the interesting discussion.

So, which is the best approach here? In your opinions, to what extent changing all parts material to SS will solve the dilemma?

Best wishes,

Ibrahim ALKANSHI

 

[r6155]

@Inchtain
1) bundle with all SS materials.
2) check actual thickness of CS shell side and nozzles.

Regards

 

[saplanti]

Inchtain,

Stainless steel with stainless steel will have galling problem. Before making decision I would suggest you think about how you will dismantle the threaded joints. The same think may happen with baffles too.

 

[EdStainless]

I see CS baffles used with everything else SS, and I see all SS. It depends on the the service and other factors.
If you go to SS (at least for tie rods) they need to be the same alloy as the tubes.
With a duplex SS I am not too worried about galling.

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

 

[r6155]

Galling in the tie rods with nuts and tubesheet, what is the problem?.

Dismantle of bundle is not a problem.

Regards

 

[saplanti]

Please check peges 8 and 9 of the following document for the galling of stainless steel.

https://www.diva-portal.org/smash/get/diva2:171318/FULLTEXT01.pdf

 

[EdStainless]

Sorry saplanti, that has no info on the duplex alloys.
The loads that it takes to gall 2205-2205 are about 10x those for 304L-304L.
Outokumpu has published good data on this but I can't find the link right now.

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

 

[r6155]

I know what is galling

Normally the tie rod have double nuts to assure they will not vibrate and come loose. Hence galling is beneficial here.
Some fabricators use one nut and tack weld.

I assume tie rod is 3/8". Duplex SS for nuts 3/8" ????

Regards

 

[EdStainless]

3/8", that sound light.
Most shops would use tubes for spacers and run 1/2" rod down them as ties.

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

 

[r6155]

Please see TEMA Table CB-4.7.1. Number and size of tie rods.
No diameter less than 3/8" shall be used above 15" nominal shell diameter.

Regards

 

[Christine74]

It's not at all uncommon for the baffle material (as well as other components that are only exposed to the shell-side fluid such as tie rods, spacers, skid bars, seal strips, dummy tubes, impingement plates, etc.) to be of the same material grade as the shell cylinder. If carbon steel is suitable for the actual pressure boundary, why wouldn't it be okay for the non-pressure components like baffles and the tie rods?

If this were a large diameter tube bundle with hundreds of tubes it would probably be worth it to salvage the existing tubes and tubesheet and replace the baffles/tie rods/spacers. But from the photos it looks like this exchanger has less 50 "U" tubes so it might not be worth the effort.


-Christine

 

[r6155]

See API 660
8.1.6
Shell side components (baffles, tie rods, etc.) shall be manufactured from material with equivalent chemical composition and corrosion resistance to those for the tubes unless otherwise specified on the datasheets.

Regards