5 replies to this topic

[kkala]

We are currently looking into existing tubular exchangers to see whether these can be used for crude desalter brine cooling (140-38 oC), using cooling water (31-41oC) or other refinery wash water. Liquid velocity in exchanger tubes should be higher than 1 m/s at the design conditions, according to a rule of thump, to avoid frequent scales in them. Does this also hold for the shell? Liebermann says so in his "Process design for reliable operations" (title may be a bit different), but there are opposite opinions. Yet it may be reasonable that velocity limits for either shell or tubes should be same. Can you please advise? What would be the consequence if brine in shell had a velocity of (say) 0.3 m/s at design conditions?

[Art Montemayor]

Kostas:

I would place a great deal of weight and respect on advice from Norm Lieberman. He has accumulated a vast amount of hands-on, field experience and what he writes is a result of many years of that know-how.

I thoroughly agree with keeping Liquid velocity in exchanger tubes higher than 1 m/s. I have gone as high as 4 m/sec in some applications, but I would not apply this in all cases. You must remain aware of any potential for negative effects from high velocities. For example, be aware that the material of construction has a lot to do with what velocities are to be tolerated. Excessive velocities can lead to destructive erosion.

My basic premise for deciding what velocities I employ in a heat exchanger has always been determined by what pressure drop I can afford to take across the heat exchanger – through the tube side as well as through the shell. You will find that it is quite easy to increase velocities in the tube side and, consequently, increase the heat transfer coefficient there. However, the shell side always presents velocity problems because you have such a relatively large cross-sectional area and you usually can’t afford to take a heavy pressure loss on the shell side. I normally always try to use as much pressure drop as I can get away with in the shell side – and that fixes the maximum velocity that I can apply there.

It has been my experience, in 50 years, that there is NO REASON to expect that velocity limits for either shell or tubes can be same. I can’t tell you what would be the consequence if you applied brine flow in a shell side at a velocity of 0.3 m/s. But I can clearly state that I would never do that without a very demanding reason. I would rather study the option of putting the brine in the tube side and increasing the velocity. You haven’t given us enough basic data to make a very detailed comment. For example, you haven’t told us what you are doing and what you have as the secondary fluid.

Higher velocities in heat exchangers will always yield better heat transfer – but at a cost of higher pressure drops. Therefore, it always works out that you have to weigh the trade-off against the potential gains. But one thing is for sure: regardless of what the pressure drop limitations are, you must first respect the fact that your fluid may ruin your heat transfer operation if left to deposit solids, create undesirable residues, or increase the fouling. And this, I believe, is what Norm is trying to warn you about. I believe this is valuable advice.

[Zauberberg]

Hello Kostas

I am attaching a photo of Desalter Brine/C.W. exchanger, employed in exactly the same service as the one you described. The photo was taken some 5 years ago during a major turnaround, while I was working in CDU/VDU.

As you can see, these exchangers are condemned to fouling due to the nature of their application: cooling the brine that contains suspended solids, oil (especially during upsets the quantity of oil can be extremely high), being quite frequent with higher/lower pH than normal etc. There are many other factors that contribute to fouling/scaling/corrosion. Also, controlling the interphase level in desalters has always been a challenge - meaning that there will be frequent oil carry-under with the brine, and that probably fouls the exchanger the most.

I think that higher velocity will not save you (or your client) from fouling. Or, putting it this way - lower velocities will not result in extremely higher fouling rates because the system is already prone to bad fouling. Being discharged from the desalter, I presume there is a lot of pressure drop that can be consumed/utilized in heat transfer but, again, the echanger is already there and cannot be modified to result in higher shell-side velocities. So you are between rock and a hard place in any way.

I would think of two options - either to consider placing the brine on tube-side of exchanger (as suggested by Art) and see what velocities you will come up with, or to consider employing two parallel filters upstream of exchangers - if fouling is a big concern for your client (high discharge temperature of brine downstream of HEX could be an issue). For such reason, you may consider employing a trim-cooler (fin-fan) downstream of heat exchanger to give you some additional margin. I have seen these things working. The key question to answer is: what operational time between two cleaning cycles of the exchanger you need to ensure? How much time the exchanger needs to be in operation, and at what maximum brine discharge temperature?

Good luck,

Attached Files

•  DSC00073.JPG   347.57KB   53 downloads

[kkala]

My sincere thanks for your answers, covering not only the main point but also other aspects of the issue .
Extending the abbreviated information of first post, purpose is to chose suitable exchangers among idle existing ones (or specify new) to cool the desalter brine from 140 to 38 oC in order to send it to Waste water Treatment. At least two exchangers in series will be used: the last using cooling water (31-41 oC, sort of trim cooler) the others using water from "sour water stripper"(SWS, max 50 oC) mixed with raw water, that goes after being heated to the crude heating train just upstream of desalter.
Your posts made it clear (before it was not for us) that brine has to pass from tubes, even if raw water can also create scales. The photo (really "a thousand words") apparently shows a case of brine passing from shell. Frequency of scale cleaning (in function of velocity) may not be possible to guess, but we have to consider a reasonably high velocity in shell and be informed of scale severity from Client. ΔP for each exchanger can be high enough (we also specify the SWS water feeding pump).
Liebermann advises 10-30 ft/s. His style is exciting, I could recall his introduction to "not less than 3 ft/s" after 25 years! And I wish I would have had his adventurous industrial experience.

[Zauberberg]

I have bad experience with dirty fluids flowing on the shell-side of an exchanger. I know many people would argue against this observation, but in reality this is what happens: never place dirty, viscous, or corrosive fluids on the shell-side. It doesn't need to mean that they will not cause fouling if they flow through the tubes, but at least cleaning of tubes can be done much easier than cleaning outside the tubes.

Desalters usually operate at pressures 7-12 barg (my experience) so there should be more than sufficient pressure drop to be utilized for heat transfer, and to maintain high velocity inside the tubes.

You are right about Lieberman. I had that lucky moment at the beginning of my career to attend one of his famous process troubleshooting courses which really give you a blast of inspiration and enthusiasm. The point is that he keeps the things as simple as they could be, and that is why it is so easy to understand the concept, and apply it on every-day problems in plant operation. I wonder if he's still alive - I guess he should be because we haven't heard the opposite.

Best regards,

[Art Montemayor]

All:

I believe Norm Lieberman is very much alive and still fighting the good fight against bad logical thinking and organizing tactical and smart process trouble-shooting. The last time I checked, he was still doing business in Metairie, Louisiana – but that was before Hurricane Katrina. His webpage is still functioning and I’ll bet if you try to contact him, he will personally answer - or his personable and cordial wife, Elizabeth (also a Chemical Engineer in her own right), will respond to you quickly. Visit both Norm and Elizabeth at:

http://www.lieberman-eng.com/personnel.htm

You will be pleasantly surprised that you can obtain Norm’s famous seminars from his Video Library. I highly recommend this source of top-grade engineering expertise for young chemical engineers out in field.