Tube Side Velocity Limit for Shell and Tube Heat Exchangers

[Pavan Kumar]

Hi All,

We are designing a shell and tube heat exchanger to be used as gas cooler for my plant. The tube side is the process gas at 26 psig to be cooled from 220 Deg C to 175 Deg C using Boiler Feed Water(BFW) on the shell side. I have a question on how much can the tube side velocity be increased while being within the allowable pressure drop without reaching the erosional velocity limits.

Based on the preliminary Aspen EDR runs the tube side velocity is 60.97 ft/sec ( 18.5 m/sec) and the tube side pressure drop is 0.19 psi while the allowable pressure drop is 0.74 psi. There is up to 0.54 psi of pressure drop that can still be utilized to convert it to heat transfer to reduce the heat transfer area but want to make sure that we don't hit the erosional velocity limit. The thread below says a velocity limit of 25 m/s (82 ft/sec) for carbon steel tubes for the air cooled exchanger. In my case the tubes are Duplex 2205. Does this limit of 25 m/sec apply to Duplex steel as well?. Also I want to know how does the tube side Rho. V^2 limitation play role in the tube side velocity limitation. Kindly quote the references which gives guidelines on these.

https://www.eng-tips.com/viewthread.cfm?qid=451490

Thanks and Regards,
Pavan Kumar

 

[EdStainless]

How clean is the gas?
Considering that I have seen applications running water in 2205 tubes at 4-5x the velocity that would be safe in steel I see no reason not to push velocities higher.
Though I have to ask if there is reason that your heat exchanger is built backwards? More typical would be the gas on the shell side and BFW on the tube side.

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

 

[Pavan Kumar]

Hi EdStainles,

The gas is reactor product gas as it is corrosive hence we put it in the tube side. The gas is not very clean in the sense there are tars that flow along with the gas. Plus the existing Gas cooler which we are DE bottle-necking has process gas on the tube side. The process gas is at 220 Deg C. We have seen tar deposition in the tubes during the operations. Can I go up to 25 m/sec as long as my allowable DP or erosional velocity is not reached?. The tubes are Duplex 2205.

Thanks and Regards,
Pavan Kumar

 

[EdStainless]

The tubes will take it.
But think of what the deposits will do inside the tubes. I'll bet that it doesn't take much deposit in the tubes to reach your DP limit.
You need to think about building a square pitch tube bundle with the dirty gas shell side.
OK, it means a 2205 shell, so what.
IT will be easier to clean and maintain.
Your velocity limit will be based on cross flow vibration issues.

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

 

[don1980]

To ensure that this exchanger actually works as intended (gas is cooled to 175C) you want the velocity (i.e. turbulence) to be as high as practical. That is, you want to generate a design that reliably consumes all of the allowable pressure drop. Gas side HTC is highly sensitive to Nre. Duplex 2205 is a very hard metal, thus it has a high resistance to erosion. This material will tolerate a far higher velocity than carbon steel, which is relatively soft. I see no problem with a velocity of 100-150 ft/sec.

 

[pierreick]

Hi,
Consider the doc attached to support your work
Good luck
Pierre

 

[Pavan Kumar]

Hi pierreick,

I have this paper and read it multiple times. It gives guidance an the minimum and max tube side velocities of 1 m/s and 2.5 - 3 m/s which is applicable to liquids. It does not say anything about gases. I will use the guidance given by don1980. Thanks don1980 for your guidance.

pierreick by the way, not related to this topic, would you be able to help me with this paper if you have it by any chance.

Leung, J.C., (1996) Easily size relief devices and piping for two-phase flow. CEP, 92 (12), 28-50

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi don1980,

To ensure that this exchanger actually works as intended (gas is cooled to 175C) you want the velocity (i.e. turbulence) to be as high as practical. That is, you want to generate a design that reliably consumes all of the allowable pressure drop. Gas side HTC is highly sensitive to Nre. Duplex 2205 is a very hard metal, thus it has a high resistance to erosion. This material will tolerate a far higher velocity than carbon steel, which is relatively soft. I see no problem with a velocity of 100-150 ft/sec.

Yes definitely. I want to make sure the tube side allowable DP us full utilized to reduce the required HT area. I wanted to know how much high the velocity can be increased by reducing the number of tubes as the current design velocity is at 60 ft/sec with duplex 2205 tubes(I can't go for multi pass due to fouling/tarring nature of the gas and the current plant space limitations and large gas pipe size). As you said up to 100-150 ft/sec if ok with Duplex is ok, so I will try that and check. Besides in this case the shell side is the controlling side as it has BFW which has high thermal conductivity compared to the process gas. Having a reduced shell diameter( possible by reducing the number of tubes) improves shell side cross flow velocity which further improved heat transfer. That is path forward I am taking. Let me know if you have any further suggestions.

Thanks and Regards,
Pavan Kumar

 

[pierreick]

Hi Pavan ,
You should have been more cautious , you have information related to Head loss ! this should help you to make your decision.
Good luck
Pierre

 

[MJCronin]

Pav.....

You first stated: We are designing a shell and tube heat exchanger to be used as gas cooler for my plant.

The word "design" implies a review of something that is not yet made

According to my lexicon and understanding of English, this means a brand new heat exchanger ..

But now you mention "reducing the number or tubes", which to me sounds like a used, repurposed heat exchanger that is so common in the third world,.... is that true ?

What exactly are you doing ?..... Is anyone else confused ?

Can you post a TEMA design datasheet or perhaps an outline drawing for this existing unit ?

MJCronin
Sr. Process Engineer

 

[Pavan Kumar]

Hi pierreick,

Hi Pavan ,
You should have been more cautious , you have information related to Head loss ! this should help you to make your decision.
Good luck
Pierre

I really did not understand what you meant. The information I shared is not proprietary.

Thanks and Regards,
Pavan Kumar

 

[don1980]

Pavan,

I assume you're using a heat exchanger software program such as HTRI, or similar. Use that to gain a good mental picture/understanding of the various sensitivities and trade-offs. Recognize that trade-off are an inherent part of exchanger design, Start by identifying the core design requirements and constraints such as the range of flowrates over which the exchanger must properly function, the permissible gas side pressure drop, extent of fouling.

Recognize that the exchanger must function sufficiently over a range of conditions (e.g. gas side flow-rate/pressure-drop), range of fouling). Identify that range of flow rates and the extent of tolerable fouling. Optimize the design by running a series of different design cases (sensitivity analyses). Us up the allowable gas side pressure drop to maximize the gas side HTC. Finally, test the proposed design to be sure that it works sufficiently at all the different conditions (e.g. min flow in dirty tubes, etc.).

You need to know the fouling resistance (fouling factor) that you're currently experiencing. Otherwise you may underestimate it in the new design (leading to under-performance and/or more frequent outages for cleaning). Model the existing exchanger, using plant data when the exchanger is dirty, to quantify this fouling factor. Your new design should use a fouling factor that is this value or higher. Using a higher value will extend the service life, but recognize that that's just one of the variables that has trade-offs - it's just one of the optimization variables.

 

[Pavan Kumar]

Hi MJCronin,

We are indeed designing a brand new exchanger. I have preliminary Thermal Design done by an Engineering company which I was reviewing. Upon looking at the design I found the tube side DP that was not full utilized. But since the tube side velocity was already 60ft/sec I wanted to check how much more can increase before we reach the erosional velocity while keeping the DP within allowable limits. The reduction in the number of tubes is from the number that the preliminary TEMA Sheet had. We cannot increase the tube passes to increase the tube side velocity due to fouling issues, so the only way is to reduce the number of tubes by reducing the shell diameter. I do not how but need to increase tube side velocity. BTW the shell side velocity was also found to very low ( laminar region). So reducing the number of tubes will serve both purposes.

I cannot post the TEMA Sheet as it is company document. I will post the Schematic drawing my next message. I can share more information if it is not clear to you.

But now you mention "reducing the number or tubes", which to me sounds like a used, repurposed heat exchanger that is so common in the third world,.... is that true ?

I am sorry but I have to say that your third world comment is making me feel insulted as this is the second time this has happened.

Thanks and Regards,
Pavan Kumar

 

[MJCronin]

I am sorry that you feel "insulted" by people who are trying to help you (for free) with your heat exchanger design

This is a good description of a useful term that has been in use for well over seventy years

https://en.wikipedia.org/wiki/Third_World

What equivalent terminology would not insult your sensibilities ? Please explain and give me examples ...

Second time that this has happened ? ..... You are keeping track

Please accept my apology .....

Politically correct terminology across cultures and countries can be such a minefield...

MJCronin
Sr. Process Engineer

 

[pierreick]

Hi Pavan,
Calm down , I mentioned in my reply that the data given in the paper about head loss ( 0.05 t0 0,2 kg/cm2) ,average of 0.1 kg/m2 should be sufficient , Head loss and velocity are correlated .

https://www.sciencedirect.com/topics/engineering/side-pressure-drop

Regarding your second question I have no access to the article as such , I found another article which may interest you and others :

https://www.hydrocarbonprocessing.c...-to-the-homogeneous-direct-integration-method

Good luck
Pierre

 

[Pavan Kumar]

Hi pierreick,

I have seen in the paper by R. Mukherjee the allowable pressure drop per shell is 0.05 - 0.2 kg/cm2 (0.71 - 2.84 psi) which 0.1 kg/cm2(1.42 psi) being typical. However when I checked against my system the allowable pressure drop for my system is restricted to 0.74 psi due to limitation of the air compressor. Too much pressure drop would create very high back pressure on the centrifugal compressor which would not deliver the required air flow. So even though head loss and pressure drop are related, I found that answer is not straight forward.

Per my TEMA sheet I saw that the tube side pressure drop is 0.19 psi with tube side velocity being 60-ft/sec. Since I still have 0.55 psi ( =0.74 - 0.19)that I could use which is possible only by allowing higher tube side velocity than 60 ft/sec. But how much high was the question. don1980 was kind enough to answer that up to 100-150 ft/sec is ok for Duplex tubes in my heat exchanger. That was specifically my question and don1980 helped me get the answer.

The paper "Leung, J.C., (1996) Easily size relief devices and piping for two-phase flow. CEP, 92 (12), 28-50" is the Chemical Engineering Progress.

Thanks and Regards,
Pavan Kumar