2 phase flows to/from heat exchangers

[Glenfiddich]

ok, this is a subject which I don't think alot of process engineers care too much about. Why would you design a series of heat exchangers so that it creates 2 phase flow before it enters a flash drum? For example, you have 3 heat exchangers in series and then a flash drum. After the 2nd heat exchanger the flow becomes 2 phase. The typical engineer would then size the 3rd heat exchanger for 2 phase flow.

Why on earth would you not put a backpressure control valve just upstream of the flash drum to keep the fluid in 100% liquid state? The extra backpressure would make the 3rd heat exchanger sizing much smaller than if it was sized for 2 phase.

 

[CJKruger]

Glenfiddich,

Adding a back pressure controller reduces the flow area but increases the heat transfer area.

But either way, these effects are not important. My view is that you shoould only use a back pressure controller if you have to. And the only two reasons I know of is:
- Prevent flashing before/after exchanger to reduce corrosion (mostly for H2S).
- Prevent slug flow in the vertical line after the exchanger.

If you do add a back pressure controller, you have to make sure it keeps the fluid in liquid phase even at clean and turndown conditions.

 

[Glenfiddich]

Thanks for the reply. I appreciate it. So you basically agree that putting backpressure to keep the process fluid in liquid state will reduce the heat transfer area requirement compared to a 2 phase fluid. Wouldn't that help to reduce the capital required for a new heat exchanger? Seems important to me. Why would this design be dismissed as a "do it if you have to" instead of doing it to reduce capital cost of the heat exchangers?

 

[EmmanuelTop]

What is your concern in having two-phase flow through the exchanger?

 

[Glenfiddich]

Applying a small amount of backpressure would make the fluid 100% liquid. That would reduce the amount of heat exchanger area required than if the exchanger was sized for two phase.

It's basically a matter or economics. Why design a heat exchanger for 2 phase when you can design a smaller one for 100% liquid phase by placing some backpressure?

It's a more efficient design unless you need alot of pressure to keep it 100% liquid.

 

[NghiaPP]

The heat transfer area A is calculated as: A = Q / U / LMTD
With back pressure the outlet temperature will be high, which lowers temperature difference LMTD. Convection heat transfer coefficient U is normal lower than vaporisation U. Heat transfer area is therefore larger with back pressure. Vaporisation may require large flow area per pass to avoid high pressure drop, but it does not relate to the total heat transfer area.

 

[Glenfiddich]

That's assuming Q is the same. Heat transfer to a liquid is more effective than heat transfer to a gas.

 

[Homayun]

Glenfidish, of course the Q must be equal in both cases...
But It's true that with a total Liq stream the LMTD will be lower but at the same time, I think the U-value will be greater for a 100% Liq stream than a 2 phase stream.

AFter all, we all know that liquids have much beter heat transfer characteristics than gases! So I think although the LMTD will drop, but the increase in U-value is probably much higher...

But, I think it's always more economical to have a 100% liquid stream than having a 2-phase stream..We need vapor only in a drum or a tower where we wanna separate it from the liquid...We don't care about vapor in a pipe!

 

[Glenfiddich]

So why the hell do so many engineers design heat exchangers for 2 phase flow when it could easily be avoided and you end up with a smaller HX and better heat transfer?

 

[NghiaPP]

The selection is dependent on many factors.
Suppress boiling give high outlet temperature, which may require a pressure higher than the available pressure. Design pressure of heat exchanger will be higher.
Heating medium must have higher temperature, which may not be available.
Of course two phase flow sometime cause vibration problem in piping. However, the problem normal can be avoided by carefully design.

I'm not agree about the U value. Please refer to:

http://www.cheresources.com/uexchangers.shtml

For heating liquid by steam : U=300-1200 W/(m2 K)
For evaporation by steam: U=900 - 3000 W/(m2 K)

 

[ChemEng84]

in our plant we faced another problem with 2 phase flow in heat exhcangers for NGL. it caused raprure and cracking problems in our Plate heat exchangers.
we solved this problem by placing a small knock out drum upstream the heat exchanger to separate the liquid from vapor, the liquid is then injected in a mist form in the heat exchanger which provided us with a single phase flow of gas in the heat exchanger.

 

[djack77494]

I view the initial post as asking about a liquid heater that might be designed to have some vapor in the incoming stream, and additional vapor in the outlet. If that's the case, then yes, I would find that non-optimal. I would prefer to separate the vapor from the liquid and have 100% liquid entering the exchanger (or vaporizer). Any individual exchanger can do a better job of heat transfer (i.e. minimum area) if vaporization occurs in the exchanger. But vapor entering the exchanger reduces heat transfer relative to liquid, as especially relative to vaporizing liquid. Therefore, in your network of heaters, all but the last exchanger should handle liquid only. If vapor is generated in other exchangers, it would be good to separate the vapor from the liquid, and send only liquid to the next exchanger.

Many times in making revisions to an existing facility, the poor process engineer who is being criticized for his "accceptance" of this poor approach really has little choice. Having 2 phases enter and leave an exchanger is certainly not a terrible situation even though I would be inclined to avoid it IF I were working on a new "grass-root" design.

 

[Glenfiddich]

Amen. I even try to explain this to folks who have 20-30 years of experience and they look at me like a deer on the highway. This one revamp had the main process fluid going into two heat exchangers in series as two phase in and out and then a trim steam heater. I placed a backpressure control valve on the outlet of the heater to only 100 psig and we didn't even need the trim steam heaters.