Coefficient on tubeside of HX

[Alwynb]

Hi there,

I am designing a shell and tube hx (Tema type BKU)to cool glycol from -17 to -21°C with NH3 evaporating in the shell side (450kW).

I get a very large hx where the tube side film resistance is 93% of total resistance. Is there a way to improve the tube side coefficient in order to reduce overall size?

I see BJAC refers to twisted tape inserts inside the tubes. When I use this option they actually make quite a difference. Is this the only method to reduce the resistance?

Regards
Alwyn

 

[speco]

Alwyn,

You have two other possibilities that I see:

1. Increase the tube velocity considerably. This will cost you pressure drop. Long and skinny designs may help, if you have the real estate available.

2. Reconfigure the exchanger to put the more viscous fluid on the shell side.

Regards,

Speco

 

[25362]

Since you aren't considering a change in HE design, there are few alternatives to improve performance.

Speco is absolutely right on his first advice. Since the HTC is a function of Re^0.8 one needs to increase the velocity to improve it.

To reduce the size of the unit another point would be to reduce the evaporating pressure on the refrigerant side i/o to increase the "delta t" to a practical maximum.
Now, i/o to cool the glycol down to -21 deg C you might have to operate under a vacuum. The atmospheric BP of ammonia is -33 deg C.

If you don't like to reduce the pressure so much, you may switch to using propane as a refrigerant, its vaporization temperature at atm. pressure is -40 deg C.

Good luck.

 

[srfish]

It appears that the glycol is in laminar flow. If this is the case, you will not be able to get anywhere near the heat transfer in bare tubing that you get with using a static mixer like twisted-tape inserts.

With bare tubes,if the Reynolds number is between 1000 and 5000, the heat transfer coefficient can be increased by using shorter tubes and more tube passes. The shorter tubes results in reducing the L/D correction in the heat transfer coefficient equation. The resulting benefit of less surface can be more than the increased cost of a shorter bundle and larger kettle.

 

[25362]

If I am not mistaken glycol (Ethylene glycol) solidifies at -13^oC. So, either this is another chemical or it is a solution. Kindly verify and explain.

 

[Montemayor]

Alwynb:

25362 raises a good point. You can lower the operating pressure of the ammonia refrigerant in the kettle's shell by simply controlling the suction pressure at the refrigeration compressor. By maximizing the delta temperature, you maximize the heat transfer rate in the BKU coil. Presumably, you are using a closed ammonia mechanical refrigeration cycle to furnish the shellside cooling effect.

However, I would not operate the shellside at atmospheric or vacuum conditions with the ammonia because of the constant need for non-condensable purging on the NH3 side. This was once a practical nuisance and now considered an emmissions problem that may necessitate vent scrubbing or other equipment. It is a practical impossibility to maintain a vacuum (or even atm. pressure) with a refrigeration compressor without sucking in atmospheric air. I am assuming conventional compressor seal technology here. That means that you're limited to a practical -25 oF (-31 oC) design with the NH3. I would not consider taking propane (or any other combustible hydrocarbon) down to atmospheric or vacuum conditions because of the hazard of sucking in the Oxygen content in the air while using conventional seals. The O2 buildup in the suction and the subsequent heat of compression in the propane refrigeration compressor can present a dangerous explosion situation. That is the reason why hydrocarbon compressors are never allowed to operate without a positive pressure control (& alarm) on the low suction side. Usually, a 5 psig suction pressure is established as the lowest allowable design pressure at that point.

I would expect a "large" BKU size - that is the nature of the beast - especially if you are using one U-tube bundle. But if you stuck with that particular type of TEMA design, you may have to live with it. Are you free to select another type of TEMA design?


Art Montemayor
Spring, TX

 

[25362]

Still looking forward to receiving an explanation on the "glycol" to be cooled below its solidification point.

 

[speco]

25362,

On your question about the freezing temperature of ethylene glycol. I believe it's a bit colder than that for 50% EG. According to Dow literature, the freezing point is around -25F or -31.7 C.

Speco

 

[25362]

to speco: neat EG, as Alwynb is referring to, is not the same as 50% EG. While EG (used as an anti-freeze) itself freezes at -13^oC, various solutions of EG have different freezing point depressions relative to water. A 60% mass solution in water shows a value of -51.2 ^oC. Viscosities are also not the same as for neat EG. Thus either, Alwynb speaks of an EG solution or of another chemical altogether. Let him clarify the issue. OK ?

 

[virk]

Decreasing evaporation temperature of ammonia not only increases LMTD but also increases overall heat transfer coefficient. In my opinion you should think about evaporating ammonia at approx. -28°C. What overall heat transfer coefficient does BJAC calculate. If it is already in the range of 0,9 kW/m2K i do not think that you can get it much higher.
I think pressure drop on glycol side should be roughly 1 bar; so we design our evaporators (just rough numbers)

Why do you use shell&tube, ...because of maintenance aspect?

Kind regards

Virk