Gas Recovery Plant (absorbers simulation) 1

[Lblopez]

Hello,

In my job, I'm learning about a gas recovery plant, this plant have three absorbers, in the first, gas feed is a C5- mixture. All C3+ are removed by absorption with naphtha. Top product is C2-

The question is, in 2 desing I've noted that the feed (gas & naphtha (liquid)) temperature inlet is lower than temperature outlet products. Is like heat generation occurs. Why is it? Can I talk of mixture heat in this process? May be absorbtion heat? Why, If I have ligth hydrocarbons? How can I explain it?

In my PRO/II simulation, this not occurs, there the gas effluent is colder than gas feed and the liquid effluent temperature is hotter than inlet. Like I think that would be :-/
What most be changed in my simulation to make in consideration this heat generation? How can I simulate it?

Thanks,

Lblopez

 

[DickRussell]

When you have gas-liquid contacting in an absorber or stripper, there is movement of material both into and out of the liquid. In a textbook absorber example, the lean oil will be considered nonvolatile, and the absorption of lighter (but not too light) material from the vapor results in the mixture getting warmer. You've taken something from the vapor phase and made it part of the liquid; the latent heat is released upon condensation/absorption into the liquid.

There also is the effect of the vapor feed stripping some material from the "lean oil." If the lean oil is really heavy, nearly none of it will be vaporized into the uncondensable vapor leaving the top of the column. But if the lean oil hasn't been stripped well downstream before being recycled to the absorber, or if the oil hasn't been selected very well and actually has a significant amount of lighter components that gets stripped out by the vapor rising up the column, then the heat required to do this change of phase must come from the mixture, and a temperature drop results.

Assuming the VLE methods have been selected properly for the modeling, the results just reflect how the two streams come to equilibrium in the various stages of the contacting, and the temperatures reported simply reflect the net heat effects that have occurred from material changing phases.

 

[Lblopez]

Ok I understand now,
What I can do in my simulation to take in consideration this heat in PRO/II?

Thanks.

 

[DickRussell]

Assuming you have selected thermo methods appropriate for a mixed hydrocarbon system, such as SRK or PR for VLE and SRK, PR, LK for enthalpy, then the heat effects are built into the calculations. The phase enthalpies are temperature, pressure, and composition-dependent, calculated by departures from the ideal gas enthalpy for the vapor and liquid mixtures. You shouldn't have to deal with any "excess enthalpy" (heat of mixing)effects for this type of system, as you might expect for a very nonideal chemical mixture.

Keep in mind also that in a sense an absorber is a counterflow heat exchanger, typically cooling the gas down to some approach to the inlet lean oil temperature and heating the lean oil as it flows down. Superimposed on this are the heat effects due to material changing phase. You might look closely at the stage compositions in the model output file, to see what is being "condensed" (absorbed) into liquid or vaporized out of the lean oil. Most of the absorption is likely to occur near the bottom of the column, and most of the stripping of volatiles out of the lean oil tends to occur near the top. If you are getting an apparent autorefrigeration effect near the top, look for volatile stuff stripped out of the lean oil up there.

Then, too, is the matter of proper design of the column. For absorption of a component, look at the "absorption factor" for, say, the C3 components (A=L/VK, where L and V are molar phase rates and the K value is for that component on the stage). If you can find a Brown-Souders chart that shows fraction not absorbed vs. absorption factor, you'll see that for good absorption (0.99, 0.999) in reasonably few stages (4-6), you need an absorption factor of around 1.2 or better. Below 1.0 won't give good absorption no matter how many trays you add. Since your gas rate is fixed, the only things you can do to improve absorption is increasing L (more oil) and decreasing K (colder oil and perhaps gas also). Similarly, the "stripping factor" is the inverse of the absorption factor, and the Brown-Souders chart shows fraction not stripped vs. stripping factor. For some component in the lean oil, it will be stripped out of the heavier stuff more for increasing V/L and K (in this column, less oil, higher temperature).

As described, you want to recover C3 but not C2. What you are essentially doing is distilling between the two, using heavy components to take up vapor pressure space and get the job done without cryogenics. You have a relative volatility between the C2 and C3 to consider (somewhere around 3, depending on conditions), so if you are getting too much absorption of C2 along with the C3, then you have to play around with the configuration (reboiled absorber, etc.). If the configuration is fixed, then you can at least explore your range of options on gas/oil temperatures, flows, etc. and see the limits on what C2/C3 rejection/recovery you can expect.

Hope this helps.

 

[Lblopez]

Thanks Dick, Very thanks.