Natural gasoline absorption with... natural gasoline

Jyrka77 · Dec 28, 2016

Hello, folks

I have a question about a curious absorption-based natural gasoline recovery system in an existing petrochemical plant.

There is an absorption column to recover natural gasoline from a gaseous hydrocarbon stream (process waste gas).
Liquid natural gasoline is used as an absorbent. So basically natural gasoline is absorbing more natural gasoline in itself. The natural gasoline is recirculating: from absorber bottom it goes to a separator (water is separated), then it is pumped back to the absorber. Another gasoline stream from upstream process(probably more heavy HC-s) is mixed to the stream before the pump and a portion of the gasoline is removed (sent to storage) after the pump. Also, there is a cooler to cool gasoline before entering absorber.

This system has never worked: after start-up of the plant, a net LOSS of gasoline was noticed, rather than increase due to recovered components. So the absorber was shut dowm (gasoline supply line was closed, gas is just passing by the column).

My question is: does such arrangement make sense? I know that usually a lean oil absorption is used, where an oil is an intermediate carrier for the natural gasoline fraction. In here, a much simpler system is built, which does not need a stripper/regenerator at all. So far I have found just one reference to a similiar arrangement from the Internet.
For me, it would make sense if the system was once-through. But since the gasoline is supposed to recirculate, shouldnt it get saturated quite fast? According to the design stream table, the gasoline makeup and draw were quite small, like 10% or less of the overall flowrate.

I tried to model this system in Aspen HYSYS, and I noticed an increase in liquid volume in the column (C1-C4 going from the gas to the absorbing liquid; C6-C8 going from the liquid to the gas), when the liquid-to-gas mass flow ratio was 1:12 with gasoline temperature 25 DegC and 1:5 with 5 DegC.

So, has anyone heard of such natural gasoline recovery scheme and what do you think of it?

Best regards
Jörgen

georgeverghese · Dec 29, 2016

The description indicates that this system does away with a regen unit to strip out the C4- fraction by the introduction of this heavy hydrocarbon stream and bleeding a portion of the resulting mix upstream of the pump.
Wouldnt this work better ( ie lower requirement for the heavy HC stream ) by removing some of the volatile rich gasoline stream BEFORE introducing the heavy HC stream ?

Have worked on rerating a lean oil absorption plant - in this case, the lean oil was pre refrigerated and a lean oil stripper was in the regen unit.

Jyrka77 · Jan 1, 2017

Hello, Georgeverghese, and thank you for the reply.

Yes, it looks like the bleeding and mixing is supposed to replace the regen unit.
Bleeding a portion of gasoline before introducing heavy gasoline makes sense, but I think then they would need two pumping units? Right now they can use only one (heavy gasoline is introduced before the pump, so it can too be pumped, and the bleeding is done after the pump, so it can be drawn under pressure.

But does the principal absorption physics make sense here? Gasoline fraction (C5-C9 alkanes) absorbing lighter components C3 and C4? What is the general rule here, heavier hydrocarbons are prone to absorb lighters? I understand that with lean oil absorption, the heavier species C5-C8 are absorbed to the oil and lighters remain in the gas? But in this case, the lean oil is even heavier, consisting of C10+ molecules?

Thank you and happy new year!

Jörgen

georgeverghese · Jan 1, 2017

In most plants, a de isohexanised gasoline stream is used as the lean oil. Since its vapor pressure is low, it has a tendency to absorb lighter (and to a smaller extent heavier) fractions till the total vapor pressure approaches the actual pressure of the system. It is easy to see this when you look at the De Priester charts for vapor pressure for light to heavy HC components found in straight run gasoline. The recovery is improved when the lean oil is lower in temperature or refrigerated. H2S tends to go along with propane if you have H2S in the rich gas feed.

There is a good worked example of this in Perry Chem Engg Handbook in the chapter on distillation which uses the simplified Kremser Brown equation to rerate the performance of a lean oil absorption column which you may find useful.

If you could enable a rich lean oil bleed before the heavy lean oil stream is introduced, this would improve recovery. Strange how the principle of operation for this plant has been compromised for the sake of mechanical simplicity.

Happy New Year to you also.

Jyrka77 · Jan 2, 2017

So basically my system can be called a lean oil system, since gasoline is used in most lean oil systems?

Okay, I will try to find the Perry handbook.

Do you know anything about the influence of gas-to-lean oil ratio in the absorber. Can it be that if there is too much gas flow, absorption isnt functioning, but with lowering the gas flowrate, it starts to function? In the current system there is no option to control the gas flowrate to absorber. My idea was to build a bypass pipe with a valve, to bypass some gas and control the amount of gas sent to column, and then just experiment with different flowrates. Does it make sense?
Also the column dimensions could be incorrect, like too small diameter and therefore too high gas velocity... the diameter is 1 m, height 5,5 m, 9 trays (valve), gas flowrate ca 5000 Nm3/h (ca 7000 kg/h), gasoline (absorbent) DESIGN flowrate 14 t/h.

Thank you!

Jörgen

georgeverghese · Jan 2, 2017

The required gas to lean oil ratio would depend on a lot of factors, some of which are:
a)Desired recovery ratio for C3 and C4
b)Absorber temp and pressure
c)No. of trays in the column
d)Amount of light ends in the lean oil
e)To a smaller extent, the tray efficiency

Details on these may all be found in the chapters on distillation and gas absorption in Perry.
Typically, one would try to keep the tray hydraulic operating point at 70% ( max 80%) or so of flooding limit for non foaming / non fouling applications - from this, you'll know if the vapor velocity through the column is acceptable. Details on calculating the flood factor (with the Eckhart charts) are also in Perry and will depend on the type of trays you've got.

You'll find details on many of these not only in Perry, but also in other mass transfer texts like Treybal.

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Natural gasoline absorption with... natural gasoline

Jyrka77

Petroleum

Recommended for you

georgeverghese

Chemical

Jyrka77

Petroleum

georgeverghese

Chemical

Jyrka77

Petroleum

georgeverghese

Chemical

Similar threads

Part and Inventory Search

Sponsor