Condensation of Liquids in Natural Gas 2

[dtai]

Hi,
When a mixture of gas with composition up to C7,C8,C9, goes through a knock out vessel, how much of the heavy hydrocarbon will be knocked out?
Will there be any heavy hydrocarbon remaining?
What will be the composition of the condensate?
Will the heaviest hydrocarbon condense first regardless of mol percent concentration?

dtai

 

[TD2K]

Maybe none, maybe a lot.

You need to pull out your textbooks and review phase equilibrium.

 

[dtai]

I am not chemical process trained. Could you please enlighten me?

 

[sshep]

Unfortunately the basics of flash calculations of even a basically ideal hydrocarbon mixture as in your case are difficult to effectively communicate to an untrained person. The easiest way to do these calculations is to obtain access to a process simulator to do your flash calcs, in which case little theory is needed if someone can help you get started.

There are numberous sources available on the web for the theory (search flash calculation), but usually the notation of nested summations makes this reading tedious. If you can get past that, you can build a spreadsheat yourself to do flash calculations using antoine's equation and itterating until the sum of vapor pressures equals the total pressure- in which case you are on significantly your way towards becoming a chemical engineer.

In the meantime I found this little tool which might help, but which seems to be missing nonane:

http://www.higgins.ucdavis.edu/webMathematica/MSP/Examples/FlashAnalysis

best of luck,
sshep

 

[Homayun]

dtai,

Doing flash calculation for binary systems (systems with only 2 componentes) is quite easy. All you need is the relative volatility of the two components.
However, as you have more components in the system, it becomes complexer to perform a hand calculation. For that purpose, we usually use a simulator.

 

[25362]

If the heavier hydrocarbons are dissolved in the gas as gas components, no knocking out should be expected just from passing the gas through a separating vessel.

However, if I got it right, the liquid laden gas stream is already the result of a previous "equilibrium" condition, such as from cooling the gas, and the pre-existing knock out vessel operates at the same T,P of the main stream.

I understand dtai refers to a liquid-gas vertical separator (aka KOD) to remove heavier hydrocarbon liquid droplets from a gas stream.

The droplet size range of the suspended heavy hydrocarbons should be defined. For example, if it is a mist, the average droplet diameter is 50-100 micrometers.

For such a drop diameter the corresponding gravity terminal settling velocity would be around 1 fps. It could be better estimated by the use of Brown & Sounders constant K, as follows:

V = K[(d_L-d_G)/d_G]^0.5

where

K: the B&S constant that can be found from published graphs
d_G, d_L: the densities of the gas and the liquid, respectively

Re-entrainment can be reduced by imparting the entering gas a centrifugal motion. A centrifugal force is generated that may be more than hundred times gravity in small units operating at high pressures. Besides, entrainment is sharply reduced as impingement type internals are added to the KOD.

There is one article in the Hydrocarbon Processing issue of June 1990, titled Selecting gas/liquid separators by P.G. Talavera, that may be of value.

But then, again, I may have misinterpreted dtai's post;
dtai is invited to tell us whether I'm right or wrong.

 

[katmar]

One common misconception I have found amongst non-chemical engineers is that all of the least volatile component condenses first without any of the other components, followed by the next least volatile component etc.

They assume that all of each component condenses at its boiling point. Unfortunately as components are mixed they condense in varying mixed compositions. As others have said, it is difficult to describe to someone who has not studied this in detail, but be aware that the separate components do not condense individually.

 

[25362]

Not to mention the retrograde condensation problems encountered in gas fields !

 

[Meya04]

You can perform the binary flash calculation for this matter by using elements rather than components. Choose the one which is important and leave the rest in the elements then perform a binary flash calculation.

 

[ProcsssDr]

This respoonse is kind of late, but I will answer dtai's questions as they stand in a general sense:
1. The amount of heavy hydrocarbon separated in the knockout vessel depends on the amount of liquid in the incoming stream and the design (size and type in internals) of the vessel. If there is no liquid in the incoming stream, then there is nothing for the knockout vessel to separate. If there is liquid, then the vessel will separate some or most of the liquid depending on the design of the vessel.
2. Yes, there will be heavy hydrocarbon remaining. I assume you mean remaining in the gas phase. Phase equilibrium describes the distribution of components between phases. The distribution factors (k-factors) are finite, hence some of each component is in each phase.
3. The composition of the condensate is dependent on the overall composition, temperature, and pressure of the incoming stream.
4. No. Some of all the components will be present in the condensate. However, the heavier components will have a higher concentration in the liquid and the lighter components will have a higher concentration in the vapor.

 

[dtai]

Gentlemen,
Thanks for the valuable input especially to a non-chemical process guy.
I have a better idea now of what's going on when natural gas passes through a KO drum.

Blessed Christmas.

Dtai

 

[arie2044]

Hi People
i'm looking for chemical & physical data for LNG and the equipment needed to analyse it properties.how do you measure the dew point of gas in presence of water

thank you for an answer
Arie