10 replies to this topic

[Nicholas]

Hi All

 

Got the following information for the inlet gas to a compressor system:

 

Flow: 4.0 MMSCFD

T: 25 degC

P: 110 psig

 

The wet gas analysis is as attached.

 

I would like your input to size a KO drum to treat this gas prior to entering the compressor. Can you please advise?

 

Thanks and Regards

Attached Files

•  Wet Gas Analysis.docx   213.75KB   45 downloads

Edited by Nicholas, 01 April 2015 - 03:00 PM.

[Zauberberg]

Nicholas, please use the Search feature on these forums. There has been a lot of threads on this subject. I typed "sizing a knockout drum" in the search window, and total 183 results came out. Some of them point at spreadsheets, discussions, etc.

 

Please review these threads and come back with any specific query you may have afterwards.

[Bobby Strain]

And you can check your sizing at my site. Just Google my name.

 

Bobby

[Teknas]

Hi Nicholas,

 

Do try out the spreadsheets available in this forum to get immediate results, but would advise you to go through the Standards / Codes being applied for your project. A good start would be to refer API 12J.

[Nicholas]

Thanks

[Nicholas]

^Gentlemen

^{Your spreadsheet links are inaccessible to me.}

 

^{Also, based on some research done, the sizing assumes a liquid volume coming in with the gas. The wet sample given or gas volume given does not have a liquid volume or composition. I did a high level check which showed that for these conditions even C6 is still in the gas phase.}

 

^{Can you please advise on this and reload spreadsheets? Is the input data inadequate? Thanks.}

 

^Nicholas

Attached Files

•  Spreadsheet Linksdocx.docx   45.3KB   62 downloads

[Virginia]

Nicholas, 

 

If there is no liquid present in the inlet gas, generally liquid flow rate is assumed as 5 % of inlet flow for sizing of Compressor KOD. The liquid properties can be taken from the upstream liquid stream. The KOD is sized considering there is a condensation of the gas and sometimes a slug of liquid (due to pocket formation upstream of KOD) can be accommodated in a Separator. 

 

Regards

[Teknas]

Nicholas,

 

Be cautious when you are taking properties for the liquid from an upstream equipment. If this gas is essentially dry and even the immediate upstream filter / separator / KOD does not show any liquid  flowrate in simulation but still gives liquid properties then do not use these figures.

 

In one of our previous projects a junior engineer considered liquid data obtained from the bottom of KOD, which although showed no flow had its property table populated in Hysys. The physical properties of this stream were not at all consistent with the liquid data. On inquiring with Hysys software team we got a reply that such streams are populated to converge the equation for the KOD and the properties should not be considered for calculation.

 

Hence, ensure that the correct stream properties are considered for KOD sizing. In our case we traced back the simulation till we found out a valid liquid stream and considered those properties assuming that in case of carryover this would be the liquid that would be found in the compressor KOD inlet stream.

 

Regards, 

[Nicholas]

Thanks for the pointers!

[wdew]

Nicholas,

 

Although the amount of liquid may be minimal or not even specified, a few questions you need to answer:

 

1) What process equipment is upstream of the new Knock Out Drum? No upstream equipment has a separation efficiency of 100%, depending on operating pressure, liquid load, gas and liquid properties and the vessel internals, the efficiency of the upstream vessel will  more likely be in the order of magnitude of 98 - 99.5%. The rest will end up in your vessel.

 

As a side note: in the event there is a cooler upstream of your vessel, have a look at how much liquid will be condensing there.

 

2) What is the maximum liquid content in the gas you require?

 

3) Is the liquid waxy or has it fouling tendencies?

 

4) What is the liquid level settings specification you wish to follow (minimum 100 mm or 150 mm per level band, 300 mm from HLL - NLL, LLLL 100 mm from bottom Tan-line)?

 

 

The above considerations will allow me to advise which demisting internals to select, which feeds into the max. K-value I can accept in the vessel, and this then will determine the vessel's diameter.

 

The height of the vessel will be determined by the value you arrive at for HHLL (keep some clearance to the bottom of the inlet device (don't use inlet baffles or half open pipes or the like), the distance you allow for gravity separation, the mist eliminator and the distance to the upper Tan-line.

 

That should be it.

 

wdew

[curious_cat]

In one of our previous projects a junior engineer considered liquid data obtained from the bottom of KOD, which although showed no flow had its property table populated in Hysys. The physical properties of this stream were not at all consistent with the liquid data. On inquiring with Hysys software team we got a reply that such streams are populated to converge the equation for the KOD and the properties should not be considered for calculation.

 

 

 

Good to know this. That's somewhat irresponsible of Hysys. Shouldn't they be flagging such dummy streams?

Edited by curious_cat, 13 April 2015 - 11:01 PM.