Inlet Separator Sizing for a Gas Processing Plant with Feed Gas in Dense Phase

[Sirius P.Eng.]

I would like to verify the adequacy of an existing inlet Separator for a Gas plant receiving gas from an offshore pipeline. An offshore FPSO exports raw gas through this pipeline to the plant. The FPSO is fitted with a production separator and in addition it carries out dehydration using TEG before compressing the gas for export to the plant.

My problem is this: the raw gas is in dense phase (above cricondenbar) at the arrival conditions (temperature, pressure). How do I size an inlet Separator for this plant when essentially only one phase (vapour phase/dense phase) exists. The souders-brown equation for instance, requires a gas/vapour density as well as a liquid phase density.

Please advise and share any experiences you have with gas plant inlet separators.

 

[georgeverghese]

Cricodenbar estimates show significant dependence to the composition of the heavy components in the gas. Heavier components may also appear in the feed in the later years when the offshore reservoir pressure has dropped. Compressor suction and discharge scrubbers would not be 100% efficient and there will be liquids carryover in mist form into the overheads gas stream. Also liquids may appear at lower operating onshore arrival pressures . Unless the gas is conservatively treated specifically for hydrocarbon dewpoint depression at this FPSO, one should expect liquids. Suggest using a liquid density of 500-600kg/m3.

 

[Sirius P.Eng.]

Thank you georgeverghese (Chemical)for your response - Suggest using a liquid density of 500-600kg/m3. Please advise on hydrocarbon liquid and water flow rates to be considered in design. Like i mentioned earlier, the gas is in dense phase and no liquid and aqueous phases are predicted by HYSYS at arrival conditions. Thank you.

 

[apetri]

assuming there is a single dense phase (no separate condensing phase as for example with high fractions of water) I calculate a phase diagram and the volumes on equilibrium lines, the tool that I use (Prode) does that automatically so the procedure is simple, I attach 2 screenshots for a 12 components mixture,

you see that, for this mixture, the values for density (the inverse of specific volume) for liquid phase are in the range 300-700 Kg/M3 depending from conditions,
you should be able to find similar values with your simulator

 

[georgeverghese]

If you change some of the heavy components to higher bp, the cricondenbar will increase, and you will get some hydrocarbon liquids settling in the pipeline. So you can see any errors or approximations made in the lab heavy component characterisation changes perception significantly. Pipeline holdup will be highest at low flow. Subsequent flow ramp up operations will sweep out these liquids at much higher rates than at steady state. Some pipeline hydraulic simulations with acceptable ramp up rates( operationally limited max ramp rate to be agreed by Plant Operations staff) with these alterations in heavy components composition will tell you what liquid condensate phase flowrate will be at this inlet sep.

Plant Operations often continue running offshore operations for maybe up to 2-3days when the TEG unit has some operational trip. This would be the basis for the aqueous phase arrival rates at the inlet sep if you ask me. They often inject liquid CI into the pipeline to counter any short term corrosion during this time, but these injection rates are small in comparison to HC/ water condensation rates in the pipeline. Likewise, TEG carryover from the contactor will also be relatively low.

 

[gaswell]

there are several papers discussing these applications, comparing different design types, etc. see for example "High Pressure Gas-Liquid Separation: An Experimental Study on Separator Performance of Natural Gas Streams at Elevated Pressures" the authors came to the conclusion that inline cyclonic separators can replace larger, conventional separators.
It could be not easy, with conventional procedures (see GPSA Data book etc.) you can estimate fluid properties with a method as that suggested by apetri but take care that at high pressures separation may be difficult (i.e. poor performance).

 

[MortenA]

If your gas is in dense phase then it will not make sense to size it for normal operation! So you will have to size it for some other condition that could happen - or you dent need an inlet separator at all! I there a free water phase?

 

[Sirius P.Eng.]

there a free water phase?

There is actually no free water phase predicted for normal operations of the gas export systems ( compressors, pipeline, etc.). I suppose the current separator was sized for an upset scenario where free water phase was present.