Composition of slop tank at proposed compressor station

[wanabe gasman]

Hello, and thanks for your time and feedback.

I am recent graduate working on design of a compressor station that will compress and dehy well gas for entry to a pipeline.
I have simulated the process in HYSYS and after saturating with water, separation, compression, cooling, and the dehydration unit: all the liquid dumps recycle back to the inlet separator and the liquids from the bottom of this separator are sent to atmospheric slop tank.
The composition I am getting for this liquid drain to slop tank is 98 mol% water. A co-worker told me this is wrong; he said too much water, that it will likely be around 60-70% water. I am not sure where I have made a mistake in HYSYS or if HYSYS is just not capable of simulating this process correctly.

I reviewed the stream and it is 99% liquid at the stream conditions. Coming from the separator it is ~100psig to the 0psig tank, so it will flash.
The 1% that is vapor is 77 mol% c1 and c2.
Is it possible that my simulation is correct and sum of the liquids (condensed underground from well, condensed on compressor skid scrubbers, open drains, glycol contactor drains, after separator drains)could be 98 mol% water in the liquid phase to the slop tank?

The MW of gas from well is ~77. 74%C1 and 13%C2. We will compress the gas to 1175 psig.
For now I assumed the gas left the well at 120F, 60psig.
With a flow of 35MMSCFD I calculated I would be sending liquids to the slop tank at a rate of 75 barrels per day. Does this seem reasonable?
Currently I am trying to size a VRU and would need this info to proceed.

Please give feedback so that I may learn where my thought process is good or bad.

Thank you.

 

[georgeverghese]

The simulator result is most probably correct, and this is a reflection of what happens in theory. In actual practice, perfect light phase - heavy phase liquid fraction separation occurs only for clean fluids. In most other cases, with gunky corrosion inhibitors, suspended solids, and with emulsions caused by strong shear effects as liquids go through high pressure letdown valves, poor L-L phase separation effects can be a nightmare for plant operations. So your co worker may be telling you what actually happens in the field.

Liquid phase fraction separation is made all the more difficult by the inability of the interface LT to "find " the interface due to an emulsion band that would develop in the 3 phase slops drum in these cases. Using a bucket and weir type internals arrangement helps to reduce the mal effects of the interface LT, as would a double overweir design also.

Even if you use more complicated internals as suggested, to be cautious, you may still want to deliberately let some light phase condensate leak into the water phase exit line to reflect field experience at this slops drum. In Pro II, there is a feature to allow a desired light phase entrainment rate into the heavy phasein a 3 phase separator. This will then help to derive a gas rate for VRU design that is more inline with what your co workers would expect.