Three Phase Separator 1

[pkpbip]

I am sizing a 3 phase separator in a gas plant that receives cooled vapor and liquid hydrocarbon from the 1st stage of a compressor and also a sour water stream. The separator will then separate the hydrocarbon liquid, water, and vapor before sending the vapor to the second stage of the compressor directly.

My question is this: what is the reason for the rules of thumb behind using a L/D ratio between 3 and 4?

I am trying to size the drum to fit in an existing structure to avoid having to build a new one. The vapor load is high and is calling for a 7.5 ft diameter. This is required for the critical velocity and also some other factors (such as minimum distance between HLL and LLL). I am using a vertical and a horizontal demister pad to allow for a smaller drum.

The problem I have is that I have PLENTY of liquid holdup time and don't need a 22.5 ft long drum for this reason. I would still have plenty of liquid holdup with a 15 ft long drum.

Is the reason for L/D ratio rule of thumb to allow sufficent time for the liquid droplet to fall before the vapor leaves the drum? The drum will have two inlets feeding through slotted T distributor directed at the vessel heads. Does this help? Can I push to have a lower L/D ratio with some other controls or is there a different reason entirely for the rule of thumb?

Thanks in advance for your help!

 

[25362]

Visit thread124-10444.

 

[NghiaPP]

The main reason for the rule of thumb is minimizing material used to build the vessel. The wall thickness is proportional with diameter and design pressure. At low design pressure, the thickness is relative small, the optimal L/D ratio is in the range 3-4. At higher design pressure, high L/D ratio (5-6) will reduce material used for the same vessel volume.
Please note that vessel diameter may not less than diameter needed for separation. Bigger diameter is more conservative in term of vapor separation. The demister pad diameter will decide the minimal vessel diameter in your case (I assume you use vertical vessel). For horizontal vessel, other criteria are applied.
I wonder why you need a plenty of liquid holdup time ? It should not be a requirement for good separation.

 

[pkpbip]

Thanks for the info.

This will be a horizontal vessel and the critical velocity has a lot to do with the sizing of the vessel but there are some other factors I am having to consider such as minimum distance between HLL and LLL (simply minimum distance b/t taps for control). Not using a vertical drum because company prefers horizontal drums for water/hydrocarbon liquid separation.

I don't NEED plenty of holdup time. I have more than enough. I would think that typically the length of the drum is what determines your hold-up time after you consider all the factors that determine your diameter. The problem I was having was that the length in this case was being determined by the minimum L/D ratio that I had set by default at 3. I was questioning this because I wanted to lower that to 2. Company does have a requirement for liquid hold-up because this is a compressor feed drum and needs to have sufficient hold-up should an upset occur so that it doesn't trip the compressor on high level.

I ended up going with the L/D ratio of 2 because space is limited in existing structure and any excess cost associated with drum is well worth not having to build another structure.

 

[chemELK]

I'm really surprised your company is so opposed to using a horizontal vessel for 3-phase separation. I use horizontal vessels for 3-phase gas, crude, oil seperation all the time for several different clients. If you are metering the products out of the vessel before additional separation, this is the only way to get accurate metering. Notice most vapor/liquid separation standard calculations are concerned with producing "liquids free" gas. However, in many applications it is important to also size the vessel for "gas-free" liquids and liquid-liquid separation. For most applications this can be done with a Stokes' Law calculation using a bubble/globule size of 500 microns. If you have additional liquids separation and this vessel is really just an inlet scrubber for a compressor, not a true 3-phase separator, then you can ignore this.

 

[25362]

The old ROT to separate light hydrocarbons and water is to provide a cross-section for a cross-flow velocity of 1 mm/s for water settling through an hydrocarbon with a range of viscosities from liquid butane to diesel. For unusual services, one should carry out lab tests on settling rates.

 

[MortenA]

If your vapour flow rate requires a certain cross sectional area optimise for this with regards to diameter and find the corresponding length required for your hold-up time. I believe that liquid hold-up time for horz. separators is usually the determening factor - not vapour velocity. The rule of thumb is therefore related to this situation. I dont think it would be economically attractive to make the vessel much larger than require just to meet the 3-4 ratio.

Maybe you should consider a (3 pahse) vertical separator in your case?

Best regards

Morten