Caustic wash carryover

[OilToil]

Hello all,

We have a remote gas sales facility that utilizes a caustic wash system, and apparently, intermittently the entire liquid content of the caustic wash vessels are carried out to the sales pipeline.

The system is configured as shown in the attached drawing, basically 2 vertical vessels operating at the same time in parallel, 20 feet tall, 80" diameter, 10 ft of liquid level in each. The gas inlet is on the bottom with a 6" diameter sparger with thirty six, 1" diameter, holes down the length of the sparger, holes alternating at 6 o'clock, 7 o'clock and 5 o'clock positions. There is a splash plate under the overhead vapor outlet, but no mist extractor.

The feed gas total is 3 MMSCFD, 70F, 30 PSIG, coming in 6" lines.

This system currently has no downstream scrubber vessel, just straight to a sales line. The treating liquid is not continuously regenerated, it is switched out and a new batch loaded when needed.

My questions are:

How to determine if the sizing of these vessels is appropriate for the current gas rates? We have none of the original design information. The closest thing I can think of is to check what vertical 2 phase flow regime we are in (but then the liquid is batch, not flowing), or a column tray flooding calculation, but I'm not sure if those correlations will work with 10 ft of liquid head. In any case, I don't think 1.5 MMSCFD through an 80" vessel is anywhere near enough velocity to entrain that much liquid?

Once I know what I should be checking for steady state design, the operations folks have hinted that the carry over problem may be more likely to occur after the field has been shut in and pressured up for a time. If the field were to pressure up to ~150 PSI and then the sparger inlet valve cranked open, well maybe that could carry some liquid out, and I will check that case too.

Is operating two of these in parallel stable? It seems to me that as soon as one vessel is fluidized, that vessel will take all the flow?

The holes on the sparger seem pretty large, where can I find design guidelines for spargers?

Suggestions to improve the design? A downstream scrubber would at least mitigate the issue and calm down the guys receiving 5000 gallons of chemical at their plant inlet unexpectedly...

Appreciate any insight you may have

 

[rustbuster]

Mist extracting pads would likely be the cheapest way to address carryover

 

[GHartmann]

That is a very crude / unusual scrubber. I doubt you have velocities high enough to develop a true froth of homogenous gas / liquid in the vessel (but you can check that). I would take a look at how the liquid inventory level is measured; something might be going on while the unit sits idle.

Not sure how your unit is pressuring up to 150 psig if you only have 30 psig gas.

With only 10 feet of free board there might be some circumstance the liquid is being drawn out with the gas. It would make more sense if the problems were at low pressures and high flows giving you high velocities.

Have an open minded brain storming session with the operations folks. Listen carefully to what they report even if its a little off the "engineering" track. When it happens, what happens, and what they think will be very usefull to your investigation.

 

[OilToil]

Thanks for your input, I have a few questions and clarifications:

" I doubt you have velocities high enough to develop a true froth of homogenous gas / liquid in the vessel (but you can check that)" - how do I check this? I have looked at drag effects for liquid droplets in this stream to see what vapor velocity can lift what size droplet, but I'm not really sure what I'm proving by doing that. Also, since I don't really know vapor flow area my estimates of velocity might not be very good.

The liquid inventory is measured with a sight glass.

The gas when the wells are flowing is 30 PSIG, when they are shut in for a while they pressure up to about 150 PSIG. The field operators use the valves just upstream of the scrubbers to shut everything in, so when they open the valves back up, they feed 150 PSIG gas into the ~30 PSIG scrubbers. I beleive this is the high flow low pressure (high velocity) situation that is most likely to carry the liquid 10 feet out of the vessels.

 

[GHartmann]

I think you can start by calculating the velocity of the gas through the holes in your sparger pipe. The pressure inlet the pipe will be roughly pressure in the tank plus the liquid head plus the dP in the pipe / orifice. The dP through the pipe holes can be approximated by:

dP = v2 x Rho / (2 x gc x C sq x 144) Where C = 0.6

I would be pretty sure that the sparger is just there to properly disperse the gas through the liquid and not create a homogenous mixture.

You can also check the superficial velocity of gas in the 80" diameter section (I believe it will be really low)

Seems to me you might need to have some type of controlled flow of the gas into the system at start-up.

Once the liqud gets into the overheads and into the downleg, perhaps there is a way the gas siphons all the liquid out of the vessel.