Gas scrubber capacity

[pyMorty]

If I make the capacity calculation of an existing vertical gas scrubber (20" x 78")already installed in a compression skid I got as a result of gas handling capacity 2.0 MMSCFD (25 psig, 0.8 SG, 90°F).

The compressor (reciprocating, one stage) is going to actually handle now 4.0 MMSCFD, same pressure (25 psig) (cylinders were replaced with bigger ones).

What does it really mean the scrubber capacity? It is my understanding that since the calculations are based on velocity required for liquid drops disengaging from gas stream it only means that the liquid/gas separation won´t be as effective as it should be, but the gas flow shouldn´t be affected, i.e., the compressor would still handle 4.0 MMSCFD despite the undersized scrubber (of course with the risk of having liquid issues)... Is this correct?

 

[zdas04]

Thanks for a detailed post, it helps.

The big issue in solving your problem is knowing which parameter on the scrubber controlled the stated capacity. I've seen capacity limited by:
[ol 1]
[li]dP across the inlet nozzle;[/li]
[li]Bulk multiphase velocity in body of vessel;[/li]
[li]Superficial gas velocity in the main vessel;[/li]
[li]Superficial liquid velocity in the main vessel;[/li]
[li]Min velocity through mist pad (although "scrubbers" often don't have mist pads, the ones I put on compressor skids do have them, but most don't);[/li]
[li]Slug size[/li]
[li]GLR; or[/li]
[li]Capicty of dump valve (this is a good one that is too often ignored, it is really embarrassing to say your vessel can handle a 100 bbl slug, and find that a 50 bbl slug trips the skid on high liquid volume because the trim in the dump valve is too small).[/li]
[/ol]

If you can find what the limiting factor is, then you can determine for yourself if it really is limiting or not in your application (e.g., it is common to set the inlet nozzle dP at 0.05 psi, you can probably go to 0.5 psi without damaging anything). On the other hand if the limiting factor is gas superficial velocity in the main vessel then going above 2.0 MMSCF/d might cause serious problems. Once you decide that a limiting component is not limiting for you, then you need to go to the next limiting component and see if you have enough capacity for that one.

I've done this analysis a number of times and only found that I could ignore the nameplate capacity about 10% of the time. The other 90% I had to change out the scrubber.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[pyMorty]

Thank you zdas04. We checked the pressure at inlet scrubber nozzle and cylinder inlet and the dP is 5 psig (quite a bit). With the remaining pressure the performance run states the compressor should be still able to handle the flow.

According to your post I´m understanding that the scrubber could actually limit the flow rate? I mean, not only the performance of the separation process but the flow rate that enters the compression cylinders?

On another post I will try to explain the whole situation we are having with the compressor we revamped. This is a specific doubt it arose as a probable issue preveting the unit from reaching the desired flow...

 

[zdas04]

If you're not familiar with "superficial velocity" you are not alone, it is a parameter that we used a lot in multi-phase flow before the CFD guys took over and started assuming flow regimes (with significantly less relationship to physical flows, but that is another discussion). It is the velocity that a species would have if it was the only fluid in the conduit (in this case in the scrubber ID). Numbers less than 4 ft/s (remembering that you cannot calculate a gas velocity directly from SCF) are generally required for gas separation. Superficial velocity numbers for liquids are less widely used for separator sizing, it is more common to use slug size, but I have one client who always checks liquid superficial velocity to calculate multi-phase drag/slip to get a handle on dP within the vessel using the Hagedorn Brown correlations. I've never seen it to be controlling, but he claims that it sometimes is more important than slug size.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[Hurmanetar]

As zdas04 said, there are many criteria to look at.

Assuming this scrubber has a mesh pad, the ID of the vessel has to be sized for the condition. If this is 20" OD with 1/2" wall, you have a 19" ID which gives a superficial velocity of 4.77 fps on the 2 MMSCFD condition and 9.53 fps on the 4 MMSCFD condition.

API 11P establishes limits on superficial velocity in three classes. For this condition and scrubber the limits would be:
Class A: 6.61 fps
Class B: 4.72 fps
Class C: 3.40 fps

A package designer would want most of the conditions to fall under the class C velocity limit with maybe a few of the cases around the Class B limit. So this scrubber with a mesh pad can handle the 2 MMSCFD condition okay, but is too small for the 4 MMSCFD condition.

You have two options: A) replace the 20" scrubber with at least a 26" (25" ID) scrubber, or cut the head off your 20" scrubber and replace the mesh pad with a vane pack if there is room.

Or just do what most people do and exceed the recommended limits until something breaks... maybe the mesh pad will blow out and your mechanics will be fishing pieces of it out of the piping, cooler, and compressor valves.

 

[georgeverghese]

If this is a vertically oriented vane pack scrubber with a liquid overflow line from the vane pack down to the liquid space, you could actually reverse flow liquid out into the scrubber gas exit line at high vane pack dp. This has ruined at least one recip compressor in my experience.
On a related matter, were the suction pulsation snubbers checked out for this higher capacity ?

 

[pyMorty]

Thank you all. I guess we are keeping the scrubber, the gas is supposed to come already in appropiate conditions, the scrubber was installed just for redundance purposes. I´m opening a new thread related to this issue, we are having issues with the compressor meeting the flow, the scrubber was one of the items to check, but as far as I understand even when it is apparently undersized the impact will be the liquid separation capacity, but the flow shouldn´t be actually has to be affected, since at the end it should work just as a "pipe".

georgeverghese, the package was revamped with no pulsation bottles, it is a single stage, 25-40 psig suction to 120 psig discharge.