Corken t291 pulling oil into pump

So we run with a huge jacketed honeypot. It has a rounded top, with various ports, and a centrally located injection port. On the recovery side we have our desiccant column, and then our pump.

We are consistently pulling hash into our pump. We’ve been checking the desiccant more regularly and find oil in it… like muffining oil, quite regularly.

When we service the corken there is a lot of oil build up in there. Sometimes to the point of clogging jt and overheating it.

Our thought processes/troubleshooting have gone like this.
1.) is someone flooding too much or too fast and the boiling solvent is too close to the port? Either flowing into the port directly
Or being sucked in?
Eliminated rhat.

2.) are minuscule amounts of hash floating up with the bubbles because our pump rips so hard/our water temps are so high?

3.) is oil/liquid solvent flowing down the sides of the injection port… rolling down the rounded top near the port and being sucked up?
—sort of eliminated. We elevated one side of the HP to encourage the solvent to move away from the recovery port.

Happy to hear any thoughts on the whole ordeal.

BUT. The question is this:

What could I do in the recovery line to possibly filter out dissolved oil before it makes its way to the pump head?

We have dessicsnt jn between. By moving to the smaller grade bb’s we’ve seen more oil be caught in the column.

We have since extended our dessicsnt tube to 4x48.
Ridiculous. I know. We use a blend of the small bb’s and medium. We use glass wool along with what I believe to be a 5 micron mesh/some plastic looking material gasket above the glass wool. But dissolved oil is still passing through. It seems like jn the bottom 18” or so of the dessicsnt some oil will be caught. When we open the bottom for he desiccant… the fuckkng oil literally muffins. Difficult to tell how many runs it’s been there for.

If more than one, its logical to assume some of it si redissolving jn the gas and moving through into the pump.

We run 24/7. Ids possible that on occasion an operator IS getting distracted and overfilling the honey pot. The water heater could go down. We charge right jnto our honeypot. They could forget to turn it off. Idfk.

Regardless. What could I do online on the recovery side of the system to catch this oil prior to hitting our pump. Pump rebuilds are expensive and time consuming

Thanks ya’ll. This community has been a godsend.
(Now If only rhe rest of you would stop utilizing all rhis info but continue posting the awesome ideas.).

1 Like

Is there a mol sieve in between the pot and pump?

1 Like

I use an inline jacketed column with a dip tube that runs to the bottom and leaving about an inch . Its a 2 inch diameter by 18 inch. I call it an expansion column which catches any extra hash before my CMEP. Never had hash on any rebuilds so they have said. I have that heated in line on my water circulation before anything else so its always hot and can make the gas evaporated from the hash. I end up with about a gram every couple months built up.


The only time I’ve had hash in a pump. It was operator error. I use a 4x48 sieve also.

1 Like

How are you measuring the volume or mass of solvent going into the honey pot?

Yea. I’ve always called it a dessicant tube. But yea. A d it’s 48” long now. Still having issues


We run an 8:1 ratio on average. 12.5p per column.

We also use a flow meter.

We have sight glasses on top of the honeypot.

Thanks. I’m not convinced that wasn’t the case.

But these guys swear it isn’t.

So idfk.

My man. That’s an idea that makes sense and would be easy to execute with our set up.

In this case the gas enters via the dip tube? Expands, abandons any remaining oil. And then rises to be recovered?


I believe @DABS-llc sells an expansion chamber

1 Like

Put a long jacketed tube on top of your recovery. Add height if still pulling oil


That is correct . Enter via dip tube and exit via the top . Hash stays in bottom . Had to find a welder to weld the dip tube on a 2in end cap with my size diameter inlet hole. Easy peasy.

1 Like

Is your recovery port next to your inlet? If so you definitely need to extend the inlet.

Calculate the volume of your vessel (the cylindrical portion) and only fill 80% of that at maximum. Use the density of liquid butane (or whatever solvent you use) to figure out how much solvent mass fits in that volume.
Don’t forget to also account for how much oil mass or volume is in the honey pot and subtract it from the maximum mass or volume from above

If your techs can’t do this math, then they probably overfilled the vessel.

1 Like

You must have a bump trap or demister before the inlet of the corken. You will consistently pull intrained organics out of your evaporator, occasionally you can pull large volumes of solvent loaded with oil.

You inlet to your evaporator should direct the solven to the heated walls. You don’t want vapor to have to travel through liquid, feeding the evaporator below a liquid level will cause excessive foaming and possible bumps.


Working on that too. Welding a funnel of sorts in. So that it creates separation between the recovery port and inlet opening


Yea. The math is tricky while nitro pushing one column clear and flooding the next. And actively
Recovering. But it’s doable.

The math will be easier when I install
The second solvent tank so that we are never recovering into nitro back pressure.

But yea. I think all this makes sense

Should be easier for you than most, considering you have flow meters. I time my liquid flow since I have no meter. And yeah, two solvent tanks is easiest.

Dealing with the extra n2 coming over all the time is a real bitch for solid recovery… I’ve been thinking of adding a trs-21 to the material column manifold, to help with scavenging the butane vapors and from the material columns as they are heated at 140°f. I would take the exhaust off the trs-21 and shove it into my recovery path above.

This would get around the extra n2 in my system fucking my collection recovery…

I would say this issue is very likely happening due to your operators overfilling/flooding the collection pot.

It sounds like the pot has been totally flooded once or twice and I would also bet the liquid level is regularly allowed to get too high in the pot during general operation. I know how it can get running monster crude systems, it’s easy to forget to check the pot if you’re not standing there watching the liquid level like a hawk. Workers often let the pot fill while doing other jobs and forget.

I run a Corken T291. You should not be getting any oil into the compressor at all. That must have been a real headache to disassemble and clean. Crazy that you actually clogged/overheated it!

To combat this issue I would suggest either not filling the collection pot as full or getting a bigger one. Or both. You also might need to alter the design of the pathway from the pot to the liquid/vapor separator to make sure you’re trapping any oil before it reaches the Corken.

You have to keep in mind the change in velocity of the vapor that happens at the recovery port on the collection pot. A large/wide vessel that necks down to 1-2" or lower will have very high vapor speed at the port and there will be a cone shaped gradient of increasing velocity surrounding the port. This will tend to pull small spits of liquid/oil into and through the port so you have to keep the boiling liquid level very low and have a liquid/vapor separator installed before the compressor. A custom collection pot with internal baffles is also a possibility I’ve considered.

If you’re getting a lot of heat into the solution, you likely have a lot of foaming, frothing and spitting happening at the liquid’d surface. I usually have 5-6" of bubbling froth on top of the liquid.

I keep a full 24-36" minimum between the very top of the boiling liquid’s surface and my recovery port. The more the better/safer. I also have a dedicated liquid/vapor separator (pic below) in my recovery line and it still gets lightly coated with oil at its inlet. My line between the separator and compressor stays totally clean, as does the compressor.

The 80% fill rule used for LPG tanks is more to assure enough room for thermal expansion and to allow enough vapor head. I personally would never fill a boiling collection pot that full. I prefer 50% max full for my collection pot. I sized my pot for this based on how much solvent I flood my standard column with so I never risk overfilling.

Also, and this is controversial, but I would suggest removing the mol sieve from the recovery line completely and replacing it with a free-flowing liquid/vapor separator or maybe two. I use a large dip-tube style cold trap design built from triclamp parts that’s about 4 feet tall. It’s the equivalent to having a 4’ tall pipe extending off the recovery port but without the inconvenience of actually having that tall of an extension on the pot itself.

Your mol sieve made from a 48" tube packed with small BB sized pellets like you describe would likely offer protection from what I would consider normal spits of liquid/oil that happen in my system but I would bet it’s likely acting as more of a restriction to vapor flow than anything else.

Also as a side question, why are you using nitrogen as your assist gas? Isn’t it annoying having to deal with removing it from the system and having to keep buying it/transporting the cylinders all the time?

IMO active systems should be using warm solvent vapor for pressure assisting.

Even if you’re running a crude monster and need lots of vapor you could setup a butane/propane vaporizer to supply as much vapor on demand as needed. Then you recover all the vapor like normal solvent, remove it from the main solvent tank and reuse it for vapor assist. No need to keep buying nitrogen and no need for the outlandish venting/vaccing required to use it.

If you run your active system such that you keep the solvent tank warm enough that it always has enough pressure to push liquid through the system on it’s own you wouldn’t even need to use any assist gas in that location, just on the columns. Then if you’re using warm solvent vapor you never need to vent. Just keep loading up columns and go.