Column overload/flooding

[Evil Wizard]

I'm hoping to get some quick info, especially from any vapour engineers out there.

I have a design in mind where I actually want to induce the vapour in my column to crash - turn to liquid and drop down where it will siphon out before it hits the heating element.

I plan on injecting the packed column with hot vapour, 60%-70% most likely, in a constant rate, temp and pressure.

A coil dephlagmator on top of the column will be tuned by water temp and flow rate to achieve removal of meths and some heads past the coil (cooling management), refluxing the majority of ethanol back into the column.

My question is, will the increasing vapour density in the column eventually cause it to crash (this is called overloading?) or will I need to set up a sensor to kill the heating element periodically?

Thanks.

[Demy]

Maybe I didn't understand, but you want to build (in simple words) a classic reflux column? You don't need to invent anything, there is great info here on the forum.

[Evil Wizard]

Hi Demy, sorry, you didn't understand.

I've built hundreds of columns. Sold them in fact.

"I actually want to induce the vapour in my column to crash"

[kimbodious]

I read this as sending the vapour to the top of the column without sending it via the packing. The condensate would return to the boiler via the packing but without mingling with any vapour that would otherwise be rising up through the packing? Yeah nah, I can see how that won’t work.

[ShineonCrazyDiamond]

I think if you get your column (packing or plates) into equilibrium, and then add additional vapor from a second source, it would act no different than if you just turned up the heat and added more vapor the ol' traditional way. It will just flood. Then, to counter act this, you want to cut your power stability intermittently.

The real question is what do you hope to achieve from all this? On the outside it sounds extremely over complicated for no apparent benefit than to say you built something different.

[LWTCS]

I'm hoping to get some quick info, especially from any vapour engineers out there.

I have a design in mind where I actually want to induce the vapour in my column to crash - turn to liquid and drop down where it will siphon out before it hits the heating element.
this sounds a bit like Liquid management? Will this siphoned out liquid be finished product?

I plan on injecting the packed column with hot vapour, 60%-70% most likely, in a constant rate, temp and pressure.

A coil dephlagmator on top of the column will be tuned by water temp and flow rate to achieve removal of meths and some heads past the coil (cooling management), refluxing the majority of ethanol back into the column.

this sounds like the type of system used to clean up fuel grade ethanol where by column bottoms is the finished product and the undesirable constituents are bled off of the top of the apparatus? What is your intended function for your design? What kind of tool are you wanting to build?

My question is, will the increasing vapour density in the column eventually cause it to crash (this is called overloading?) or will I need to set up a sensor to kill the heating element periodically?

killing heat input will kill your gradient. No gradient=no fractionation. On a continuous feed system you'd manage feed rate to control plate behavior and keep btu/watt input steady. I would imagine you'll need to control the feed rate of your injection? However the heat source for the hot vapor injection should be independent of the constant heat input needed to maintain your gradient in the column imo. Any interruption in the heat needed to maintain the column will make the system very temperamental. Are there multiple columns on this system?

Thanks.

[LWTCS]

Here is an example of a system with 3 columns. Each column has it's own reboiler so heat input can be tuned to that column's specific requirements.

You can see that column #3 shows the finished product as bottoms. Vapor off the top is the undesirable.

Each column acts to basically scrub the shit out of dirty ethanol molecules.

[ShineonCrazyDiamond]

Good point Larry. Didn't think to consider the multiple column application. It came off as intending to create that system in a one column set up.

[LWTCS]

Good point Larry. Didn't think to consider the multiple column application. It came off as intending to create that system in a one column set up.

Yes it did. Based on the info within the question, I would say getting a single column to behave or otherwise be a superior design that does stuff a more conventional design cant do is a no go.
But the poster's question doesn't have enough information to understand what he is actually trying to accomplish.