Reflux Splitting

[Frosteecat]

Ok I was just operating on the anectdotal reports that end of run issues arise due to therm to height ratios. Your trap design(s) obviously retain more heat and are apparently more efficient.

[LWTCS]

Heat loss occurs with the lateral rigs

[myles]

I am hoping that by reintroducing some higher ABV liquid into the packing and onto the surface of the distributor plates that this will depress the temperature in the lower part of the column and encourage re-condensation of the high BP tails fractions that I don't want.

OOPS!! I forgot about dew point. I had hoped that "evapourative cooling" would play a role but it just wont work. The liquid reflux is at a lower temperature than the vapour at this point so there might be some cooling effect.

However, on reflection I think wetting the packing may have more of a contribution towards increasing the efficiency than anything else.

[flyingdutchman]

What if you built it with some form of condensor on the top and used a bokabob style take off just under it like in the diagram in his paper for the reflux control. The reflux return could be valved and have it run to some point near the base of the column dircetly to the packing/plates. It should be up high enough that the reflux would be vaporized some safe margin above the boiler but far enough down to keep the packing and or plates wet. Maybe incorporate a centering ring right at the bottom where it returns like the type you spoke about earlier. The valve could control how much reflux goes to the lower section of the column if you want, so that at the start it could reflux at the top but later relux to the bottom (not sure if this is an advantage or not but its an option). Above the bokabob take off (perhaps an inverted reducing bushing) you could go straight up and through to a dephleg/coil a sort of modified CM or have a second take off at the bokabob also valved for product LM (like his drawing) . I guess you could even configure it like a VM.

[myles]

From my perspective it has always been intended to be a VM / LM combination. The product will be on the VM side and the heads and reflux will be controlled in the LM side. I will split the reflux with the 2 return points at the top and mid-point of the column positions.

Top half of the column will be normal scrubbers with 1 reflux distributor, lower half of the column will be scubbers with reflux distributors set at column diameter spacing. In effect the bottom half will be a perforated plate column with packing between the plates.

I have thought about what I want the plates to do, as they are intended to redistribute the reflux and to act as a perforated plate, so have come up with this suggestion.

Lets call it a 52mm dia plate (for my 54mm tube) with 2 rows of perforations centered on Radius 10mm and Radius 20mm, the inner row are 5mm holes to return reflux to the packing (flow drilled to retain fluid on the plate untill it overflows into the holes) the outer row are 3mm holes normal drilled to act as perforations. With an additional 3mm hole in the centre of the plate. (just wish I had the skills to illustrate it) I know the recomendation is for 2mm holes for this diameter tube for a perforated plate but the combination of 6 x 5mm and 13 x 3mm works out close to the 10% plate area, and should be a compromise between a perforated plate and a reflux re-distributor.

The inner diameter of the tube is actually 51.6mm so the 52mm plate will be quite thin (0.4 or 0.7mm sheet) and have the outer edge raised up to be a compression fit into the column.

[flyingdutchman]

Myles,
Nice design! I would think the combination of reflux split and VM would be super efficient. Might I ask is the return at the middle of the column because from that point you can be sure the reflux will vaporize b-4 it can return to the boiler (while keeping the lower packing wet). Or is it so that you will be creating a sort of hybrid reflux stripper/rectifier combo like we see with continuous designs, the twist being its the reflux and not the wash that enters midpoint? Or both maybe? Or am I just way over thinking that one.

ABAAAfMgEAK-1.jpg (5.51 KiB) Viewed 2311 times

[myles]

No the mid point entry is just to counteract the tendancy of the column to dry out. I always run low wines or stripped wash on my column as I tend to accumulate feints from the pot still.

Any simmilarity to a centre feed stripper is coincidence, and I have no intention of charging with wash.

[flyingdutchman]

Hi Myles,
I promise I wasnt trying to steer this towards a continous type column. I was just curious why the middle instead of a little further down. Not that I mean to imply the middle is wrong as I really have no idea how far down it will travel once it enters (and of course once it hits the boiler its starting from scratch).

[myles]

At the moment it is just speculation. Really what I need to do is measure the ABV of the reflux at different points in the column during the phases of the run. Or find documentation from someone that has already done it - a simmulator would be nice. The whole idea is based on the understanding that as reflux gets depleted, so the column has less to work with. What I wish to do is counteract this but at the same time I don't want it to go in too low into the column or the reflux will still contain alcohol when it falls out the bottom.

At least by mounting the column over a thumper instead of the boiler this would not be a crisis in energy terms. I want to shift all the alcohol into the thumper and column so that the column is working in a "richer" environment. There is not much point in fractioning out alcohol and then returning it back to the wash - just a waste of energy and time.

I can see the simmilarities with a continuous column, although I am still in batch mode mindset - if most of the alcohol has been transfered to the thumper / column, the primary boiler is relegated to the function of a steam generator - just producing steam with a very low alcohol content as a heat source.

[flyingdutchman]

Myles,
Its a tad off topic but I just have to ask .....

It will be interesting to see if this is achievable, we take the old guideline of 20" per second for mesh/scrubbers for granted.

Do you happen to know if such a guideline exists for plates (either bubble or seive)??

Thxs
FD

[myles]

Myles,
Its a tad off topic but I just have to ask .....

It will be interesting to see if this is achievable, we take the old guideline of 20" per second for mesh/scrubbers for granted.

Do you happen to know if such a guideline exists for plates (either bubble or seive)??

Thxs
FD

I am sure there are but they are beyond my level of maths. I know for mesh/scrubbers that the figure of 20"/sec is based on the speed in an EMPTY column, so the actual speed when packed will be higher. This vapour speed is only really used to calculate suitable power levels. Taken to extreems and the vapour will just blow any liquid present right out of the column.

With plated columns one of the determining factors will be plate spacing. Wider spacings will allow faster speeds. This might be why some of the packed column configurations include an empty buffer zone between the top of the packing and the take off point. Dare i sugest a vortex generator might also be a possible solution that could potentially permit higher vapour speeds.

I am a bit wary of going too fast though as increased speed also reduces the "contact time" between vapour and liquid for segregation purposes.

[myles]

I had another think about the reflux distributor / perforated plates. With the increase in reflux returned to the lower section of the column, these will probably retain sufficient liquid anyway, so I have reversed the flow drilled holes to direct the surplus reflux into the packing.





This is cut slightly oversize and will be finished later, but this is just to give you the idea. The lower section of column will have these interspersed with the packing at column dia spacings to start with.

The perforations will be finished to final diameters once the plates have been rounded off propperly.

[flyingdutchman]

Hi Myles,
Thanks I have seen a couple of texts and indeed the math for just a 1 component system is tough (it can be done but thats for another thread I am planning).

This vapour speed is only really used to calculate suitable power levels. Taken to extreems and the vapour will just blow any liquid present right out of the column.

Couldn't agree more! There is a big difference between knowing the limits of a system and pushing that vs just building something and hoping it works.
I love the vortex idea I am guessing you could get a big gain in effective path length.
ps
just saw your plates they look great! Is the center hole for a threaded (SS) rod (modular, adjustable, sandwiches ??)

[mash rookie]

I have not had time to keep up with you guys recently but I think you are talking about three distinctive improvements for a basic packed column. Considering I am working on a four and a six inch packed column I probably should be paying closer attention.

(1) I am on board with the reflux return line. It is something I do with my flute. My experience says heads must be taken before any return or smearing occurs.

(2) Reflux centering and distribution plates. I like where you are going there. What about a cone shaped perforated plate? A funnel with perf holes?

(3) Vortex. Interesting thought. Yes FD, You could increase theoretical path length if you could affect vapor flow pattern causing a vortex without restricting volume flow. Angled fins? If over done it would be the same as making a taller narrower column.
Clockwise, Counter clockwise? Reversible for guys in Australia where toilet flush backwards? Just being funny.

Myles,. What if your centering plates were triangular and twisted like a fan blade? If they were concave wouldn’t they still collect and center distillate? Possibly create a vortex?

Mash Rookie

[Prairiepiss]

With what I've seen with my returning reflux from a packed column. The centering rings I'm starting to question. I know use of them have shown performance increases. But from what I've seen the reflux is really directed by the packing placement and not diverted to the walls. More so when it is at a higher ABV. So it leaves me wondering if the centering rings putting the reflux in the center is really doing something? It is there something more going on here?

Just an observation from running mine and getting the reflux to change paths just by adjusting the packing. I haven't had a chance to really play with this yet. I plan to. But I haven't really dug in and thought or played with it yet.

[mash rookie]

With what I've seen with my returning reflux from a packed column. The centering rings I'm starting to question. I know use of them have shown performance increases. But from what I've seen the reflux is really directed by the packing placement and not diverted to the walls. More so when it is at a higher ABV. So it leaves me wondering if the centering rings putting the reflux in the center is really doing something? It is there something more going on here?

Just an observation from running mine and getting the reflux to change paths just by adjusting the packing. I haven't had a chance to really play with this yet. I plan to. But I haven't really dug in and thought or played with it yet.

Do you have windows to observe the packing in the column while running? Have you been able to see it change paths? Are you manipulating packing or speculating?

I think the reason behind the thinking is to balance against the passive reflux occurring at the column walls. While not critical at the beginning or middle of the run it will come more into play as the run nears the end. Following Myles "wet the packing theory"

[Prairiepiss]

Have you looked at my build and videos? I don't have Windows in the column. I have them in the thumper ball. But I can see the bottom of the packing. And the returning reflux. You van see it in my videos. You can also see the wetting effect.

[mash rookie]

I was hoping you had seen vapor path change. I was able to observe quite well my packing through my site glasses when I did my tests but was unable to discern vapor path by observation of more or less dripping throughout the packing. It was quite even.

[myles]

Well MrP it could EASILY be "something more going on". It has been shown often enough before that you can observe an effect but come to the wrong conclusion about WHY it happened.

It has been shown with the centering rings for example that if you use a single big hole in the middle, you can improve the efficiency of the column. However, make the hole too small and the opposite occurs.

Now I am not able to say if the improvement occurs because of the reflux redistribution, because there is an orifice causing vapour turbulence, because we have generated a falling curtain of liquid or some other reason.
We can see the effect on surface tension as the alcohol is depleted and the reflux tends to run down the side walls instead of wicking into the packing. But it could be irrelevant. Now I am trying to redistribute my reflux evenly across the packing and not just directing towards 1 big hole, but I don't know if it will do anything. I am also trying to prevent the vapour from being 'focused' by the plates into a path running straight up the centre of the packing.

I am also hedging my bets by trying to make the plates behave somewhat like a perforated plate. I can visualise reasons for doing all of the above, but they could be completely wrong.

At the end of the day it is a bit of scrap copper sheet and time that is potentially being wasted. There is also that little voice (that I usually ignore) that says "there are lots of designs that already work, why bother? And when I do occasionally reply I usually say "why not?"

Even if it is no better than anything else, it probably won't be any worse either, and there is always that slight possibility that it might be better. And if it ever is better I know it will probably be due to something I had not even thought about. But I get as much enjoyment from "tinkering" as I do from running the still so I don't mind.

Now where is that flux cappacitor?

[Prairiepiss]

Well if it gets us closer to understanding why it works better. I'm all for it. It has just gotten me real curious as to why it works. Because as you stated centering the reflux is not the only thing happening there. By placing a centering ring in a column. You are changing many things. But I'm not real sure how to isolate them to see what actually increases performance.

A packed glass column would defiantly help to observe it. MR?

I will be testing different things with mine. But one man can only do so much with his spare time.

[mash rookie]

A packed glass column would defiantly help to observe it. MR?

Hmmm. Maybe. But it wont be done till next week. Maybe I will share. Not big on videos though. They are slated for plate trees. The planned test is to evaluate my packing plan for my stainless 4” & 6”
Who told you?

Myles, I have never seen you jump to conclusions as though they were fact. I see you as a carefuul observer and thinker. Even when I do not agree I find your posts thought provoking.

[myles]

Thanks MR, I follow your posts with interest too. That comment about the vortex generator I mentioned earlier - what I had in mind was the possibility of putting one between the top of the packing and the take off point - to spin any entrained droplets out towards the side of the column, if you were pushing the upper vapour speed limmit.

I have 19" or so of 2.25" clear actylic tube. Obviously no use for alcohol, but the worst case scenario in still performance would be with very low ABV vapour. Possibly pushing pure H2O steam through packing would be of visual benefit. I might try rigging up a small steam test rig. I had bought it in the 1st place to look at condenser coil performance, but just never got around to it.

Perhaps now would be a good time, well possibly next week as I have commitments this week. All this real life work keeps intruding on my leisure time.

[Frosteecat]

Moved to hardware...

[flyingdutchman]

I would think at the very least the plates would reduce channeling which occurs in all kinds of columns (not just stills) and plays a big part in reducing column efficiency. I would hope that properly constructed and fitted plates would prevent it from not just running down the column sides but break up/reduce channeling w/in the packing. I like the idea that it could act like a combo perf plate. I mean it seem like the punch through on the plate would determine liquid height. Sure I expect vapor to take the path of least resistance but wouldnt it split so that atleast some portion will flow through the small non punched holes and so not only wick onto the packing but also act like a perf plate (just like current flow in a parrallel circuit)? I think a glass column might show alot.
PS

But from what I've seen the reflux is really directed by the packing placement and not diverted to the walls.

I agree proper uniform packing goes a looong way towards preventing channeling/directing reflux flow.
PSPS
I would think if it was a modular design like a series of sandwiches, you could add or reduce the number of plates, increase descrease or elimnate the spacing between the sandwiches, have very good control over the unifromity of the packing, and take direct observations of what differnce occured with repect to separations??

[myles]

From what I have read and seen not much liquid can get through the perforations provided vapour is flowing, that's why the perforated plates need a downcommer. A layer of liquid is trapped on the plate and the vapour passes vertically through it. Inneficient (compared to the horizontal vapour path) but simpler to build than either bubble cap or valved plates.

With the flow drilled holes pointing downwards I suspect there would still be a layer of liquid on the plate. Pointed upwards they would try to maintain a depth of about 2
mm. I think I will try them pointed down in reflux distribution configuration first. I just wish I could put one in a school lab wind tunnel to see the vapour passage through the plate.

[flyingdutchman]

What i find really cool is that it seems (to me atleast) that alot of people are looking at the same basic concept you have described and seem to be envisioning it in different configurations. Its a great post

[Prairiepiss]

Myles have you thought about duel flow perforated plates? Basically perforated plates with bigger holes and no downcomers. Vapor speeds and such come more into play. But may be an option?

[myles]

Hi Myles, just saw your plates they look great! Is the center hole for a threaded (SS) rod (modular, adjustable, sandwiches ??)

That was not the origional plan, but I have just taken delivery of some 1 meter lengths of M3 threaded stainless steel rod, stainless crinkle washers and stainless nuts. So it looks like a centre threaded rod is on the cards. I already had 10% plate area perforated so loosing 1 perforation is not a big deal.

I will just have to make up a clamp to spin the plates in the bench drill so that I can round off the edges, and to reduce them enough to just fit inside the tube. One day I might get that lathe.

[flyingdutchman]

well...
Just flattened 17 pieces of 1 1/4 type M copper tube over the weekend. Anxiously awaiting your results or impressions before I move forwards. Im thinking of making a sandwich with the bottom plate facing up acting like a sieve plate, then say 2-4 cm of packing, followed by a top plate to act like a centering ring. then perhaps 4" - 6" of spacing between sandwiches. Hoping this might fix some of the off flavor issues with plated 2" columns and sort of act like a combo packed/plated still??? Just fooling with ideas so far.

[myles]

I got all the discs rounded off and reduced down to be a close fit inside the tube. In hindsight it would probably have been easier to do this before drilling the starter holes, but at the time I thought it would be more convenient to drill all the holes BEFORE cutting the discs out of the big offcut of copper sheet.

Memo for next time - remove backing sheet before drilling loads of perforations

Anyhow I have 2 circles of holes on each plate. The inner circle was drilled to 4mm and then flow drilled to 6mm, to create small funnels to direct the reflux back into the packing. The outer circle was drilled to 2mm (innitially) and these will now be opened up to 3mm. I don't expect much reflux will be able to penetrate these perforations due to the vapour passing through the small holes.

I have 21 discs available, so I could alternate with a disc between each stainless scrubber. I will not know for sure until I start to put it together. This will be the lower section of the column, topped off with straight scrubbers through to the top of the column. Pictures to follow.