Idea: Thumper array w/ forced reflux and reflux return

[kerinin]

Strap in, it's another one of my crazy still ideas! It's the weekend and I've been entertaining myself thinking about different ways to achieve the same process as those gorgeous flutes. This started because I was pondering building a 2" flute, but increasing the condensate depth do the vapor speed could be kept the same as a 4" flute without loosing the performance.

Here's what I'm pondering:

thumber_array.jpg (7.57 KiB) Viewed 2489 times

It's basically just a series of thumpers - the difference is that it has forced reflux at the end of the chain (the reflux condenser) as well as reflux returns between each thumper. The idea is that the reflux flow will always be from the end of the chain towards the beginning (since we're forcing reflux at the end), so the reflux returns should act like the downcomers in the typical plate column design.

Rather than using bubble plates, this design uses thumpers, so the heat transfer between incoming vapor and condensate happens as the vapor rises through the condensate. Nothing special there.

I was thinking it would be cool to make the thumpers out of glass tubes so you could see the amount of condensate in each one, and watch as the vapor bubbles up through it.

So yes; I know I'm trying to 'fix' a design that isn't broken, and I absolutely *love* the flutes that everyone is building. I was just thinking about different approaches that might also be interesting. I don't know what the cost of glass tubes are, but it's conceivable that a rig like this could be built for less than the larger flutes.

As to the validity of the design or the quality of product it would produce - I have no idea! Suggestions? Criticism?

[LWTCS]

OMG that is the craziest idea eeeeeeever. It can't be done. It is just beyond reach. It is impossible.......It's crazy talk.
R U OK ??

Have a look


http://homedistiller.org/forum/viewtopi ... 16&t=14994

[kerinin]

There it is! I kept seeing references to it, but I couldn't find the thread. Thanks!

BTW, after some poking around, it turns out you can get glass tube from 'art glass' suppliers, and it's unbelievably cheap; $6 for 2" x 59" borosilicate tube:

http://pacificartglass.com/index.php?fi ... cat_id=603" onclick="window.open(this.href);return false;" rel="nofollow

Goes all the way up to almost 3" for $6:

http://pacificartglass.com/index.php?se ... &filterby=" onclick="window.open(this.href);return false;" rel="nofollow

Now just gotta figure out a connection detail - hopefully something requiring less machining than the glass Ponu I saw a while ago...

[kerinin]

Well, 2" copper fittings should be 50.4mm ID, and if the glass is 50mm OD, I'm wondering if you could wrap the glass a few times with teflon tape and just shove it in the socket. Wouldn't withstand much pressure without something to hold it in place though

[Samohon]

Take the tip from LWTCS before you go off and spend hard earned $$$...

Ah well, back to the drawing board...

[LWTCS]

I was just being a smarty pants cuz it has been done already. Your thought process is on point.
Miles and Barney are also putting together versions of this. Myles version is more similar to your drawing,,,only with the single thumper.

I hope you continue to think more about this. I think Rock is right however. With an array like that it will be much Harder to heat......Further more you do not get the benifit of the middle thumpers getting loaded with high proof vapor from up stream and high proof alchohol liquid draining back to the previous thumper. This dynamic action is what catapolts the abv to 95%.
Your drawing shows each independant thumper draining back to the primary boiler. That means lower abv and much more energy loss.

Hopefully Loneswinger can carve out enough time to tell us weather or not a 300ml liquid volume per chamber will out perform a 400 ml liquid volume.

[kerinin]

Now it's getting interesting

I should mention that I'm not thinking of these as thumpers in the traditional sense of being pots filled with liquid prior to firing up the still, I'm thinking of the 'thumpers' as being fed entirely from reflux, if that makes any difference

Without supplemental heat there will not be enough scavenged heat to support more than 2 thumper. The dephlegmator will have little to no effect as there will be no vapor making it to the dephlegmator

At one point I was considering heating each of the thumpers at the base and cooling each of them at the top, but still capturing vapor and condensing it into the next stage and returning overflow to the previous stage - might this type design address the issue? I moved away from that because it seemed wasteful to use coolant to force condensation at each stage, when the heat of condensation could be used for boiling the next stage. From another angle, why would a stacked column work but this wouldn't? Seems like the heat transfer is fairly similar in both cases - is heat loss due to lack of insulation the issue?

Take the tip from LWTCS before you go off and spend hard earned $$$...

I'm actually shocked at how cheap something like this *could* be. I just ordered a stick of the glass tube (seems like a cool thing to have around, even if this design is useless) for $12 including shipping. A stick of 3/4" and a stick of 1/2", plus a few reducers to seal the glass - I could see the entire material lest costing less than $80. But you're right, no point in wasting time and money on something that's fundamentally flawwed.

Your drawing shows each independant thumper draining back to the primary boiler. That means lower abv and much more energy loss.

That wasn't my intention, maybe I should have said more about what I'm hoping would happen here. Each 'stage' has a return going to the previous stage, and the vapor from each stage gets fed into the next stage. I'm envisioning running this at similar power levels as a 4" flute, so hopefully the vapor inlet wouldn't cause surging (maybe small perforations around the base or something would help here?). At the end of the chain there are two condensers, one draining back into the last stage (essentially a dephleg) and another draining to the takeoff. My intention was to use the same operating principals as a flute, but to use the depth of each stage to increase takeoff rates (rather than the width).

Hopefully Loneswinger can carve out enough time to tell us weather or not a 300ml liquid volume per chamber will out perform a 400 ml liquid volume.

I read through the discussion your linked to in the flute thread, and I'm working on some modifications to the design. I like the idea of cascading from larger volumes at the start of the chain to smaller volumes at the end - even if it didn't help with the %ABV, it seems like having a smaller 'thumper' at the end of the chain would be useful for heads separation. The revised design would have two 'thumpers' with 600mL volume, one with 400mL and the final stage would have 200mL. I've been considering making the 600mL thumpers modular so you could go from a 3-stage (600/400/200) setup to however many you wanted (600/600/600/600/400/200).

I just have to say it again - I can't believe how cheap 2" glass tube is; there has to be a good use for it!

[olddog]

Are you thinking of something like this.

OD

[LWTCS]

Yeah K.

Put a dephleg on top of the Evil Twins. Then put the Twins atop a Humper Thumper to mitigate the energy (and abv) loss created by having return back to the primary boiler.

Bet she would crank like a scalded dog!!?

Hahahahaha "The Scalded Dog" Put that on a Tshirt aye Mike

Someone build it please.

[olddog]

Put a dephleg on top of the Evil Twins.

There was a couple of turns of 1/2"around the top of the second column, but it needed more. I would have benefited with a proper dephleg.


OD

[kerinin]

Are you thinking of something like this.

Yes! That's the evil twins still isn't it? How fast can you run that thing?

I'm still curious about my central assumption though; that increasing the condensate depth will have the same effect as increasing the plate area.

Aside from that, the only real 'idea' i'm working with is a horizontal array of glass tubes. Mostly an aesthetic choice, but I do think it would look pretty cool. Sorta like a mad scientist's lair.

[olddog]

Yes! That's the evil twins still isn't it? How fast can you run that thing?

That used to run at potstill speed, about a quart every 15 minutes.


OD

[LWTCS]

Maybe better it didn't get done at that juncture Mike.......

Who knows how a more successful Twins build may have impacted the thought process for the Flute concept.


In hind sight....it would have been a game changer for the twins build.

Shame to dump all that high proof back into the primary though.

[olddog]

Shame to dump all that high proof back into the primary though.

That was just a tails return back to the boiler.


OD

[LWTCS]

That was just a tails return back to the boiler.

Oh yeah,,,,, but with a bonafied dephleg running?????

[kerinin]

After some more thought, I think the design can be simplified a bit. Rather than trying to set the condensate level using the top of the return pipe, and since they're horizontal, I'm thinking it would be easier to just connect each thumper, and control the level of each thumper with the height of the return pipe to the boiler. I also think it would be easier to use caps at the top of the thumper instead of reducers (also cheaper).

thumber_array2.jpg (6.54 KiB) Viewed 2195 times

One issue I see with this approach is that if the vapor bubbling through the thumpers is turbulent, you could end up with pressure differentials that would end up pushing reflux 'forward' in the chain, even if the aggregate reflux flow is backwards (since reflux is being generated at the end of the chain. This could lead to undesirable mixing between adjacent thumpers that would reduce the separation speed.

Out of curiousity, I tried modeling the behavior of one of the thumpers at the end of a stripping run I just did. I ran it pretty far into the tails so the vapor was mostly water, cut off the coolant to my Liebig, and stuck it in a jar of water. The first couple minutes were really interesting; the water was much colder than the vapor, so vapor kept spurting out of the bottom of the Liebig then immediately collapsing into liquid - made a strange popping sound each time; pop! pop! pop!. Eventually the water temperature rose to the point that the vapor was bubbling to the surface. I had the wife take a picture:

photo.JPG (7.13 KiB) Viewed 2195 times

This very unscientific experiment made two things clear; if I'm going to run the still at the speed I'm hoping to achieve, there's going to be a *lot* of turbulence in the thumpers, and I'm not sure that it will be possible to establish a clear 'water level' as shown in the schematic above. I'm not sure how the turbulence of the thumpers would affect the return system I'm showing, but I suspect it won't work as I'm intending. Maybe I could throw some ball/butterfly valves in the condensate return line so reflux can only flow in one direction, but I'm not sure how well that would work - most of those type valves allow some backflow before they close and that backflow might end up being more than the actual reflux flow I'm trying to facilitate.

[kerinin]

Here's a crazy idea; what if the heat transfer between thumpers was done with water / glycol rather than vapor motion? The basic idea would be to put a heat exchanger in each thumper (ie a pipe running from top to bottom) and cycle fluid from the start to the end of the chain. Assuming glycol, the glycol would start in the first thumper, absorb some heat from the condensate, then flow into the second thumper. Since the second thumper would have a lower boiling point, the hot glycol would transfer some heat to the condensate in the second thumper, ideally boiling off some of the condensate. It would then move to the next one and do the same thing. This approach would also require a condenser at the top of each thumper, and the condensate from each condenser would need to be added to the next thumper.

What you'd end up with would be a series of condensate reservoirs, each of which was boiling due to the glycol running through them. All the boiled vapor would be captured and sent to the next reservoir. There wouldn't be any reflux return; product could be collected as it arrived at the last condenser, at which point it would have gone through a fixed number of cycles. There would be no way for vapor to transit the entire array without going through those cycles, as you would be full condensing everything that boiled.

Probably more trouble than it's worth, but I think it's an interesting idea anyway...

[LWTCS]

Your definitely gonna get lots of smearing based on that diagram.

Keep thinking Kerinin. Your doin good.