Question about a Liebig for a 3" Boka

[joeymac]

In a typical slant plate LM the bottom plate just overflows and reflux falls back into the column. Its pretty simple. Is there an advantage or theory to what sergiolis is doing by having the product valve lower with the tee fitting return the reflux to the column using a tube?

[der wo]

Yes. If you read the whole thread you will find the answer. And also links to similar constructions.

[joeymac]

I saw bearrivers thread. His modification looks like it stemmed from originally having the top of his bottom slant plate at a level slightly below the top of his takeoff tube which was also horizontal.

If someone had a typical slant plate with a slanted takeoff completely below the top edge of the bottom plate - I'm wondering how much advantage there might be to a reflux return tube. I like the looks of it.

[der wo]

Sorry joey, I don't see the difference between bearrivers and sergiolis construction. For me it's the same. Both are normal bokas with an external return added. I don't know what's "Dee", perhaps this is the problem.

[sergiolis]

Yes der wo, both tube and valve are 1/2". I bought the SS tube from UK so I trust it must be a calibrated one. Anyway I wrapped teflon around cones and tube and looks tight enough....
Hi joeymac, I'm not the expert here but one of the advantages about reflux return is you get a smaller pool, therefore you minimise smearing. Another advantage is that you center completely the refluxed liquid, therefore you minimise channeling. Another advantage bearrivers mention is that some drops splash when they collide over the top liquid pool formed by the slanted plate, therefore your reflux ratio is not so accurate...

[joeymac]

Sorry joey, I don't see the difference between bearrivers and sergiolis construction. For me it's the same. Both are normal bokas with an external return added. I don't know what's "Dee", perhaps this is the problem.

Sorry, I hadn't had my morning coffee... I had a typo and fixed it above. I was meaning that I wonder what the major performance difference would be with sergio's or bear's design (both similar) compared to a simple Boka with reflux that overflows (like mine pictured here) where the takeoff tube is properly slanted and well below the bottom plate's top edge. Because it looked like bear was having channeling issues with his which is why he converted to eternal reflux return.

I actually was so curious that I decided to test mine with water in the sink like bearriver did in his thread. To simulate a worst case scenario, I put my 2" boka head in the sink and sprayed into the top of the head with about 10mL/sec of water which is more I can ever produce on a 5500 watt boiler. 5500W should only produce a maximum of 8.88 mL/sec condensate at 78degC. So this would simulate full reflux at full power. I only got about 60-70% of the return stream coming off the center point of the bottom plate and 30-40% channeling off the sides. Not great but not too bad either. If I backed the water flow off by half (to simulate a 50:50 reflux ratio) probably 85-95% of the water falls of the center of the plate. When I open the output tube to simulate a potstill mode (no reflux), I'd say nearly 95%+ of the water was coming out of the 3/8" product takeoff tube and only a very little was splashing or overflowing out of the bottom plate.

I'm sure I might get some slight benefit from external reflux return, but after testing I'm fine with the overall results and don't think it's worth the trouble for me to convert to an external reflux return tube. Anyways, sorry for the small sidetrack... back to sergio's thread.

[der wo]

Another benefit of an external return is the better temperature measurement. Before I build my current LM with external return, I had a VM/LM-combo without return, and the reflux affected the measurement. So 80°C at the beginning of the run with only little reflux had a higher abv than 80°C at the end with a high reflux ratio. With return is more distance between the cold reflux and the thermo probe.

But yes, a normal boka is good too of course. But I wanted to fix this problem . I don't claim, spirits taste better with an external return...

[rad14701]

To simulate a worst case scenario, I put my 2" boka head in the sink and sprayed into the top of the head with about 10mL/sec of water which is more I can ever produce on a 5500 watt boiler. 5500W should only produce a maximum of 8.88 mL/sec condensate at 78degC

This is a misconception... Your boiler is going to be a lot hotter than 78C when the top of your column is reading 78C... It might be within a few degrees of 100C towards the end of a run and will always be above 78C if the column reads 78C... So you'll be knocking down a lot more vapor and handling more distillate than what you are speculating... For example, a 10% boiler charge won't even start producing vapor until it reaches 93C... And once the boiler charge is down to 5% the temperature will be 96C+... Just because the top of the column gives a specific reading it has no bearing on boiler or vapor temperature anywhere other than the cooled top end due to cooled reflux produced by the reflux condenser... Theories and fundamentals... Yeah, I say that a lot - because they mean a lot...

[joeymac]

The only reason I mentioned temperature is that the liquid density of ethanol (and water) varies slightly with temperature and that's about the temp of the liquids on the plates. But temperature doesn't really matter much. When only looking at condensate volume off reflux coils, all that really matters is power balance. Assuming a well insulated boiler and column in equilibrium with no parasitic heat loss and no product removal, the power into the boiler should equal the power removal at the condenser. Once in equilibrium, energy in must equal energy out. Considering a worst case scenario of a heating element outputing 5500W, then the condenser coils must condense 5500W worth of vapor. Meaning the liquid falling down the reflux head can't have more volume/mass than can be produced by 5500w of cooling. What happens in the boiler or column doesn't matter if there's no energy loss between the heat input and reflux coils... and parasitic heat loss or product removal as seen in the real world only works to decrease the reflux coil condensate volume anyways.

Heats of Vaporization: EtOH +841 kJ/kg , Water +2257 kJ/Kg
Liquid densities at 78C: EtOH 0.740 g/mL , Water 0.963 g/mL

These properties combine to make the following volume-per-energy:
EtOH 1.607 mL/kW per second
Water 0.460 mL/kW per second

The maximum pure ethanol condensate that can be produced with 5500W of cooling: 8.84 mL/sec
The maximum pure water condensate that can be produced with 5500W of cooling: 2.53 mL/sec
So the real-world condensate off the coils for any ABV% wash should be somewhere between about 8.8 and 2.5 mL/sec with 8.8 being "worst case".

[sergiolis]

der wo, glad to know there are still more advantages!!!
I will comment on my first run soon... Let's see how my slim liebig works!!!
Nothing new but this is my reflux condenser.It's about 30cm long. I tried to build a double helix but it wasn't possible. Anyway I hope it will be fine..
Thanks very much to all for your great help on building this vodka squeezer!!!!

[sergiolis]

Hi guys,
I'm now at the alcohol cleaning run stage, but something is wrong.
Some vapour leaks from the top of the column, however, the water out from the reflux condenser is cold. So the reflux condenser is not working properly. I think it is too narrow for the ID diameter of the column and some vapours are able to scape because they are not in contact with the coil.
Even working with a good water flow the take-off product is pretty warm (35ºC). Maybe because my liebig water layer it is also really thin.
So i need to change some elements but I don't know where to start.
I cannot build a double helix with corrugated SS because it doesn't fit well, so I guess I have to wind a double helix copper condenser, maybe OD: 10mm or 8mm
I could change my liebig too if I modify the outside tube for a wider tube made in copper. Now there is only a 1.5mm water layer...
what do you think?
Still there is much work to do

[joeymac]

Jam copper scrubbers into the voids of the coils. Not enough to hold liquid... but just enough to swirl and mix vapors. This will slow up the vapor, cause more turbulence, and increase efficiency. This is an easy fix and much simpler than making a new condenser. There's nothing wrong with that condenser - looks like it should knock down at least 4kW with proper vapor flow.

Also be sure you're running the cold water INTO the outer coils and the warm coolant should EXIT the center cold finger. The concept of running coolant opposite to the vapor is that the coldest coolant should be the LAST obstacle vapor must overcome. This is generally the most efficient flow direction for leibigs, shotguns, and dimroth (cold finger coil) condensers. Double helixes are fairly indifferent to coolant flow direction.

35C product takeoff is not a problem. Slant plate columns already produce a liquid product - you're just aiming to cool it a bit more so that your product STAYS liquid is not evaporating quickly out of the jars. When I'm stripping I remove my entire LM valve and leibig from the column so that product just streams out as liquid into large jugs as quickly as possible. My collection jugs get HOT enough that I keep the jugs in a water bath to keep them from getting too hot. But they're still probably 50+ deg C and it's not a problem. If I'm interested in taking a alcoholometer reading, I set 75mL or so aside to cool off to a reasonable temperature.

[der wo]

Sorry, that it doesn't work everything well...

Which direction is the water flow? Correct would be: Water down into the product liebig. Then into the coil. On the picture into the left, that the coldest water is on the top of the condenser. To plug scrubbers into the upper half of the coil would increase the efficiency.
But generally I can't help you here, because I always used liebigs for reflux.

I don't think the layer is the problem. Perhaps the thick stainless steel of the product tube is the problem. Such a 1.7mm thick walled stainless tube is beyond all our experiences I fear. SS is a very bad heat conductor. So ss better has to be thin walled to become a good heat exchanger.
I fear you have to make it longer or replace the inner tube.

But first get the other problem fixed. Perhaps in practice the 35°C warm product won't disturb much. And for a neutral you will never open the valve 100%, so perhaps in practice you get the desired 20°C.


Edit: Just see, joeymac has the same answers for you...

[sergiolis]

Nice answers and really helpful because looks like I love to buy new stuff, but I really don´t.....
Perfect! I have some feints to reflux so I will do my next try with them...
Maybe I can try to make the CSS condenser wider and I will put some scrubbers....
About the liebig if I have some issues, as it is not expensive, I could buy a new SS tube (0.91mm wall) so I would increase the water layer and the heat transfer, but as you recommend I will try it again first with a real run.
I really appreciated these quick answers.... thank you very much

[der wo]

BTW, sergiolis, where do you buy the SS tubes? Do you have a link?

[googe]

I did a mini liebig on a 4" worked well http://homedistiller.org/forum/viewtopi ... 17&t=42591

[sergiolis]

Here it is
http://www.ebay.com/itm/1-2-OD-X-20SWG- ... J7NBnewYZw" onclick="window.open(this.href);return false;" rel="nofollow

[der wo]

Ah, from the UK.
Are you sure it's seamless? For cleaning seamless would be good.

[sergiolis]

The link above is welded, but what I have (1.63mm wall) is seamless.
Welded would be a problem?

[der wo]

All SS tubes and pipes are welded as far as I know.
But some (for construction) have the seam only removed on the outside. The ones for drinking water have removed it inside too.
If it has a seam inside, it is very sharp and perhaps you will not get all the dirt and chemicals from welding away.

[sergiolis]

I asked the seller about the tube and this is the answer: This tube is welded not seamless however there is no raised weld inside and the bore is smooth, similar to larger hygienic tubes.
I have some more issues to improve.
I tested the slanted plates, once during the alcohol cleaning run and then under the water tap and there is some chanelling because some water is falling down from behind the lower plate. My plan is to leghten the upper plate at least one cm. In this way the problem would be solved but I would like to know what is your opinion.
Another issue is in the Reflux Return. As liquid mixture is dripping from the lower point is not centered, in fact liquid falls to the side close to the sight glass wall.
My plan is to maintain the vapour lock as it is but adding 2 more 90º elbows to displace horizontally the dripping point. In this way, visually the reflux return would be off-center but actually the dripping point would be centered. What do you think?

[der wo]

Did you see the channeling also while the cleaning run? If the channeling problem occurs only under the water tap, perhaps you used more water than you would have distillate? Or you did not hold it perfectly upright?
Generally this is a problem only, if you want to run it as a pure potstill too, and don't accept any reflux for this. Sounds a bit anal. And of course it's only a problem, when you have this issue while the valve is 100% open.
But yes, lenghten is a solution, if there is really a reasonable amount of unwanted reflux.

Perhaps try soldering a tip (copper wire) or a wall (copper sheet) to the lowest point of reflux return.
Or how does it flow, if you simply remove the last 90° elbow?

[sergiolis]

So I filled in 3 SS scrubbers in the upper half of the reflux condenser, and its shape now is wider.
I centered the lowest point of the Reflux Return (In case is not completely centered a piece of ss scrubber around the lowest point might help)
I extended the lenght of the upper plate because during the cleaning alcohol run there was channeling too... now under the tap water there is no channeling..
let's see how it works in a feints run.

[der wo]

Looks good.
You could put packing also in the sight glass, that the return is directly over or perhaps slightly in the packing.

[sergiolis]

Thanks der wo,
Yes, this was my idea. I'm only scared about the reflux condenser because probably I'll be working at more than 3000W. I can't wait to test it!

[sergiolis]

Feints run succesful!!! I will comment on SPP thread....

[sergiolis]

I'm thinking about to build a Shotgun condenser (or dephelgmater) as a reflux condenser. I already have the spool tube 18" height, so I would need only some 12mm OD copper tube and a copper sheet.. It would be not expensive and I think it will improve much more the efficiency.
This data ( http://pracowniametaloplastyczna.pl/en/advices" onclick="window.open(this.href);return false;" rel="nofollow ) suggest 4600W for a 3" column and my dimroth condenser cannot hold such power....
Is it a good idea?

[der wo]

Sorry, I can't help this time. I don't have experience with dimensioning reflux coils or shotguns.
If you can get smaller diameter corrugated tubing, perhaps a double coil is another option?

[sergiolis]

Hi der wo, I hope you are well ! ...
I tried but the smaller diameter I could find for CSS is DN8 (12mm OD) ... that's why I'm thinking in alternatives like a shotgun.

[der wo]

Hi sergio,

I have in the German speaking forum a member with almost the same design (also the same reflux return) and with massive problems: With 100% open valve the reflux goes down the horizontal return path. He only gets 2 drips per second product regardless of the heat he uses. He even rised the wattage to almost flooding the column. Were you able to control with the valve something like 100% output and 0% reflux?
Here a picture:
http://www.schnapsbrennen.at/popup/imgW ... age317.jpg" onclick="window.open(this.href);return false;" rel="nofollow

Here a video with sightglass under the return:
https://www.youtube.com/watch?v=aNtcP7U7qs8" onclick="window.open(this.href);return false;" rel="nofollow
High wattage. Vinegar run. He says, the valve is 100% open.