Continuous distillation - best source of heat

[dragonfrog]

I've been thinking about a continuous still design, where heated wash would be injected midway up a column, with steam coming up from below, like a couple we've seen on the site (Manu built one, Absinthe built another, maybe some others around).

What I'm not sure of is the best source of heat to preheat the wash - heat exchange with the spent wash, or using the wash to condense the distillate. The continuous stripper Manu built uses the spent wash as a heat source, as does Absinthe's. The other examples out there are the Armagnac and Coffey stills, which take the opposite approach, using the wash to be distilled as the coolant for the distillate vapours.

I did some calculations, but I know I oversimplified a lot - maybe too much. Perhaps someone more knowledgeable than me could look over this for any obvious errors...

heat capacity of water: 2.44 J/deg.g = 2.44 J/deg.mL
heat capacity of ethanol: 4.18 J/deg.g = 5.29 J/deg.mL (density of ethanol = 0.79 g /mL)

specific heat of vaporization of ethanol: 841 J/g = 1065 J / mL
specific heat of vaporization of water: 2257 J/g = 2257 J / mL

If the wash is 10%, the distillate 70%, and the spent wash .5%

heat capacity of a 10% wash: 2.73 J/deg.mL (ignoring reduced volume of ethanol-water mixtures...)

to cool 9 mL of water from 100 to 40 C yields up 60 * 9 * 2.44 = 1317 J
1317 J will heat 10 mL of 10% wash by 1317 / 10 / 2.73 = 48 deg, from 20 to 68 deg

to boil off .95 mL of ethanol and .51 mL of water takes 1065 * .95 + 2257 * .41 = 2162 J
1937 J will heat 10 mL of 10% wash by 1937 / 10 / 2.73 = 71 deg, from 20 to 91 deg

The vapours would take so much heat to condense, you'd probably need a second water-cooled condenser after the wash-cooled one!



If the wash is 5%, the distillate 55%, and the spent wash .5%

heat capacity of a 5% wash: 2.53 J/deg.mL

to cool 9.5 mL of water from 100 to 40 C yields up 60 * 9.5 * 2.44 = 1391 J
1391 J will heat 10 mL of 5% wash 1391 / 10 / 2.53 = 55 deg, from 20 to 75 deg

to boil off .45 mL ethanol and .37 mL water takes 1065 * .45 + 2257 * .37 = 1314 J
1314 J will heat 10mL of 5% wash 1314 / 10 / 2.53 = 52 deg, from 20 to 72 deg


So, in conclusion, there's a break-even point where it doesn't matter which heat source you use, and that's somewhere around a 5% wash - anything stronger than that, and there's more heat available in the vapours than in the hot spent wash, possibly even to the point where there's more than you can use. That's also ignoring the small amount of heat it would take to cool the wash down.

This makes some intuitive sense to me. On my last stripping run it took 2 hours to get a 6.5% beer up to boiling, and 2.5 hours to strip out most of the ethanol, using a 950 W burner on max the whole time - so more heat went into boiling off the distillate vapours than went into heating the wash to its initial boiling point. Of course, that's complicated by radiant heat from the boiler, and the gradually rising boiling point of the wash, and probably a host of other things I never even thought of. But you get the idea.

Anyway, comments or suggestions welcomed...

Cheers
DF

[manu de hanoi]

- About spent wash heat exchanger, the temp of both liquids is the same at best at the end of the heat exchange.
So roughly because the volumes are about the same you'd have a final temp midway between 20 and 100 = 60 C. More likely between 50 to 60. If you use steam heating and distill at high ABV the spent wash becomes a much more interesting option.

- Another source of energy is the reflux condenser (I dont think you'll get 70% straight out of 10 % wash)

- Yes you'd need a second condenser for either reflux (especially if you do high abvs) or output if you use them as pre heaters
here is a piece of art from 2 years ago :

the gorgon !

- I have still to figure how to conveniently reuse the extra heat from all the sources to heat the second heads rectifying column.

[dragonfrog]

Thanks for the reality check Manu!

It occurs to me that using the wash as coolant can't give you more heat than you can handle - otherwise the Coffey and Armagnac stills just wouldn't work! I don't know what strength of wash the Coffey still is meant to work with, but the Armagnac still is for distilling wine, so upward of 10% I'd think.

The only way I can think of its working would be that you start to get evaporation in the coolant of the condenser - no reason why evaporation has to happen exclusively in the still column, you'd just have to design the mechanism that takes the heated wash into the column to be able to handle some vapour already being present. Maybe what I need to do is actually find the patent for the patent still and read it...

(edit) - I think with a Liebig condenser you should be able to get more complete heat transfer than split-the-difference, since the fluids flow in opposite directions - if the condenser was long enough and the fluids moving slowly enough in both directions, you should theoretically be able to have the coolant leave the condenser at very nearly the same temperature as the distillate entering it, and vice versa.