Predicting the ABV of your distillate

[prankster1590]

When you have a binary system of ethanol and water in your still, then it's possible to predict the final alcohol content of your distillate. I'm new to this so I was wondering if i'm doing it correctly. I like calculating stuff and wanna improve my control over the distallation process.

This is the rayleigh equation:



I'm gonna solve this integral numerically by using simpsons rules. But first i need a x(molar fraction ethanol in liquid phase)-y(molar fraction in vapor phase).

For example: 1% ethanol in liquid phase (x) corresponds to about 15% ethanol vapor (y) when the 1% ethanol liquid is boiling.

basically gonna make a table with x values from 1% ethanol in water until about 88% ethanol in water. I make a third column in which 1/(y-x) is calculated.

L1=inital molar content of the still (number of moles ethanol + number of moles water)
L2=total number of moles left after distillation

So of i have to start with a solution which contains 21 mole% of ethanol I integrate the 1/(y-x) column from 21% to 1% ethanol left in the liquid (or 2%, or 3% left).

The number that you get is ln(L1/L2)

From my data going from a 1 mole (L1) 20% ethanol solution to a 1% solution results in 0.51 (L2) moles of a 1% solution left in the still.
This also means that the distillate contains 0,49 moles.

So the amount of ethanol distilled to the distillate is L1*0.2(20%)-(L2*0.01) = 0,2049 moles. The distillate contains a total of 0,49 moles so the distillate. Is 0,2049/0,49*100=42 mole% ethanol in water.

Dont know yet how to convert to ABV but my first try calculating that resulted in 81 V%, which seems a bit high.

So anybody know more about this. And does this calculation look good? Or. Are these numbers not realistic?

[Steve Broady]

I don’t mean to stop you from having fun, but I can’t help wondering why you’re going this route? There is a very helpful chart on the site which correlates I put ABV, vapor temperature, and output ABV. I believe it was generated using the equations you’ve mentioned above.

However, I don’t understand why you need these numbers on the first place. The advice I was given was to run the still and you get what you get. If you really need to know the exact proof at any given stage, it can be measured with a refractometer or a hydrometer, or you can do some simple calculations based on what you started with. If you start with 100 liters of 10%, then you have 10 liters of ethanol. Assuming you get all of it out of the still, you can figure out the resulting proof easily just by measuring the final volume. If you have 20 liters, then it’s 50%, for example.

[Hügelwilli]

Where do you have the numbers of the table?
At the low alcohol strengths the distance between y and x is too high according to the data I have and according to measurements distillers I know have done. Instead of 0.1564 it is more like 0.113.

Such calculations can be done without integrals by loops. For example the simple distillation calculators on this site work with a rough loop. Also the calculators on my website work with loops. Mathematically it is not correct, but computers can do it with enough precision, that it is not recognizable. You can loop a distillation in 100000 little steps without waiting more than 0.1 seconds for the result. Same as iterations: there is no need to reverse mathematical formulas sometimes.
The problem is, that you don't have a good data of x and y and probably not from the alcoholic strength in mol or weight to the density and %abv. This ruins every attempt of precision. This is the key question: What precision do you need?
And then there are factors like passive rectification and air pressure. Do you want to skip them? And what measurement tools you use? Do they have the precision you want? I mean, for example it makes no sense to measure the steam temperature with a bimetallic thermometer and then calculate the alcohol strength from it with one digit after the comma. Or to measure the alcohol strength with a 15cm long alcoholmeter and then dilute it using a calculator that is able to calculate with contraction.

[Mr Sippy]

At the low alcohol strengths the distance between y and x is too high according to the data I have and according to measurements distillers I know have done. Instead of 0.1564 it is more like 0.113.

I think it is more like 0.103 But the other numbers look pretty good.

You are on the right track. Using an integrated form of the Rayliegh equation is a perfectly acceptable method of doing a mass balance.
It's a bit of fun to generate an instantaneous picture of volumes and ABV in both boiler and collection vessel all related to temperature. It won't give you "better control" but it is fun.

Willi makes good but addressable points. You will have gross systematic errors if you don't account for pressure or non-ideal mixture. Another significant error source is the calcs being based on a binary mix. Since you seem comfortable with some maths I'd suggest to calculate your own VLE (Px-y) table. Instead of molar fractions use temperature increments of 0.2 degrees. For abv conversions I suggest reading Edwin Croissant's paper "On the conversion of alcohol". Non-ideal mixtures are corrected with an 'activity coefficient' applied to Raoults equation.

If you are coding this, I have some C code that may be useful to you. You will also likely find good info by searching any combination of mass balance + Rayliegh + single stage batch distillation.

[prankster1590]

Where do you have the numbers of the table?
At the low alcohol strengths the distance between y and x is too high according to the data I have and according to measurements distillers I know have done. Instead of 0.1564 it is more like 0.113.

I found a paper with experimental values but with x in steps of 10%. So I made a six degree polynomial with a r^2 of 0,998 to have values with steps of 1%. Buts thats ok. I know its not accurate. The mix is not a binary system. I was looking for a good approximation.

I like the pragmatic way from Steve broady.

If you really need to know the exact proof at any given stage, it can be measured with a refractometer or a hydrometer, or you can do some simple calculations based on what you started with. If you start with 100 liters of 10%, then you have 10 liters of ethanol. Assuming you get all of it out of the still, you can figure out the resulting proof easily just by measuring the final volume. If you have 20 liters, then it’s 50%, for example.

[Hügelwilli]

What I don't understand is that you think it is not a binary system.
The mix is of course a binary system. At least at the precision we can get. The congeners are negligible regarding the VLE.
And "Calculating a VLE"? How? The best way is to find the best measurements you can get and draw a curve from it.

[Mr Sippy]

The mix is of course a binary system. At least at the precision we can get. The congeners are negligible regarding the VLE.

I've always wondered about that. I'd have thought it measurable.


Thread title 'Predicting ABV..." requires an a priori approach. Given the equations to generate a VLE were carefully derived; it's sensible to use them and then control for other variables depending on your tolerance for error. They will get you to accuracy quicker than any other method. Using Antoine and Raoults with van Laar yields values functionally equivalent to Perry.

Using ones own observational data is perfectly fine. No arguement there.

OP started off showing the Rayliegh equation. It is the predictive method, by integration, for determining amount and abv in boiler, amount and abv collected, and average abv over any given increment. It's another fun thing for those interested in physical properties. That's where I though this thread was headed.

Regards

[prankster1590]

I found this website. Here you can get the values voor x and y

http://www.vle-calc.com/phase_diagram.html

[Hügelwilli]

The mix is of course a binary system. At least at the precision we can get. The congeners are negligible regarding the VLE.

I've always wondered about that. I'd have thought it measurable.

In case someone wonders why at his potstill the first drops are coming before the temperature is at the boiling temp of ethanol:
It is not because of the congeners. It is because of the air. Vapor mixed with air has a lower temperature. Not because air is cold but because the vapor is spread out at more space what is similar to under-pressure. And the vapor temps are lower at low pressure.
For example if you have wash with 10%abv and in the vapor is still 10% air (by volume), the vapor will have a 2.8°C lower temp than the liquid in the boiler.


prankster,
the link doesn't work at least for me.

[prankster1590]

What I don't understand is that you think it is not a binary system.
The mix is of course a binary system. At least at the precision we can get. The congeners are negligible regarding the VLE.
And "Calculating a VLE"? How? The best way is to find the best measurements you can get and draw a curve from it.

Because of foreshots and heads. Methanol and acetone and such.

[Saltbush Bill]

Whats got me beat is how all of this is going to make your booze taste any better, regardless of if you drink it as white dog or oaked for years in a barrel.
Or is it in someway going to make your still run better , be easier to run? Cuts easier to make maybe ? Or in some other way improve the quality of the final product ?

[squigglefunk]

hey man a lot of this stuff makes me scratch my head!

[NZChris]

Probably ok in theory and for entertainment but, in practice, passive or induced reflux will mean the distillate will come off at a higher abv than calculated. The algorithm is about as much use for making rum, whiskey and neutral strips and cuts as a thermometer or a parrot.

[Yummyrum]

Maybe this might be of some interest
https://homedistiller.org/wiki/index.ph ... e_Strength

[Hügelwilli]

Whats got me beat is how all of this is going to make your booze taste any better, regardless of if you drink it as white dog or oaked for years in a barrel.
Or is it in someway going to make your still run better , be easier to run? Cuts easier to make maybe ? Or in some other way improve the quality of the final product ?

More knowledge doesn't give you a better distillate, because it's all about tasting?
So all the knowledge that for example homedistiller.org has collected over the years didn't improve the quality of the distillates of the members?
Not all knowledge leads to a benefit in practice immediately. But sometimes it will. We know from the past.

[Steve Broady]

Knowledge is not always useful or applicable to practice. For example, I do not need to know exactly what chemical reactions are going on during combustion in order to dive my car, nor do I need to know the exact gear ratios inside the transmission. All I need to know is how to take whatever results those things give and put them to use most effectively. This seems analogous to me.

[Hügelwilli]

You don't. But someone else had to find out what chemical reactions happen for building the car in the quality you now enjoy.

[Steve Broady]

Actually, I think that the practical, experimental knowledge came first. Someone tried it and it worked, and they made incremental changes to improve it. It wasn’t until much later that anyone tried to explain in detail what was going on inside. And even then, that knowledge is really only useful to a small handful of people who are working on the cutting edge of the technology.

All of which I would argue applies perfectly to distillation of alcohol on a hobby level. Unless someone is trying to design a better still, and can do so with a design that is no more complex, expensive, or difficult to build and operate than what is common right now, the theory is little more than a possibly interesting intellectual diversion to a hobbyist. Which is why I originally stated that I don’t really understand the desire to calculate things which have already been calculated. But, as always, you do you. Enjoy, and stay safe!

[prankster1590]

I think it will benefit quality, reproducability and efficiency if you know the science behind the process

[Hügelwilli]

- For example if you use silicone or PTFE should be science based, you can't find out by taste.
- Or if you think you should cut foreshots because it makes you blind or for other reasons, you can't find out by taste.
- To calculate abv and amounts can answer questions like, if you should build your column longer or not.

[NZChris]

I think it will benefit quality, reproducability and efficiency if you know the science behind the process

Quality, reproducibility and efficiency comes from your own record keeping. Every wash and distillation should have records kept in such a way that you can build knowledge of how your ingredients and stills behave. E.g., when I do multiple stripping runs, I only pay careful attention to the first one, then use the temperature of the wash to shut down the rest of the strips at the same point without having to do any checks. That temperature is not useful for the next ferment, or for different products, or for another distiller.

[StillerBoy]

Quality, reproducibility and efficiency comes from your own record keeping. Every wash and distillation should have records kept in such a way that you can build knowledge of how your ingredients and stills behave. E.g., when I do multiple stripping runs, I only pay careful attention to the first one, then use the temperature of the wash to shut down the rest of the strips at the same point without having to do any checks. That temperature is not useful for the next ferment, or for different products, or for another distiller.

Thank you Chris for the clearest explanation of enjoyable distilling and product making..

At the hobby level, that all that is important, the rest is wasteful time in creating confusion..

Mastering that simple method (note taking) and applying the principal at all time.. that's where the majority fail in learning the hobby..

Mars

[haggy]

prankster,

Experience in running your still and keeping good records is most valuable.

But we do know something more about your distillate ABV question.

To get a good estimate of the vapor ABV % coming off the pot and/or the thumper, go to the Calculator Section ( off the Parent Site ) and click on Haggy's Potstill +Thumper calculator.

Use the input data already there or add your own, click on Calculate Results and you will see results. At the bottom is a table of key info as the run proceeds. On the left half of the table is the pot info with both the instantaneous distillate vapor ABV% and a cumulative ABV% coming off the pot as the run time increases. This info is valid for a pot only stripping run.

On the right side of the table is the thumper information. All the vapor from the pot goes into the bottom of the thumper, condenses, gets mixed and exits the thumper as a more enriched vapor. Both the thumper instant vapor ABV% and cumulative are given. Changes in the input data let you see the effect on the distillate ABV.

The amount of vapor from the pot and thumper at each time increment is also given. This is in US gal, I have another version of this calculator using either US gal or liters.

See my thread "Hail the Mighty Thumper " for more information and for the basis of the calculations.

haggy

[Ben]

Knowledge is not always useful or applicable to practice. For example, I do not need to know exactly what chemical reactions are going on during combustion in order to dive my car, nor do I need to know the exact gear ratios inside the transmission. All I need to know is how to take whatever results those things give and put them to use most effectively. This seems analogous to me.

Your right, it is very analogous.

A gearhead would use all this information to make the experience more enjoyable for themselves, they would use the knowledge to diagnose problems and make improvements, they would enjoy fantasizing about what is going on in the engine and gearbox at every roll out from a stop light. They would consider you unenlightened for the lack of understanding, and imagine your life to be less fulfilled. Stilling is the exact same thing, the OP is learning about how his still is running (even if it is only theory), getting under his hood and tinkering... most car guys start with an oil change, this is no different. The fact that he is aware of changes through the run puts him far ahead of most people with 8 posts.

If mathematics and chemistry brings joy to him then more power to him. If it enhances his experience that's what matters, we aren't all content with making a ferment into a clear liquid and getting sauced off it. Different ways of finding joy. We all bring something to the table, if his strength is math then lets be thankful that it's one more person to solve the equations we don't understand.

A big part of scientific experimentation is understanding why and where the experiment deviates from the model theory. OP has short-cutted a bunch of that by posting this thread and getting results from other peoples experiments.

Distilling is extremely complex, fermenting is extremely complex, the understanding has to start somewhere, we aren't born with this skill. You don't need scientific understanding to make a good product, but it might help some people find another layer of challenges and interest.