# Alcohol Dilution

## Recommended Posts

I wanted to get some clarification on how commercial operations are diluting their alcohol, as I keep reading conflicting info. Which of the following are you using?

1. 500ml of 80% Alcohol + 500ml Water = 1L 40% ABV

2. 500ml of 80% Alcohol + q.s. Water to 1L = 1L 40% ABV

3. Use trial and error, by repeatedly adding water and testing with spirit-meter

4. ???

##### Share on other sites

You must use the gauging tables provided on the TTB website

http://www.ttb.gov/foia/gauging_manual_toc.shtml

##### Share on other sites

Thanks Bryan, I appreciate the patience with this Beginner's section of the forum.

##### Share on other sites

From a theoretical calculation, due to the non-ideal mixing behavior of alcohol and water, you will need .519 liters of water to produce 1 liter of 40%ABV. (for your example scenario.)

I have written some calculators which compute these relationships. But I am not using them, right now, in a commercial distillery.

Glenn

##### Share on other sites

We start with this bit o math: Known volume x known proof divided by desired proof = desired volume

But you have to adjust for temperature as hydrometers are typically calibrated at 60 degrees ferenheit.

##### Share on other sites

The calculations can get you in the ballpark, but given the minuscule tolerance allowed by the TTB, the only option is getting it close and then "repeatedly adding water and testing with spirit-meter." The first few percent can go pretty quickly, but that last 2% can take a while.

In the words of Erik Martin at Artisan Spirits: "Proofing Is A Bitch."

##### Share on other sites

Once you understand the gauging tables (which account for "contraction") you will choose to proof by weight.

Contraction is the reduction in volume when alcohol and water are combined. 500 ml water and 500 ml alcohol when mixed will make about 950 ml of something nearer 82 proof (that's an educated guess, not the result of a calculation).

Alcohol expands 5 times as much as water for the same temperature change...and the change is not insignificant. For example, 500ml of the 160 proof you suggested occupies 505.6ml at 80F. Also, graduated cylinders are remarkably inaccurate devices - check the specifications. You can't use them to do any meaningful metrics. The limits of error for bottling proof are +0, -0.15% (or -0.3 proof), or 79.7 proof to 80.0 proof if you're bottling at 80 proof. If you (1) use an expensive graduated cylinder, (2) correct for temperature, (3) read accurately, (4) adjust for the fact that the "metric system" measures at 20C, while the IRS measures at 60F (15.56C), you'll be lucky to get within +- 0.5% - 6 times the error you can tolerate. Add to that the fact that you can't accurately measure volume in large tanks conveniently, and you arrive at the conclusion that proofing by volume is to be avoided.

By contrast, the weight of alcohol and water do not change as a function of temperature, AND scales are very accurate and easy to read. An inexpensive scale can divide its range into 6000 parts. That's resolution of .016% of full scale. If I make batches at about 1/2 of full scale, and assume +- 1 LSD inaccuracy, I'm right in the zone where I can hit 79.7 to 80.0 proof every stinking time just by gauging, making a calculation, and adding an exact amount of water. No screwin' around.\

Let's go back to your example. You can do this with a \$40 table top scale: 519g of water, and 431g of 160 proof alcohol will combine to make exactly 950 grams of spirit at 80 proof - 1000mL at 60F.

Max, you're asking good questions. Here's a shameless plug:

You might want to consider attending the Week-long ADI Distilling Workshop

You may also want to consider applying for the Michael Jackson Internship

##### Share on other sites

I'm a bit confused.

We will gauge by weight and I know how to figure PG from hydrometer reading, temp readings, and weight.

In your example, you got those figures from the gauging manual?

950 grams is 2.094395 pounds, and at 80 proof (.10093 PG/lb) that's .21139 PG...at 60 degrees.

How did you figure how much is water and how much is alcohol? I can't seem to figure it out.

So, to check proof prior to bottling, I will take the hydrometer reading and adjust for temp. That will give me accurate proof, correct? Then, to check for an accurate fill, can't I use a graduated cylinder for the 750 ML?

##### Share on other sites

How did you figure how much is water and how much is alcohol? I can't seem to figure it out.

Table #6, after taking corrected proof measurement at 60*F and knowing your wine gallons.

##### Share on other sites

I'm a bit confused.

Then, to check for an accurate fill, can't I use a graduated cylinder for the 750 ML?

This should be in another section about filling, but whatever.

(and I'll get to the other questions after you've reviewed the PM thread from earlier)

Correct. The graduated cylinder is useless for measuring volume.

Measure the dry weight of each of a set of bottles, mark each one with its tare weight. Then before and during the bottling run, fill these bottles, and weigh the result. If you're bottling 80 proof at 750mL, then the weight will be 712 grams higher than the tare weight. (Why? Because 750mL at 80pf weighs 712 grams - table 6)

Now record those results for your QC paper-trail. Make corrections to the bottling line if needed.

Here's a quick trick. After weighing each bottle, dump the contents back into the bottling tank, and re-weigh each bottle, noting the residual weight of spirit that is still in each bottle - usually 1 to 2 grams. Record that as well, and subtract that value with the tare weight when you run the next test...because now the bottles are wet, and if you add exactly 712 grams more, you'll be at 714 grams...and you wouldn't want to correct the bottling line because you didn't account for the bottles being wet.

This is even worse: You decide you'll rinse all of your bottles before putting spirit into them, but you don't dry them out. Try the same test above, but with water. Weigh the dry bottles, rinse them and drain them. If your shop is like ours, you dump the water, but don't drain them for very long - only long enough to put them on the filling table. Weigh them wet. You'll find they weigh 1 to 2 grams more wet than dry. What does this do to your bottling proof?

We find (and the manual confirms) that this small amount of water will reduce proof in the bottle by .1 proof per 1 gram water. Therefore, we (a) dump the bottles a second time just before we stick them into the filling machine, and ( we adjust our filling tank upward by .1 proof to account for this, and © when we do the Standards of Fill QC Check (mentioned above), we also re-check the proof.

Last week, we found that my youngest son was omitting step (a) above, and when we checked, we were below 79.75 proof (our lower limit) on two bottles tested, so we dumped the entire bottling run up to that point, and then continued. The small number of bottles (about 100) reduced the bottling tank as well, but the combination was still high enough to test-out at 79.8 proof on the remainder of the run.

Will

##### Share on other sites

I'm a bit confused.

950 grams is 2.094395 pounds, and at 80 proof (.10093 PG/lb) that's .21139 PG...at 60 degrees.

How did you figure how much is water and how much is alcohol? I can't seem to figure it out.

TTB Table 6 -

First, you need to know that one gallon of water weighs 8.32823 LBs. You should throw-away everything you have that's in the metric system and learn to work in GALLONS and LBS - but you won't do that (neither did I), so one LB = 0.45359 KG, and one Liter is exactly 0.264172 Wine Gallons. These are TTB accepted values. These are different values than you will find in the scientific literature because the scientific literature is not a tax authority...and they're working at 20C with standards that have been adjusted several times during the last century. These values are the values you will use.

Examine Table 6 carefully...just look at the left-most 3 columns. Just look at the rows where the proof ends in 0: 10, 20, 30... Note that the alcohol volume is always exactly 1/2 the proof, but the water volume is always some funky number...and it's somewhat larger than you would expect. The numbers always add-up to something just over 100. This is contraction. The volume decreases when you mix alcohol and water, and the make-up is ALWAYS WATER, not alcohol. So, even if we learn in the future that it's the alcohol that's "at fault" for the contraction, we'll still blame it on the water.

Now, let's look at the numbers for 80 proof:

40.00 volumes of (pure 200 proof) alcohol and 63.42 volumes of water are needed to make 100 volumes of finished product. 3.42 volumes extra are needed to account for contraction.

You want to reduce a given amount of high-proof spirit to some other proof for bottling.

You can approach this problem two ways: (1) you can break the original spirit into its alcohol and water equivalents, and do likewise for the ending proof, then compute the differences, or (2) you can read the instructions for Table 6 in the Gauging Manual and do exactly what they do. Both methods arrive at the same result. (1) may be more intuitive at first, but (2) is faster.

This is the passage in question for (2):

To do this, divide the alcohol in the given strength by the alcohol in the required strength, multiply the quotient by the water in the required strength, and subtract the water in the given strength from the product. The remainder is the number of gallons of water to be added to 100 gallons of spirits of the given strength to produce a spirit of a required strength.

Now, scale the 100 gallon result to fit the situation you're working with. It's really easy after you've done it a few times.

(it's also really easy after you send the entire Gauging Manual through an OCR program, validate the results, and put the tables into a spreadsheet.)

Do it a few times,

Will

##### Share on other sites

I still don't understand the math behind this....

(1) you can break the original spirit into its alcohol and water equivalents, and do likewise for the ending proof, then compute the differences

Which I'm sure would tell me how to do this:

519g of water, and 431g of 160 proof alcohol will combine to make exactly 950 grams of spirit at 80 proof - 1000mL at 60F.

##### Share on other sites

Notwithstanding the fact that method (2) is the prescribed method, I was referring to the somewhat more intuitive approach (for me) which is to compute how much alcohol and water will be present in the final proof, and how much you already have in the given proof, and compute the differences.

Later, I'll show a shortcut.

500mL @ 160 pf -> 1000mL @ 80 pf

160 pf is 80% alcohol by volume, so we can break it down to its component parts.

100% alcohol weighs 6.6097 LBS / Gallon

100% water weighs 8.32823 LBS / Gallon

for the 160 proof (using Table 6)

.80 x 6.6097 = 5.29 LBS / Gallon alcohol

.2287 x 8.32823 = 1.91 LBS / Gallon water

for the 80 proof

.40 x 6.6097 = 2.64 LBS / Gallon alcohol

.6342 x 8.32823 = 5.28 LBS / Gallon water

Let's look at the alcohol:

5.29 / 2.64 = 2.00 - so every unit of volume at 160 proof will make 2 units volume of 80 proof.

For simplicity, let's think through making 2 gallons - we'll adjust the quantity later.

2.00 * 5.29 = 10.58 LBS water total in the final batch.

Every time I use one gallon of 160 proof, I'm also contributing 1.91 lbs of water toward the amount I need in the final product. So:

10.58 - 1.91 = 8.67 LBS water extra needed.

So let's fire-up the scale:

the 160 proof weighs 5.29 + 1.91 LBS per gallon, so measure-out 7.2 LBS of 160 proof stuff.

Now add water until we have 2 * (2.64 + 5.28) = 15.84 LBS total.

You can scale this to any size batch, or convert to SI units using the coefficients I posted earlier (but don't use coefficients you find in the scientific literature - use the TTB values).

That walks us through the details or mechanics of what's happening, but did you notice the shortcut we could have taken?

Once we know the ratio of alcohol to alcohol (given proof / ending proof) we can just multiply that ratio by the starting quantity, which equals the ending quantity we will be making, then look-up the weight of each gallon of the ending proof (using Table 5), multiply that by the ending quantity, then add water until we get that weight. Duck soup.

Like this:

8 gallons of 140 proof -> 80 proof to bottle:

140 / 80 = 1.75

so 8 gallons times 1.75 will make 14 gallons of final stuff.

80 proof weighs 7.93 LBS / gallon (Table 5)

14 * 7.93 = 111.02 LBS total

Now tare your scale and dump the 8 gallons of 140 proof. According to Table 5, that will weigh 7.41 LBS/Gal or 59.28 LBS (though we don't really need to know that). We simply add water until the whole thing weighs 111 LBS.

See, no screwin' around...

Will

##### Share on other sites

Ok. I worked on this for 3 hours yesterday and an hour today. I must be an idiot. It's making me crazy....really.

I understand your post and the PM from last night. I understand figuring out how much water to add to cut to another proof.

But for the life of me, I can not figure out how you took 950 grams of 80 proof spirit and figured out how much was 160 proof alcohol and how much was water.

##### Share on other sites

John,

Max Action's original question asked about diluting 160 proof to 80 proof. He asked (multiple-choice style) how this is done correctly, and used 500ml as the starting quantity. Several of us did the math, and determined he would need 519 grams of water, and the result would be 1 liter with a weight of 950 grams.

Your question above does not reflect what he asked... You're looking at the problem from the other side...how much 160 proof is in 80 proof, and how much is water? (unless you simply phrased the question incorrectly).

Converting the 950 grams back to it's source components is easy.

Start by converting 950 grams of 80 proof to volume (hint: it's one liter)

Now, how much pure alcohol is in there? (hint: 40% - 400mL)

How much alcohol at 160 proof can I make if i have 400mL of pure (200pf) alcohol? (hint: 400/.8 = 500ml)

The somewhat trickier part is the determination of exactly how much water is going to be needed, and that's where the TTB tables come in...

80 proof is 40% alcohol, and 63.42% water.

Use the coefficients from earlier posts for conversions between Imperial and SI units.

Will

##### Share on other sites

I was thinking of it backwards.

Thanks for the lesson. Send me a bill.

Seriously...Thanks!

##### Share on other sites

My son is named Bill - I'd send him to you, but he's far too valuable to me here!

When you feel like it, conquer this one:

You're diluting 350 LBS of 135 proof to 86 proof for bottling, so you figure you'll be making 73.6 gallons of finished product, which will weigh 580.6 LBS. You tare the scale, dump the stuff into the proofing tank and turn-on the water...but then you get distracted. By the time you regain consciousness, the scale reads 603.2 LBS.

What's the present proof?

How do you fix it?

##### Share on other sites

Yes Will, Thanks.

Bob

##### Share on other sites

• 1 month later...

Will....

On your example...83 proof? Not sure how to fix it other than add some more alcohol or to distill it again.

Could be wrong, but I figured out PG then divided it by total weight to give me PG/lb.

##### Share on other sites

• 2 weeks later...

Hey, Will.

I just wanted to thank you for your incredibly specific explanations in this thread. Really - amazing. It should not be "amazing" since this is something that any prospective distiller needs to know, but...that's a lot of really valuable information and impressive knowledge knocked out here for us.

Thanks again.

Andrew

##### Share on other sites

From the original information we can determine that there was 31.66 gallons of pure alcohol. Knowing we overshot the weight by 22.6#, with water weighing 8.33 #/gallon, we know there we added 2.7 gallons. So the final volume was 76.3 gallons. That means (31.66/76.3) * 200 = 82.9 proof.

I'm going to assume that I have some of the original 145 proof spirit that was not part of the reduction in proof. Basically you'd want to how much (by weight) of that 145 proof spirit to add to bring it up to the original target of 86 proof. But I haven't completely figured out how to calculate that.

##### Share on other sites

One of the challenges is to understand and deal with contraction. The amount of water required to reach a given volume reaches a maximum at 121 proof. Your approach does correctly determine the excess volume of water added, but refer to the tables to find the solution.

Let's accept the notion that we do have remaining spirit (at 135 proof, not 145), so yes, let's add some.

Note that the intermediate result (present proof) is not relevant to the underlying question: "how do we fix it?"

##### Share on other sites

exactly the type of mis-information we hope to avoid here...that dilution calculator is simply wrong. any commercial distiller who does his process in that manner will be paying large fines.

if you don't have a basic permit, ask questions, but please don't act like you know the answers.

will

##### Share on other sites

• 1 month later...

Will,

Awesome details on proofing! What is 'required' for the proofing vessels and scales? I'm looking to purchase a vessel for proofing and a scale, but don't want to waste time and money purchasing 'non-approved' items. It is very well detailed what is required for the hydrometers and thermometers, but I can't find much about vessels and scales.

Thanks!

##### Share on other sites

• 1 year later...

Ok I had to stop reading 3/4 of the way down on page 1, does anyone know a good optometrist? I am so cross eyed now. If I had actually paid attention in any of my math classes would I have actually seen this problem?

## Create an account

Register a new account