Boiling off the alcohol

[_john]

I am trying to find out the best way to evaporate alcohol from liquids without changing the flavor of the original liquid. In my case this applies to "real" mirin which is 14% a.b.v. I have tried boiling it at a rolling boil until I perceive no alcohol in the vapor and I can so longer set the vapor alight. The resulting mirin is noticeably caramelized in taste and has lost quite a bit of volume. If time is not an issue is it better to keep the liquid at say 70°C (above the boiling point of alcohol I think) until it cannot be lit? If I boil off the alcohol from 1000ml of mirin can I add water at the end to bring it to 860ml (minus the 14% alcohol) and have the taste remain similar to the mirin before boiling?

A scene from my everyday life

[heidih]

What are you trying to achieve - is it an alcohol issue or a flavor destination?

[nickrey]

The boiling point of alcohol is 79C, of water 100C. It seems easy: boil the mixture and it goes away. Unfortunately, this is very wrong. Check out this document from the USDA. Retention codes 5001 to 5010 look at % retention of alcohol across different cooking methods. Basically flaming it leaves behind a massive 75% of the alcohol, simmering it for 1 1/2 hours leaves 20% of the alcohol, and so on. It takes two and a half hours of simmering to remove 95% of the alcohol.

Bottom line, you will have to cook for a very long time to remove all the alcohol. At a simmer, which is below boiling and should have less effect on your mirin and minimise water loss, it will take around 2.5 hours to remove 95% of the alcohol. Hope this works for you.

Edited January 24, 2012 by nickrey (log)

[_john]

@heidih - The goal is to be able to use the mirin in dishes that are not heated and where I don't want the alcohol. So a flavor issue and an alcohol issue.

@nickrey - Thanks a lot this looks like the data I was looking for. Obviously I will have to do a lot more experimentation to find out what the tolerable retention is. In the document is says that alcohol stored overnight without heating retains %70 of the alcohol. Could just leaving it uncovered for several days cause a lot of the alcohol to evaporate?

[nickrey]

Personally, I'd simmer it for three hours to achieve what you are looking for. If it is losing water as you cook, add some back in.

[Mark Donnelly]

The boiling point of alcohol is 79C, of water 100C. It seems easy: boil the mixture and it goes away. Unfortunately, this is very wrong. Check out this document from the USDA. Retention codes 5001 to 5010 look at % retention of alcohol across different cooking methods. Basically flaming it leaves behind a massive 75% of the alcohol, simmering it for 1 1/2 hours leaves 20% of the alcohol, and so on. It takes two and a half hours of simmering to remove 95% of the alcohol.

Bottom line, you will have to cook for a very long time to remove all the alcohol. At a simmer, which is below boiling and should have less effect on your mirin and minimise water loss, it will take around 2.5 hours to remove 95% of the alcohol. Hope this works for you.

Very informative and completely counter-intuitive. I'll just live with the consequences of not removing alcohol entirely as I, like most people, have better things to do than to watch a pot simmer for several hours. I mean, I could knock off a 1,000 tweets in that time.

[Norm Matthews]

Water will carry flavors such as in making soup or coffee, Fat carries flavors that water won't which is one reason why fried food tastes different than boiled foods and alcohol brings out flavors that neither of the other two will which is one of the reasons why the flavor changes when heating alcohol with food.

I know about the USDA report on alcohol but I'd like someone to explain why an experiment I did apparently show different results. I found a hydrometer in an antique store and thought it was calibrated to measure sugar levels as for determining how alcoholic grape juice should be after fermenting. When I got it home I discovered it was a customs house hydrometer that was calibrated to measure alcohol content of a liquid. Out of curiosity I measured brandy and it showed 70 proof. After adding two parts water as one might to when adding alcohol to a sauce, it measured 20 proof and then after only simmering for three minutes, it measured no alcohol.

Edited January 24, 2012 by Norm Matthews (log)

[nickrey]

The concept of proof is different from percentage akcohol. Proof was when you could put the liquid on gunpowder and the gunpowder would still light. This is 100 proof, which equates to around 57% alcohol by volume. It is also measured at a particular ambient temperature. Move any of these variables and you have a problem estimating percentage alcohol. I suspect the authors of the USDA scientific study had more sophisticated measuring equipment than the hydrometer shown.

[Norm Matthews]

I know that the hydrometer is calibrated to be accurate at (I think) 70ºF and my experiment did not by any means attempt to be scientific but I would suspect that since it does roughly measure alcoholic content that there should me some kind of rough correlation.

It also seems to me that USDA numbers are misleading by quoting alcohol by percentage of itself instead of talking about amount by volume. If you add a tablespoon of wine at 12.5 percent alcohol to a cup of liquid, you have automatically reduced the amount by volume by 16 times. USDA seems to consider that to be 100 percent alcohol but even if it isn't heated the amount of alcohol in a cup of chicken broth is now 1.56% by volume. See what I mean? Even if my math figures are not accurate, the amount of alcohol by volume is reduced greatly. If a cup of sauce is a serving for 4, that is now less that 0.4% per serving by volume even though it is still 100 % alcohol. Heating it may reduce the alcohol by only 25% that is now 75% of the alcohol remaining but it is also now .3 by volume. Unless I am mistaken, that is legally considered non-alcoholic.

If my thinking is out of line, I'd like someone to explain it to me.

Edited January 24, 2012 by Norm Matthews (log)

[nickrey]

Thank you for pointing this out Norm, the original article was published outside the range where the journal made papers available so it was not accessible to look at their method.

Mirin, which was the original object of discussion here, contains 20% alcohol. Flaming mirin, therefore, would leave 75% of the alcohol or leave it as 15% alcohol by volume. This is not legally non-alcoholic, so you need to look at different approaches. In the US, it seems that non-alcoholic means less than 0.05% alcohol by volume. To get to this, you'd need to reduce the existing alcohol to 25% of its starting point, ensuring that the water content does not change as well. The time required to reduced the alcohol content to 25% is one hour of simmering, as long as you can make sure that the water content remains the same.

[_john]

Is there any way to measure alcohol content that is accessible to the home cook or professional kitchen?

[nickrey]

You'd run into a problem with this. Typical measuring devices (eg. an Alcohol Refractometer, hydrometer, or vinometer) measure sugar content and use this as an estimate of the alcohol content. In beer, for example, they use a hydrometer before and after fermenting and work out how much of the sugar has been converted into alcohol.

As Mirin has glucose added as well as alcohol, it's really going to throw any measurement off. This problem is also found in measuring alcohol content in sweet as opposed to dry wine.

Check out this link for a description of different measures.

[threestars]

Very nice topic. I learned a lot from this.

[Mikels]

I think the reason it is so difficult to get rid of the alcohol is that it forms an azeotrope with water. An azeotrope consists of two or more compounds in a specific ratio that can't be separated by simply distilling it. The ethanol-water azeotrope is about 4% water and 96% ethanol. This boils fractionally lower than pure ethanol. That is one of the reasons you lose water when you boil the mixture. Even then, water is evaporating at about half the amount when it is boiling (vapor pressure at 79C is 341mm Hg and, of course at boiling it is 760mm Hg at sea level). If you don't want to lose too much water, you would need to run a still that has a column to condense the water with the mixture kept above 79C (don't let the Feds find out or they will send Eliot Ness's counterpart). Those are the physical reasons it is so difficult to get rid of the alcohol and why you lose so much water.