I have a secret for you: Many blogs and websites purposely exaggerate facts, and hide certain very important aspects of a subject, in order to make their posts more exciting, provocative, etc. The idea is not for you to actually learn the facts, but to be “shocked” into sharing the article. This happens with cooking myths all the time, especially when they are based on what passes as common knowledge. It is common knowledge, for example, that when you put alcohol into a cooked dish, the alcohol “burns off” completely, through heat evaporation, so that no alcohol is left in the dish.
Well, thanks to some research published by the USDA, including the study, Alcohol Retention in Food Preparation by Augustin, et al. 1 we know that this is not true. It takes a lot of cooking to burn off all the alcohol in a dish. After these new facts came to light, a lot of articles started appearing letting us in on the “shocking” news that the intoxicating qualities of alcohol in a dish would still be there with normal cooking times!
Can You Get Drunk Off of Wine in Cooked Foods?
The idea that you will get intoxicated from alcohol used in say, a vodka sauce for pasta, or some dish with wine in it, is poppycock! Why? First of all, let’s look at what the research says. If you add wine, or any other source of alcohol to a boiling liquid, and then stop cooking it by removing it from the heat, 85% of the alcohol will remain. This is the point where most people seem to stop recounting the facts. However, if you continue simmering for 15 minutes, 40% of the alcohol will remain. That means that 60% of the alcohol is gone (this counts as most, in my book). After around half an hour, 35% will remain. After one hour, 25%. One and a half hours, 20%. Two hours, 10%. Two and a half, 5%. As you continue simmering, there are diminishing returns on the alcohol removal, but after 3 hours it is safe to say that most of the alcohol is gone. Different methods of cooking, and when, precisely, you add the alcohol, will change the amount lost and the time it takes.
Alcohol Burn-Off Chart from USDA
The USDA actually gives this data in chart form. 2Note that various other conditions are given. Most noticeable is that different values are given for baked items where alcohol is used but not stirred into the mixture. Also, you’ll notice that the alcohol evaporation for flamed dishes is lower than you might expect. You find out below why this is so.
|Preparation Method||% Alcohol Retained|
|Alc. added to boiling liquid then removed from heat||85%|
|No heat, stored overnight||70%|
|Baked 25 minutes, alc. not stirred into mixture||45%|
|Baked, simmered dishes, alcohol stirred in|
|15 minutes cooking time||40%|
|30 minutes cooking time||35%|
|1 hour cooking time||25%|
|1.5 hours cooking time||20%|
|2 hours cooking time||10%|
|2.5 hours cooking time||5%|
Shocking? Not so much. What most people completely ignore is the actual amount of wine or spirit we use in dishes, and thus, the actual volume of alcohol in the overall dish. Wine has about 12 to 14% alcohol by volume (ABV). You will tend to use more wine in dishes than spirits like vodka or whiskey, which we will say contain about 40% alcohol by volume (80 proof, but spirits vary).
While most people, including restaurants, over-pour wine, a “serving” is considered about 5 ounces, which is about 0.7 fluid ounces of alcohol on the high side (14% ABV). Would you use that much wine in a dish? If it was a wine reduction used to build a sauce, maybe, but then you’d be reducing down the wine until it was like a syrup, cooking off a great deal of the alcohol. Even so, a wine reduction will probably end up being the most concentrated source of alcohol in cooking, but you’d use so little of it on the dish the actual amount of alcohol would be negligible. But lets look at something more typical for a home cook. Say, Marsala sauce, which uses Marsala wine, a fortified wine from Sicily, which may contain up to 20% alcohol by volume.
Typically, to make a Marsala, the wine is added to a pan that something else has been sauteed in, such as chicken, onions, mushrooms, etc. and used to deglaze the pan, whereupon chicken broth or other broth is added. Even if you were to make a large volume of sauce for four people, you’d probably not use more than 3/4 cup of wine. Let’s be generous and say you use a cup of wine. And let’s also say your Marsala wine is 20% ABV, meaning the wine you use in the dish contains 1.6 ounces of alcohol. This means that one ounce of it contains 0.2 ounces of alcohol. You would then, typically, add the same volume of broth, if not a bit more. So we have two cups of liquid, plus whatever other moisture is already in the pan. This means that the alcohol is diluted by the same amount of liquid. One ounce of this mixture would contain 0.1 ounce of alcohol.
But you do not stop cooking Marsala sauce right after you add the wine. You continue cooking until the sauce is reduced by half. This may take 5 to 10 minutes, or longer, depending on the pan (sauces reduce faster in wider more shallow pans). Although you are concentrating the sauce, and so concentrating the amount of alcohol, you are also evaporating some of the alcohol.
Let’s say, conservatively, when we’re done, we have one cup of sauce, which contains 85% of the original alcohol; in other words, 1.3 ounces of alcohol, assuming you do not add anything to the sauce before serving. Now, we spoon our sauce over grilled chicken. Each person gets 4 tablespoons of sauce. That is 2 ounces of sauce. Each person gets 0.3 ounces of alcohol. So, each person gets less than half a “standard” serving of alcohol. If we’re being quite generous in our interpretation. Chances are, in reality, the actual amount of alcohol will be much less. Remember, we started with a fortified wine that was very “dry”…on the high side of alcohol content. Most of the wines we use will have a good deal less alcohol. It is quite unlikely that you would become intoxicated from a wine sauce. In this example, we used a lot of wine and we ladled it on generously! In reality, the amount of alcohol in any dish made with wine would be negligible, with the amounts typically used.
If we were making something like Penne with Vodka sauce (I mentioned vodka sauce) we would end up cooking it for at least 20 minutes when all was said and done, and the most vodka you’d see in a recipe is 1/4 to 1/2 cup. So, for our 80 proof, 40% ABV vodka, we’d have, at most, 1.6 ounces of alcohol, and we’d cook off at least 60% of that, meaning the whole dish would have only 0.6 ounces of alcohol. To that we would have added a whole lot of tomatoes (say a 28oz can of crushed), and at least 1/2 cup cream. This means that our 0.6 ounces of alcohol is diluted into about 3 cups of sauce. You’d get 0.2 ounces of alcohol if you ate a whole cup of sauce. That is less than half a drink of alcohol. So, again, you are not likely to become intoxicated by vodka sauce. And, of course, depending on your cooking method, you might cook off more of the alcohol. I, for example, add the vodka to the butter, onion, and garlic before I add any tomatoes, and allow the vodka to cook down first. Then I cook it for up to 20 minutes, or longer.
Besides all this, if you are concerned about the amount of alcohol left in a cooked dish, most of the time you should be able to simply cook it longer. If a sauce is reduced more than you want, you can add more of whatever other liquid is being used, such as broth or cream.
Flambé to Burn off All of the Alcohol?
To flambé is to add an alcoholic liquid to a dish and then light it on fire. This is sometimes called flaming. Although most people would expect burning the alcohol off would remove more alcohol, you may have noticed in the chart above that flaming retains more alcohol than simply cooking for 15 minutes.
Since wines or beers do not contain enough alcohol to ignite in this manner, only alcoholic spirits (liquors) such as rum, whiskey, or vodka will work. This is most famously used for flambéed desserts such as Bananas Foster, Cherries Jubilee, and Crêpe Suzette . Some people think that this will completely burn off all the alcohol in very short order, much quicker than just simmering the food, but the flavor will be left behind.
However, in order to actually ignite the dish, it must be heated to around 180°F, so that the alcohol begins to rise as a vapor. When you ignite it with a match or other source of flame, it is actually the vapors you are igniting. These vapors will have risen whether or not you set them on fire, and the fact that they are burning makes no difference. The question becomes how much of the liquid fuel is lost due to evaporation, and how much due to combustion. While both may be possible, we do not know, just based on observation, what is occurring.
So, are you really “burning off” more alcohol than can be liberated by the cooking process? Many say no. For those, the real question in flambéing is how it affects the flavor of the food, not whether it burns off the alcohol. Some say the chemical changes that take place due to the heat of the flame, including caramelization, makes for a better tasting dish. Others say that it does nothing, and is nothing more than an overly-dramatic and flashy bit of presentation that some chefs use to wow people. Some of the people who say it is just flash and has no effect on flavor also say it ruins the dish! On the other hand, those who say it elevates the dish can rarely explain why it does so. It is not clear at all that people can typically discern the differences in taste between a flambéed and non-flambéed dish.
This study, by Hansen, et al. 3 examined the amount of alcohol (ethanol) burned off by ignition and non-ignition normal cooking, the amount of browning caused by ignition and non-ignition cooking, and taste differences as reported by panelists. In all instances, no significant difference was found between the flambéed and non-flambéed methods, except that when a simple vodka-only preparation was ignited, and compared to a heated but not ignited preparation, the ignited vodka showed only a 24.7% loss of ethanol, compared to a 34.7% loss in the heated and not-ignited vodka. In both, the alcohol loss was reported to be due to heating, rather than combustion.
Flambéeing Makes No Difference to Flavor
In a caramel sauce preparation, no significant differences in alcohol loss, browning, or sensory reactions were recorded. It seems, according to this, and other well-carried out studies, flambéeing really is nothing more than a dramatic and quite dangerous bit of show-business. However, in terms of taste, the presence of other ingredients, and the amount of aroma present in the dish other than the aroma of the spirit, may have an impact, so that other aromas might mask any small difference in flavor caused by ignition. In the real world, this would mean that most of us will probably never notice any real flavor difference.
It is worth noting, by the way, that in the study linked above, the non-ignited caramel sauce, was made by melting butter, and then adding sugar and heating on a 130°C hotplate (266°F) for two minutes. Then, the vodka was added and heated for 15 seconds before the pan was covered and removed from the heat. Yet, the authors reported a 34.7% loss of alcohol. That is significantly different than our 15% loss, above, in similar conditions. Does this have to do with the ingredients themselves? Butter and sugar is a lot different than a mostly liquid preparation. Or, should we expect more alcohol loss than we assumed from the previous research? Does it matter?
- Augustin J, Augustin E, Cutrufelli RL, Hagen SR, Teitzel C. Alcohol retention in food preparation. J Am Diet Assoc. 1992;92:486–488
- United States. USDA. Nutrient Data Laboratory, Beltsville Human Nutrition Research Center (BHNRC), and Agricultural Research Service (ARS). USDA Table of Nutrient Retention Factors Release 6. Dec. 2007. Web. 7 Apr. 2014.
- Hansen, Christine E., Misha T. Kwasniewski, and Gavin L. Sacks. “Decoupling the Effects of Heating and Flaming on Chemical and Sensory Changes during Flambé Cooking.” International Journal of Gastronomy and Food Science 1.2 (2012): 90-95.