Saturday, January 31, 2015

Filtration: a Way to Improve Whiskey, or Destroy It?

Filtration is a part of whiskey-making that is rarely discussed. Virtually all whiskeys are filtered in some way before bottling. Mostly cosmetic, it is the last chance to make sure your whiskey glistens on the shelf.

Filtration also affects flavor. Some believe that's a bad thing and a few whiskeys choose to be slightly less beautiful so they can be slightly more flavorful.

The manipulative filtration of whiskey turns out to be a special skill of Michter's Willie Pratt, which he demonstrated through ten tastes of a 10-year-old rye whiskey Michter's was preparing to bottle. By varying the filter material, its density, the use of charcoal, and other factors, he produced ten distinctly different flavors from the same whiskey sample. The tasting panel's job was to select the one that showed best.

Everyone does this sort of thing, Michter's was just nice enough to provide a demonstration.

'Rectification' means to fix or correct. In the late 19th century, 'rectifiers' bought whiskey from distillers and 'fixed' it, i.e., they made it more palatable and, therefore, more marketable through blending, redistilling, filtering, and flavoring.

Some rectifiers took good whiskeys from different makers and batched them together into a product greater than the sum of its parts. Others created what came to be called compound whiskey which was, in fact, not whiskey at all but neutral spirit flavored to resemble whiskey. In those days before truth-in-advertising regulation it was passed off as the real thing, much to the chagrin of whiskey purists.

All of these techniques still exist but are regulated and must be disclosed when used. Filtration is the exception. It has always been considered a legitimate part of the production process. Since some kind of filtration almost always takes place as one of the final steps before bottling, it doesn't have to be disclosed. Even when it is deliberately used to alter the flavor of the spirit, the producer does not have to reveal what was done, how it was done, nor even that it was done at all.

The most ubiquitous type of filtration is chill filtration, the purpose of which is to remove certain fatty acids that can cause the whiskey to appear cloudy, especially when the bottle is cold. The whiskey is cooled to about 32° F and passed through a fine filtering medium (e.g., silk), usually coming into contact with activated carbon. This is supposed to have only a minor effect on flavor but it can be manipulated to deliberately affect taste in a significant way.

One reason filtering is not regulated is because it is a subtractive process, taking flavor (and color, odor, etc.) out, but not adding any. However, in the complex mix that is a whiskey's flavor, the elimination of one flavor can highlight those that remain. Long thought to mainly diminish a whiskey's flavor overall, skilful filtration can alter a whiskey's taste (presumably for the better) more than you might think.

Skill in the art of filtration is important for anyone who deals in older whiskeys, which in the case of bourbon and rye is anything that has spent more than a decade in the barrel. Old bourbons and ryes have their charms but because they are aged in new barrels, wood flavors tend to overpower, beginning in about year ten. The older a whiskey gets the more tempting it is to tone it down a bit through filtration, but an unskilful job can ruin the whiskey with blandness. Rectify a whiskey too much and you get vodka.

There is a subset of bourbon and rye drinkers who subsist almost entirely on whiskeys in their teens and twenties. Unless a whiskey's maker states that it is not filtered, it probably is, and a denial of chill-filtering doesn't tell you what is being done to it at room temperature. Presumably, the drinkers of these whiskeys want them to be all that they can be and if they can be improved by a little filtration, why not? Or maybe that's not how they feel, but that's how it is.


Sam Komlenic said...

I'm aware first hand of Michter's satisfaction with their chosen method(s) of filtration and can't argue with the result.

Chuck, were you part of the panel? If so, what was the verdict? Which method was most preferred by the panel?

Chuck Cowdery said...

I was on the panel. I was not, however, privy to the methodology.

Anonymous said...

Richnimrod said;
Interesting as usual, Chuck.
I have personally been pleased more often than not by non-chill-filtered Bourbons. That is of course, my own taste preference. Many are put off by the 'frost-haze' that is instantly obvious when adding ice or refrigerating a Bourbon. I see it as a promise of extra flavor, and especially a 'thicker' mouth feel.
It would sure be enlightening to know the various methods of filtering, along with their expected resulting changes to the whiskey. Of course, without a great deal more info from all the distillers as to the methods used by them, we'd be no better off in our efforts to select by expected taste profile according to filtration. Still, it would be fun to know more about it.

Sam Komlenic said...

Do they plan to share the results with the panel, or was this purely for their own benefit?

I'm a bit confused as to the intent of the exercise, I think.

Chuck Cowdery said...

The 'result' is the current expression of Michter's 10-year-old rye.

Anonymous said...

Might there be a list of non-chill filtered bourbons? If so, could someone post it?

Kyle said...

It's only a single study (there may be others), but according to their results even "experienced" whisk(e)y connoisseurs are basically guessing as to whether a whiskey is filtered or not. Obviously filtration makes a difference on a chemical level, but at least according to this study it's not really perceptible to humans. Is it possible that the perceived differences in the Michter's experiment were only because those participating were expecting to find something different? It seems fairly common that when you're led to expect something, like the smell of "dark fruit" in a whiskey, you're probably going to find it thereafter. Just some interesting stuff to think about.

Sam Komlenic said...

Now I understand. I thought this was more open-ended.


Anonymous said...

I'm sorry, but not much in this particular blog post is correct. Whiskey filtration is in fact very much regulated,: (b) Extractions. The removal from any distilled spirits of any constituents to such an extent that the product does not possess the taste, aroma, and characteristics generally attributed to that class or type of distilled spirits alters the class and type thereof, and the product shall be appropriately redesignated. In addition, in the case of straight whisky the removal of more than 15 percent of the fixed acids, or volatile acids, or esters, or soluble solids, or higher alcohols, or more than 25 percent of the soluble color, shall be deemed to alter the class or type thereof.

The reasons for this are multi fold, but primarily it must be remembered that the distillation proof maximums that define the standards of identity are set specifically so as to allow "or require" that certain amounts of known congeners remain in any given product.

In other words you are not allowed to take a bourbon to 160 by distillation and then take out in excess of 15% of the remainder of non ethanol components through filtration "or otherwise", as it changes what the product may be called, as for all practical purposes it would be closer to light whiskey.

Conventional distillation is not the only way to alter and extract ethanol and congeners from a mash, it is just the common way. Many things can also be removed through processes of different filtration protocol and medium.

Of course the company referenced in the blog can alter flavors, that's what they do :) But it is clearly a regulated field of process .

Alex said...


The study doesn't seem to say filtration is imperceptible, it only says that those tasters couldn't reliably tell which was which. But it doesn't mean the two didn't taste different.

Here, we're not talking about identifying filtration but it's actual effects on flavor are where the consumer only tastes the final result.

Kyle said...

Eh, I think that's kind of splitting hairs. If you can't reliable tell which is which, I think it's fair to call that imperceptible. I agree, it doesn't mean that they don't taste different and it definitely doesn't mean they aren't different on a chemical level - obviously they are. But it does mean that, statistically, even "experienced" consumers aren't able to do better than guess as to whether the effects on flavor are present in the final result or not. Which, as they point out in their summary, basically means it's not much more than a piece of marketing they can put on the bottle to influence what you think you're tasting and your purchasing decisions. But again, it's just one study.

Alex said...


I disagree. An extreme example: If you had two bourbons identical except for filtration and one tasted strongly of strawberries and one tasted strongly of bananas, you might only be able to randomly guess which one is filtered because you don't know what the effect of the filtration is. Thus, statistically, no one can identify which is filtered. But that doesn't mean the filtration is imperceptible since the two taste very different from each other.

If someone outside of the industry is not familiar with the specific effects of different types of filtration on a whisky (and the whole point of Chuck's write-up is that there are different types of filtration with complex effects on the resulting product), I could see how it could be impossible to tell which one is the filtered sample between two samples. Add the extra complexity of different whiskies, and who knows if the smoothness you taste in one or the woodiness you taste in the other is the result of the underlying whiskey or of the filtration.

In other words, if you don't know which is the unfiltered sample, your preconceived notions about the effects of filtration on taste may simply be wrong. But that doesn't mean the filtration can't have dramatic effects, just that you don't know which is the before and which is the after sample.

We can agree to disagree, but for me the study doesn't address this point.

Kyle said...

I don't entirely disagree - it's certainly possible that filtration has a significant impact on the flavor/texture of a whiskey. What I question is whether that effect can actually be detected by a human or if people only taste something different because they've been told that a whiskey is filtered in some way and they're therefore expecting to taste something different. Suggestion is powerful (and sometimes deceptive).

Chuck Cowdery said...

Let me chime in here. Nobody is selling products based on their filtration. Nobody is saying 'we improved this through filtration.' So nobody is asking consumers to 'detect' filtration or 'taste something different.' I wrote about filtration because it is not usually in the discussion except for the handful of brands that boast 'not chill-filtered.' There, of course, you can't taste 'the difference' because there is nothing to compare them to.

Although filtration has gotten more sophisticated, it is still a blunt instrument. You can't fine tune a flavor profile as you might like, turning this up and that down like some kind of multi-channel mixer.

Dan said...

You need to filter whiskey to some extent. Whiskey straight from the barrel is nasty, unless it has been lying absolutely still with the bung uppermost for a good long while (days, weeks), waiting for someone to extract a sample.
There is a lot of char that ends up in the whiskey, but also environmental particulates and weird shit that precipitates out of a spirit over time. And they do negatively effect the taste.
Add to this that alcohol (or a high %abv mixture of alcohol and water) refracts light rather differently than plain water, making it "sparkle", but also highlighting every single bug-hair suspended in it.
So, yes, it is all filtered, to some degree, unless it is one of those wierd, murky and very limited releases.

Chuck Cowdery said...

Very true. Even products that advertise that they're "not chill filtered" are filtered at ambient temperature to remove solids. As for it being nasty, though, that's only true of the dregs. I've taken many barrel samples and haven't had a murky one yet.