Over two years ago, we set out to answer a question that does not seem to have a good answer: “What happens to bourbon when stored in different real world conditions and fill levels over time?”

We set up this experiment to try and answer the following basic questions:

1) At what point is there too much air relative to liquid in the bottle? In other words, at what point do you invite your friends over to finish it off, transition into smaller bottles, or apply an inert gas filler like Private Preserve?

2) What effect does temperature have on bourbon over time? Is room temperature best, or should you consider refrigeration for long term storage?

3) What effect does opening and closing a bottle (“drinking it down” and introducing oxygen repeatedly) have over time relative to the various storage conditions or simply keeping a bottle sealed?

4) What effect does sunlight have?

5) If your bottle is exposed to sunlight, is there any noticeable benefit to an amber glass versus a clear glass container?

6) At what specific time intervals do changes start to take place given a range of real world storage conditions?

To try and answer those questions and more, we tasted at 6 monthsand 12 months, making this our last and final tasting at 24 months.

How We Tasted:

• The samples were all moved from their storage locations to a common location 2 days prior to tasting in order to allow them to acclimate to room temperature.
  
  
• There were 3 tasters - Nick, Eric, and Jordan.
  
  
• We tasted blind sets of 3 samples at a time. We included 2 random leftover samples from the 12 month bourbon storage experiment tasting to round out the numbers so that 3 could be tasted per flight without any samples being tasted twice by the same person.
  
  
• We purposely varied the blind sets between the 3 tasters so that different direct comparisons could be made. Once complete, we cross-referenced our notes in order to piece together the findings so they could be presented here.
  
  
• We each poured for one another (while the taster was not looking), and aside from fill level at the time of pouring (which changed as we progressed through the flights) we did not know what we were pouring.
  
  
• The tasting was recorded by a video production company, currently working on a documentary about bourbon. We hope to see the documentary released sometime in the near future.

What We Noticed:

• With the exception of the drink down sample, the sunlight samples did not seem much worse than the 12 month samples did from memory. While I thought the sunlight samples might be a lot worse, they were still “drinkable” just like at 12 months. However these were quite bad, with terrible tasting notes including rubber, plastic, rotten apples, and paint thinner.
  
  
• The sunlight amber sample was not noted as tasting characteristically “bad” by any of us. Only Jordan noticed a very slight rubber taste in the finish. Eric liked it better than the refrigerator ⅔ and refrigerator ⅓ samples by direct comparison. Nick noted it as tasting a little off, generally speaking.
  
  
• The closet full fill sample, our control sample, was only selected by one of us as the favorite in its respective comparison flight. Nick preferred the refrigerator drink down and refrigerator full samples to it, Eric preferred it over the refrigerator drink down and sunlight drink down samples. Jordan preferred the closet drink down sample over the closet full sample (the other sample in his particular flight was a random 12 month sample, which he liked least).
  
  
• We each selected the dark closet ⅓ sample as the favorite within its respective flight. It was not directly compared with closet full fill sample in any flight, but was chosen over closet ⅔ and refrigerator full fill samples.
  
  
• With respect to the dark closet and refrigerator samples, the results were a little scattered, but generally so close it was difficult to pinpoint differences (with the exception of ⅓ fill noted above). In reviewing notes, preferences were more aligned to desired potency versus balance of flavors for the individual tasters within the individual flights.
  
  
• Sunlight ⅓ fill and sunlight drink down (now at ⅓ fill for the 24 month tasting) samples were consistently selected as the least favorite within their flights. This includes a direct comparison with the sunlight ⅔ fill sample.

‍

After over two years of researching, waiting, tasting, and documenting our findings, I’ll leave you with a few conclusions, some advice, and some things to consider when it comes to storing your bourbon.

‍Store your bourbon in a cool, dark place and most importantly out of direct sunlight. I’ve probably beaten this concept to death by this point, but the fact is sunlight, for whatever reason, did some some serious damage to the bourbon inside the sample bottles. Interestingly, the amber bottle stored in direct sunlight fared much better than the clear bottles, though amber bourbon bottles are few and far between in the real world. Additionally, while the refrigerator stored samples fared well, even over a 24 month period, they were only nominally different tasting than the room temperature stored samples at each of the sample intervals. They were also sealed with screw caps. Potential long term cork seal issues might exist as the low humidity environment inside a refrigerator could cause a cork to dry out more quickly.

‍Bourbon changes over time relative to the introduction of air within the container, but maybe not as much as we might have thought. Whether the bourbon gets better or worse will depend on the particular bourbon, the drinker, and the duration of storage, but this experiment proved it can change even if only a minimal extent. The changes from 6, to 12, to 24 month intervals suggest the interaction with air inside the container may not cause a steady degradation over time, but simply a metamorphosis of the bourbon’s flavor profile. I had hoped to analyze the samples on a mass spectrometer, but the stars didn’t align so our results are completely by taste. But this helped answer one of my most burning questions, which was if a ⅓ fill bottle meant it was time to finish it off, transfer to smaller containers, or consider an inert gas like Private Preserve. After running this experiment and taking notes on other real world long term and low fill level situations, the evidence suggests there is only a minimal impact on the bourbon itself and I’m personally much less worried about it. But would it be much different after 5 years? 10? That I can’t answer at this point.

‍Periodically opening bourbon and re-sealing over time is OK. The continual introduction of air causes excessive oxidation and ultimately ruins bourbon. The drink down samples were intended to represent a real world situation where a bottle was periodically opened, poured, and then sealed again. Based on our findings this yielded minimal change to the flavor of the bourbon over time, suggesting that it’s not something to worry too much about. Good news for many of those dusty bar pours.

‍Does the closure material matter? As far as closures go, all of our samples utilized conical seal screw cap closures. The intent was to ensure an airtight seal for the experiment, ultimately focusing in on the effects of the conditions and fill levels thereby eliminating any bias due to closure failure. The non-sunlight stored samples all fared pretty well. But not so with the sunlight samples. It is possible that the air within the direct sunlight samples heated up (the “hot car effect”), causing some kind of interaction between the air and the plastic material the conical seals are made out of. This might explain the rubber / plastic / paint thinner scent and flavors we all noticed with those samples, which became increasingly more prominent as fill level decreased. With that in mind, the samples stored in clear bottles and in direct sunlight also lost a great deal of color over the course of time, a change that would not correlate to the interaction of air with the conical plastic seal but had an effect on the bourbon nonetheless. Could the conical seals have been the reason for the flavor degradation as we noticed? If the bottles had been corked would the results have been different? Would the bourbon have smelled and tasted like cork?

‍The volume of liquid inside along with relative humidity of the outside air might affect the viability of a cork. A friend of mine liked the style of his bottle of Blanton’s so much that he reused it to store Macadamia nuts. Blanton’s has a natural cork, and it became quite loose over time to the extent the cork shrunk so much that it provided no seal at all. This got me thinking - what effect does the existence of liquid, or a specific volume of liquid, within a bottle have on a cork and its ability to expand enough to create an adequate seal? After all, wine is stored on its side to keep the cork from drying out but that’s not the case with bourbon.

What better way to test than with an informal experiment. I started with an empty bottle of 1792 Ridgemont Reserve placed on a shelf in my office, which gets indirect sunlight. Left untouched for about three months, I discovered the cork had shrunk to an extent that it no longer created a seal - I could pull it about halfway out freely. I filled it completely with water and a few weeks later it was back to normal, the cork expanded again due to the presence of water within the container. Finally, I emptied most of the water from the full bottle leaving only a few ounces. After about three months I observed the cork was still fairly tightly sealed.

I’ve also noticed my bottles that go untouched for long periods of time sometimes have notoriously hard to pull corks that feel as if they’re going to break off. Most likely because the corks are starting to dry out, my simple remedy for the situation has been to flip the bottle and wet the cork for a few seconds, then try again. Repeat as needed. Within a few attempts the cork is wet enough to move freely and I can pull it out without breaking it off inside the bottle.

Plenty of references to cork failure have been noted, and there have been reports of completely full unopened bottles experiencing loss of fill level due to evaporation over time. Old corks completely crumbling isn’t a tall tale either, we experienced it ourselves with a vintage bottle of Wild Turkey from the early 90’s.

Is the predominant culprit for bourbon that’s gone bad simply failed or tainted corks? And if so, are corks more likely to fail when fill level along with the relative humidity of the surrounding air decline?

Maybe a second bourbon storage experiment is in order...

***************************

Over these past few years I’ve been overwhelmed and appreciative of the tremendous positive response we’ve received as a result of this experiment, and thank everyone for waiting patiently as the results have unfolded.

But I would like to leave everyone with this: Our Bourbon Storage Experiment does not provide a final conclusion when it comes to bourbon and whiskey storage, and is not without its flaws. It is (hopefully) a very informative and interesting reference point, but it is representative of specific conditions, a specific set of taste preferences, a specific way of measuring change over time, and the transformation of a specific bourbon over time.

I hope this inspires others, and invite anyone to share their own experiences or experiments in the comments below or by dropping us a line.