Category: Testing & Blind Tastings

This is a guest post written by a local bourbon enthusiast friend Kevin Wyss. It also draws on some earlier research done by Mike Jasinski.

(If you’re just here for the results and not a history and science lesson, scroll to the bottom)

The allure of “dusty” bourbon is seemingly universal in the current whiskey landscape, and while there’s no definition for what makes a true “dusty” bottle, many would argue that to qualify, the bottle must be at least 25 years old. Indeed this liquid from another century is held in such high regard that bottles which retailed for $10 in 1990 can now fetch well above $500 on the internet in backwater Facebook groups. Many purveyors of dusty bourbons argue that the flavors in these bottles simply cannot be matched by bourbon currently on the market, hence the hype and demand surrounding these bottles.

The history
In many cases, current production methods of bourbons have drastically changed in the new century. For example, a massive bourbon-fueled fire at the Heaven Hill distillery destroyed their entire production facility in 1996, sparking the transition to a more automated and modern distillation and barreling system. Now bottles of Pre-Fire Heaven Hill bourbon can fetch obscene amounts on the secondary black market. Whether this hype surrounding old Wild Turkey, Heaven Hill, Old Grand-Dad, Old Forester, Old Crow, Stitzel-Weller or Jim Beam dusties are worth the chase and cost is a discussion for a more experienced bourbon drinker, and not the topic of today’s article.

One of the cheapest and most common current forms of dusty bourbon is ceramic or porcelain decanters. These decanters were produced regularly starting sometime around the 70s, when the “bourbon glut” was just beginning. The “bourbon glut” resulted from a shift in Americans taste towards vodka, resulting in a drop in bourbon sales. Producers, which require 4-12 years to make their product, predicted that the drop in consumption would be temporary, and did not decrease production. Thus, a few years later, producers were sitting on thousands of gallons of well aged stock, with no demand for it. In a marketing ploy, many producers started releasing their bourbon not in standard glass bottles, but collectible and limited edition ceramic decanters. These decanters ranged from serious to downright goofy, taking the form of college mascots, animals, cars, guns, couch sized chess sets, spark plugs, wizards, clowns, states, and truly anything you can imagine. Importantly however, the bourbon inside these decanters were generally well aged stock, provided at rock bottom costs.

The science

To circumvent these shortcomings, I decided it would be best to use a machine called an ICP-MS: an inductively coupled plasma mass spectrometer. These machines, used by the EPA and FDA to study the content of heavy metals in water, food, food containers, and soil, are incredibly sensitive – able to detect less than 1 ppb of lead (or 0.0000001%) in water. Basically an ICP-MS works by injecting the liquid sample into a super-hot flame (17,500 °F). This basically strips everything down the bare atoms (ions) present in the sample. The different atoms (ions) such as lead, are then sorted by weight and measured with very high precision. One slight complication here is that the tannins, ethanol, and other organic compounds which make bourbon taste so good, will make soot upon injection, which can complicate the readings. So, to overcome this, I added an acid and peroxide digestion to break those compounds down. Then, I could centrifuge the samples to remove the tannins and flavor molecules that had settled out after digestion – with all the lead remaining in the solution. I then removed the ethanol using rotary evaporation, so all I was left with is a solution of acid and water, with all the lead still dissolved in the samples. I used an internal standard during the test to make sure that the testing of the samples was consistent and reproducible. To figure out the unknown concentration of lead in the samples, it was compared to a calibration curve – where you make solutions of a known lead concentration and extrapolate those results to the bourbon samples. The results of the bourbon samples will only be as accurate as the calibration curve is – the calibration curve I made had an accuracy of 0.996, so I stand by the results.

The results

So, using the ICP-MS I have access to, I ran some samples of dusty decanter bourbon to screen for lead content. Many, many thanks my fellow bourbon friends for providing the dusty samples – I know how valuable and sentimental they are to be simply given away for free.

I also found one other bourbon-loving chemist that had done a very similar study in 2013 – Mike Jasinski who under user name michaelturtle1 had posted data on the straightbourbon.com message boards. So, here is the data table combining all of the samples we tested, with the samples I tested highlighted. These show the amount of lead in one 50 mL pour. The amount of lead in one pour of bourbon is compared to the amount of lead in 1 gallon of bottled water, Houston tap water, and tap water from Flint, MI during the ongoing lead pipe crisis in the bottom of the right table.

The average pour of the 33 dusty decanters tested contains 4.9 μg of lead. However, the deviation from decanter to decanter is huge – ranging from 41.3 μg/pour to 0.3 μg/pour.

Modern Wild Turkey Rare Breed from a glass bottle was used as a control – showing a lead content of 0.03 μg/pour. Some notes on two of the samples: the Old Overholt 1915 was a glass bottle, not a decanter. This was included to see if significant amounts of lead were in distillers pipes, solder, stills, or tanks, which might carry over to the final product. The Ezra Brooks 1971 sample is not the original juice: I got that decanter as a gift, empty and have turned it in to an infinity bottle. I’m not sure how fast the leaching of lead happens, or how much lead was in the original bourbon, but from a sample size of 1, it seems like using old ceramic decanters as an infinity bottle may be a safe practice.

Here is a plot showing the lead content versus the year of the sample. Keep in mind that US lead laws changed a lot in the 20^th century – lead solder for pipes was used well into the 1970’s, lead paint was outlawed in 1978. Lead in cookware and ceramics was a problem well into the 1980’s.

The biggest outliers with the most lead, all came before 1974, but besides that I was really surprised to see that there wasn’t a strong correlation between the year the decanter was produced with lead content. Large differences between decanters produced in the same year and by the same company was also interesting to me, but may result from storage conditions or differences in the decanter production process. I was intrigued to see the massive difference in lead content between a glass bottle of modern bourbon (containing barely more lead than bottled water) and the 1915 glass bottle of Old Overholt BIB. Despite the bottled-in-bond distinction and lauded history of the distillery, significant amounts of lead were present in the whiskey. This speaks to the lack of laws and checks in place in the early 20^th century regarding lead, which I find interesting historically.

The takeaway

My main take home message here is that if consumed in small amounts, all decanters I tested are safe. Drinking 1 gallon of Houston tap water, you ingest ~15 μg of lead. A 50 mL pour from a decanter I tested will give you 0.3-41.3 μg of lead, at an average of 4.9 μg of lead. The CDC says to safely stay below 250 μg of lead ingested per day. Typical developed countries usually ingest less than 50 μg/day. So even drinking a pour of the highest lead containing decanters isn’t going to get you close to CDC daily limits, assuming you have a normal diet and life. Lead is purged from the body with a half-life of about 3 weeks. So, if someone did multiple decanter pours per day, every day for two weeks, you might encounter some issues. But even that I’d say is unlikely – so hunt on and enjoy those dusty decanters.

Today’s post is a drinking for science twofer.  First, I’ve added to my testing projects on “Do spirits change in the bottle once opened?”, but this time with rum.  Second, is a blind tasting of modern era Wild Turkey with bottles from 2010, 2015 and 2020.

I’ve tackled the question of “Does bourbon change once the bottle is opened?” 3 different times.  My conclusions are that it does not change enough to be perceivable when stored up to 1 year with a good seal, out of bright light, normal room temp, and a fill level at around 50%.  You can read about this info here – https://tater-talk.com/2018/08/30/did-my-bourbon-change-in-the-bottle-1-year-test/

Given my past conclusions, I have wondered if heavily peated scotch or higher ester rums would result the same.  Are these esters more volatile?  One way to find out is to test it. I purchased 3 bottles of Stolen Overproof Rum 375ml from the same store on the same day. On Jan 1, 2021, I poured one of these into a 750 ml bottle.  Over the next 12 months, I opened that bottle 28 times and swirled it around exposing it to fresh air.  The bottle was well sealed each time and stored in a dark cabinet at normal room temperature.  On Jan 20, 2022, I had 11 tasters over for a triangle blind tasting.  The other 2 bottles of Stolen were opened right prior to the test.  Each taster had 3 Glencairn glasses without knowing which glass had which sample.  The objective was to see if they could identify the odd sample.  I also gave them a choice of saying no difference, though none elected that option.

If the testers randomly guessed 33.3% would luckily identify the odd sample.  In my test 6 of the 11, 54.5%, were able to correctly identify the odd sample which is roughly 1.6 times the random guess odds.  My group of tasters are all folks who routinely drink spirits neat and would be considered above the average consumer in a test like this.  Still, 45.5% were not able to identify the odd sample.  Of the 6 that identified the odd sample, most thought it was the worst of the three, but all commented the differences were subtle. Unlike my bourbon tests, I think the results here indicate that some type of small changes likely occurred.  I think storing the bottle half full and opening it a total of 28 times is well above normal and makes for a good test.   I still think spirits are remarkably stable once opened if you follow best practices of storing bottles upright, tightly sealed, out of any bright lights, and in normal room temps.  If you get down to the last few pours in a bottle, either finish it off or, if you want to hold on to some, transfer to a 50ml glass sample bottle and properly label it.

For you stat nerds and other trolls, I’m well aware I didn’t have a N number of 15.  Take what you want from the data given, but no need to give your opinion until you have ran similar tests, which are easy enough to do, and published it.

For my fellow Wild Turkey 101 fans, I put together a blind tasting of modern era WT 101 with bottles from 2010, 2015 and 2020.  This was the same group of 11 tasters. I provided a score sheet and asked the tasters to rank in their order of 1st, 2nd and 3rd place.  I will say I learned my lesson and never again will I attempt a twofer test in one night, some of my tasters left indecipherable markings instead.  Perhaps vanity was the cause, as maybe they did not want to be perceived as having scored this wrong?  The scores of the 8 tasters that correctly followed instructions are as follows:

I’ve since had a chance to sit down with these 3 multiple times, although not blind.  Honestly, I go back and forth between favoring 2010 and 2015.  Both are better to me than the 2020.  This is not knocking current WT 101 as I still think it’s one of the best values currently in bourbon.

I recently came across a newspaper article about a company, Persedo Spirits, with a patented process that discussed polishing whiskey.  The article mentioned turning bad liquor into good. Over the years I have seen and tasted several other spirits produced by some type of contraption, they were poor substitutes for and easy to detect apart from properly aged spirits.  I would not have thought twice about this latest company except for the fact they were in my backyard in Alvin, TX which is about 35 miles south from me. 

They had convinced the newspaper reporter that their polished spirit “was considerably smoother with richer flavors”.   I wondered if they would entertain a local group of bourbon enthusiasts for a visit. I sent them an email and received a phone call back that day from the company President Ricky Ford.  We had a nice chat and I was honest about being skeptical, but I committed to approach his products with an open mind. Ricky understood and said he welcomed the challenge and would love to show us their facility. We picked a date and I arranged a group composed of myself and four other members from the Houston Bourbon Society to visit.

For me to be convinced that something positive was really happening I wanted to bring my own whiskey and watch it go through their process then conduct a triangle blind taste test with it.  Our group decided to use Evan Williams 100 proof Bottled in Bond for this test. It is a solid economical pour, but in my opinion is on the young side with some rough edges. Ricky told me their process required a minimum of 1.5 liters.  I purchased a 1.75 liter bottle and a 750 ml bottle of the Evan Williams and blended the two together to ensure homogeneity. Upon arriving at Persedo Spirits, we tested the proof and temperature of our test Bourbon to compare to the post process.

I knew they were located in an industrial warehouse space so I had a picture in my mind of what the place would look like.  I was wrong about that. About one quarter of the location is built out into a beautiful man cave environment with plenty of TV’s, seating, and a full bar.  Their crew welcomed us and the bar was fully set up with snacks and Glencairn tasting glasses. Ricky started by telling us the history of the process that his father had invented and how they progressed from that to their more complex patented process.  We took our Evan Williams and started the process on their small-scale device, which took about 30 minutes. He then showed us a much larger version that they use in production. Their business is modeled to handle having others send them spirits in totes for processing or they can license their equipment to be installed at their customers facility.  They are a federally registered DSP.

One thing they were very clear on is they respect traditional barrel aging.  They are not trying to replace it. They have a slogan of “Where Technology Respects Tradition”.  So, what exactly does happen in this process? You can visit their website for more detailed information, but I’ll give an overview here.  The spirit is placed under a mild vacuum while being subjected to ultrasound waves; it’s also subjected to the addition of food grade Nitrogen and CO2 gas.  The parameters can be adjusted based on the desired outcome. Other companies have used ultrasound before in an effort to simulate traditional aging, but the other elements Persedo claims help them be more selective about what changes.

While we were waiting for our sample whiskey to go through the process, our group tasted side by side a few other whiskeys they had previously polished.  They have a bar with a large selection of various spirits they had polished. For the first sample, I spotted Wild Turkey 101 on their bar. I drink enough WT 101 that I know the profile well.  I could easily tell the difference between the 2 samples. I preferred the before version, but that could be due to how I expected it to taste. We also tried samples of before and polished MGI 95/5 Rye in the form of Bulleit Rye, a local produced bourbon, and Weller SR.  For me, I’m not a fan of Bulleit Rye before or after. All five of us agreed that the local Bourbon was much better in the polished version.

By this time our Evan Williams was ready.  We proofed the after version and temperature corrected the proof. It had dropped 1 point, which in a 100-proof bourbon is also 1%.    My testing in all my drinking for science projects, where possible, is done in triangle blinds. Each taster is given 3 samples which might be 2 of the processed and 1 of the standard or vice versa.  Regardless 1 of the samples will be different, so the first step is to determine if the 1 different sample could be identified. The follow up question is, was the odd sample better or worse overall?

3 of the 5 tasters correctly identified the odd sample in round 1.  Of the 3 that correctly identified the odd sample, 2 preferred the unpolished sample and 1 commenting they liked the samples the same.

Ricky stated that the processed whiskey tends to taste better a few days after polishing.  Based on what I have learned about marrying whiskies I can accept that claim. Ricky had polished another large bottle of Evan Williams 100 proof 1 week prior, so we conducted round 2 with it.

2 of the 4 tasters correctly identified the odd sample in round 2.  Of the 2 that correctly identified the odd sample, 1 preferred the unpolished sample and 1 preferred the after polishing sample.

I know this was a small sample set, but 5 out of 9 times in the triangle blind the odd sample was correctly identified.  For me, this shows Presidio’s polishing process does change the whiskey. However, in the case of Evan Williams 100 proof, most of our group preferred the unpolished whiskey.  In the case of this Evan Williams, for me the difference was subtle but the polished whiskey developed a minerally note to it that makes the whiskey taste different and somewhat artificial.  I was able to identify it in the triangle blind. From the visit I did bring home the Evan Williams bottles in hopes to have more participate in a future triangle blind.

In tasting through the other samples provided, I tend to think their polishing process works better with younger spirits.  If you are a craft whiskey distillery that needs to sell young whiskey for cash flow, I’d suggest seeing for yourself what your whiskey tastes like before and after polishing.

Time for another drinking for science post.  This time it’s a blind tasting conducted within the Houston Bourbon Society, HBS, with 15 tasters of 18 straight whiskies.  This is the 3rd time I’ve done a blind tasting with this group. Here is a link that covers previous blind tastings – https://tater-talk.com/2018/01/23/blind-tasting-of-20-american-whiskies/.  This time when I recruited the tasters I specifically ask for volunteers that had not participated in the previous 2 blind tastings.  HBS has over 6,000 members and if you are a part of this group you’re likely a bourbon enthusiast so you know a little more than the average consumer.  Are these folks expert tasters? Perhaps or perhaps they represent and score whiskies like the average bourbon enthusiast. 

In the past blind tastings, I’ve used a 100 point scale system.  In this system most whiskies score in the 60-100 point range and so below 60 becomes meaningless.  This time I switched to a more simple 0-5 scale, with 5 points being the best whiskey possible and 0 points being the worst.  2.5 points would be the middle of the road. Tasters scored each whiskey in either whole or half points. Tasters turned in scores of 3 whiskies each week, so this tasting lasted 6 weeks.  I instructed the tasters to score each whiskey on its own, preferably tasted on different nights.

The theme of this tasting was straight whiskies; specifically Straight Bourbon or Straight Rye.  Whiskies that are finished in a secondary cask that previously held wine or another spirit are no longer straight whiskies and did not qualify for inclusion.  Straight Whiskies are the product of only 1 state and aged at least 2 years. I also tried to keep the proof range much narrower than in my previous blind tastings.  This time the lowest proof was 90 and the highest proof was 116.8.

The results:

Sorted by Median:

link to graphs in larger format – https://photos.app.goo.gl/UgurGZBgge83ErFt8

So should you draw some sort of line at the top and never try any whiskey below that line?  Keep in mind that for every one of the whiskies at least one of the 15 tasters scored it 3.5 points, meaning they really liked it.  You like what you like; it does not matter if this particular groups overall score is lower than you expected. For example, I particularly enjoy OF Rye which scored very poorly; I’m still going to buy another bottle, and soon as my open bottle is finished. 

Most of the whiskies selected are current versions purchased right before this tasting began.  The GB Balmorhea were 2 375ml bottles, which I have been told was the first batch. Barrell Bourbon HBS Mr Blue was a 14 year pick of TN Straight Bourbon done in 2018.  The OWA bottle was from my bunker; I included this as a thanks to the volunteer tasters for their time. They were a great group providing very good tasting notes and completing scores on time each week.  

My personal observations are current Wild Turkey Rare Breed is a great Straight Bourbon, easy to find and at a reasonable price.  Henry McKenna, at least the normal single barrel that us every day folks randomly buy off the shelf, is a solid bourbon, but it’s not the world’s best whiskey no matter what some asinine award show proclaims.