Monday, July 27, 2015

Bruichladdich, Islay Barley, 2007 Rockside Farm

stats: single malt Scotch, Islay, 6 years old, 50%, $65

When a distillery finds itself under new ownership, it’s not uncommon for some changes to occur regarding the methods of production. Whether these changes are an attempt to improve consistency and quality or to increase output and boost profit margins, the character of the whisky is bound to be transformed by them.

While researching a couple of recent posts about Ardbeg, I learned that after it was taken over in 1997 by Glenmorangie a lot of long overdue maintenance and equipment upgrades were finally attended to. At the same time, having Dr. Bill Lumsden at the helm meant that there would be a much sharper focus on cask management going forward. But, I’ve also read (from the Ardbeg distillery profile on the Whisky-Online site, which I consider to be a reliable source) that more powerful yeast strains and shorter fermentation times have been employed during the latest era of Ardbeg ownership. In this example the positive and negative effects of these changes probably cancel each other out in terms of overall quality. Certainly though, the character of the whisky produced after 1997 is markedly different than that produced before.

While researching my piece on Bruichladdich’s 10 year old, I examined the direction that the distillery had moved in since coming under new ownership in 2012. Early in 2013, just six months after being bought by Remy Cointreau, it was announced that production was set to be doubled. The group that saved Bruichladdich in 2001 had been very proud of the quality of the distillate produced after they got the place up and running again, touting the long fermentations and slow distillations they employed. When a distillery is pushed to rapidly make a large increase of its output, corners are often cut to achieve the desired production numbers. I certainly had concerns about the changes that could have potentially been going on under the radar recently at Bruichladdich.

I did a little digging and came across an article announcing that they were going to start making twice as much whisky, where a distillery spokesperson claimed that they were determined to maintain the traditional methods of production at Bruichladdich. He went on to state that doubling output was possible without making any changes at the plant, aside from adding warehouses. At that time two new warehouses had been built in the previous two years, one was under construction, and a fourth was in the planning phase.

Being the skeptic that I am, I still had my doubts. I decided to dig deeper and take a look at some numbers to see if these claims would hold up to scrutiny.

The distillery had stated that it would be going from 750,000 LPA (liters of pure alcohol) per year to 1.5 million LPA and doing so with 24 hour production for five and a half days per week. Most of the other numbers I’m using are coming from the Misako Udo book The Scottish Whisky Distilleries, which was published in 2007. Bruichladdich has the following equipment: A single mash tun which takes 7 tons of grist and produces 36,000 liters of wort, 6 Oregon Pine washbacks that can each take a charge of 36,000 liters, two wash stills that each take a charge of 12,000 liters and two spirit stills that each take a charge of 7,100 liters. Fermentation times are stated as 60-67 hours during the week and 100-107 hours over the weekends.

I started off with the assumption that they were using a production cycle similar to that of Springbank which I was familiar with from their Whisky School. Springbank also has a single mash tun and six washbacks, and they were running five mashes per week. Bruichladdich’s spirit yield is 401 liters of alcohol per ton of grist. If you multiply that by 7 tons per mash and then multiply by 5 days per week and 52 weeks per year, you get roughly 730,000 liters per year. That’s close enough to 750,000 for me to assume that five mashes per week is the correct number.

Going over my notes from my time at Springbank, it looks a mash cycle should take four to five hours. Both distilleries run four batches of water through a mash (three is more typical of Scottish malt distilleries). Bruichladdich does produce more wort from a bigger mash (36,000 liters from 7 tons of malt vs. 21,000 liters from 3 tons of malt), but even if their mash cycle takes twice as long as Springbank’s, there is still plenty of time to run two mashes in a day without changing the process or the equipment.

With the given fermentation times, the mash done on Monday gets distilled on Thursday, Tuesday’s gets distilled on Friday, Wednesday’s on the following Monday, Thursday’s on Tuesday and Friday’s on Wednesday. Once his sequence has been established, there should never be more than three washbacks filled at any given time.

If the number of mashes doubles to two a day, then two washbacks are being filled each day. For the cycle to work out properly, you have to split the washbacks into two sets of three. A washback from one set gets filled in the first half of each production day, and a washback from the other set gets filled in the second half of each production day. Running like this, the above mentioned fermentation times are maintained and each of the six washbacks is getting filled six to twelve hours after it is emptied on an ongoing basis.

While there’s a little extra capacity here in theory (you have one washbacks that’s empty for six to twelve hours on each of the five production days), it’s probably not practical to put an extra mash through they system each week. I haven’t accounted for the time it takes to fill and then drain the washbacks, or the time it takes to sanitize them between fillings (this is done by pumping them full of steam). Adding an extra mash would have the staff constantly filling washbacks the moment they were ready. The best practice is to pitch the yeast as soon as possible after the wort is cooled to prevent any wild yeast or bacteria from establishing a foothold. That would mean the mashing cycle would have to be matched to when the washbacks became available, rather than a regular schedule of mashing at the same two times each day. I’m sure that would not be very practical.

On to the stills; the wash still takes 1/3 of what each washback holds (and the spirit still takes what the wash still puts out), so they would have been doing three distillation runs (each through a wash still / spirit still combo) each day that they were operating. The wash still run lasts five hours and the spirit still run lasts six hours and 40 minutes. A little bit of time is also needed to pump the spent lees and pot ale out of the stills after distillation is complete, but that shouldn’t add much ore than 20 minutes to the times above. With Bruichladdich’s two sets of stills, it is possible to do six distillation runs during a 24 hour period. There is a bit of lag time as the spirit stills can’t start running until the wash stills have done their first run. But once the spirit stills get started, they should both be able to push through three runs each for every 24 hours that they keep running. That lag at the start of the week’s distillations is probably part of the reason that Bruichladdich is running five and a half days a week rather than five.

I have to say that I was pleasantly surprised to find that my suspicions were unfounded in this case. Bruichladdich is now running at maximum capacity, at least in terms of washbacks. They could increase output further by running seven day a week; with two mashes and six distillation runs every day. If they did that without adding washbacks, that would mean fermentation times would be 60-67 hours for every single mash. As it stands now, for every two mashes that ferment 60-67 hours there are three that ferment 100-107 hours. That gives an average fermentation time of 87-88 hours. A reduction of 27% would be significant. If we hear of any future increases in production at Bruichladdich let’s hope the news is accompanied by an announcement of the distillery itself growing.

Okay, enough of the number crunching, let’s see how this relatively young whisky from Islay grown barley tastes.

The nose is quite fragrant, with floral notes and a slightly soapy quality (but not in a bad way). More subtle aromas are reminiscent of a salty, coastal breeze pushing across grassy dunes. On the palate it is somewhat full bodied and brings a nice range of flavors right up front. Gentle malt and stony minerality lead the way, with hints of fish nets and driftwood rounding things out. A youthful edge of green malt appears briefly as it transitions from the mid-palate to the finish. Warming spice notes take over at the end and are balanced by an oh-so-subtle touch of peat smoke.

In comparison, the 10 year has a more dense malt character with notes of gingerbread and American oak.

This Islay Barley bottle carries a 6 year age statement as well as the notation that it was distilled in 2007 and bottled in 2013. The bottling code, however, indicates that it was bottled in July of 2014. All of the lettering is printed directly onto the glass, rather than on a label. I suspect that they printed up too many bottles and didn’t fill them as quickly as they expected to. I would think it is actually 2007 distillate; why would they print more bottles than they had whisky to fill. So this should be between 6 ½ and 7 ½ years old, depending on when in 2007 it was distilled.

Friday, July 10, 2015

Ardbeg, Uigeadail vs. Corryvreckan

Uigeadail: single malt Scotch, Islay, 54.2%, $80
Corryvreckan: single malt Scotch, Islay, 57.1%, $75

I recently compared Ardbeg’s flagship 10 year old to their moderately peated, limited edition Blasda bottling, while at the same time going over the history of the distillery. Now I’m going to have a closer look at Uigeadail and Corryvreckan, the two Ardbeg bottlings that, along with the 10 year, currently make up the brand’s core range. But first a quick overview of how the range has evolved during the last 18 years.

As you’ll recall from the post linked above, when the mothballed Ardbeg distillery was bought and restarted by Glenmorangie in 1997 there were stocks of whisky from two distinct periods in the warehouses; the early 1970’s through March of 1981 and mid-1989 through mid-1996. The latter period was limited to two months of production per year. The new owners would also have the whisky they began producing themselves from mid 1997 onward as that spirit came of age. Production methods differed for each of these three periods giving three distinct styles of Ardbeg that would shape the evolution of the brand’s offerings for years to come.

The only bottlings put out by Ardbeg in 1997, 1998 and 1999 were their 17 year old and a series of vintage releases dated to the mid 1970’s. The 17 year is often said to be made up solely of distillate from 1980 and 1981, but I have seen a quote attributed to Glenmoranie’s Dr. Bill Lumsden stating that the 17 year also contained distillate laid down between 1975 and 1977.

The 17 year old and the 1970’s vintage bottlings continued on until 2004, but they were joined by a new 10 year old offering in 2000. This bottling was the first use of whisky from the period of limited production between 1989 and 1996.

Another annual release was started in 2001; Lord of the Isles was a vatting of whiskies from 1976 and 1977. It was part of the lineup until 2007 and much like the 17 year, its label stayed the same but the whisky grew older with each subsequent bottling.

The next significant addition to Ardbeg’s lineup was Uigeadail, which first appeared in 2003. It was described at younger bourbon barrel aged whisky vatted with much older sherry cask matured whisky.

Then there was a series of bottlings which tracked the progress of the whisky that the new owners began distilling in 1997. First was Very Young in 2004 which was followed by Still Young in 2006, Almost There in 2007 and finally Renaissance in 2008.

There were two very limited releases of lightly peated, cask strength Ardbeg Kildalton. The one in 2004 was distilled in 1980 and put into 700 ml bottles. The 2005 release was distilled in 1981 and only bottled in miniatures.

The next addition to Ardbeg’s standard lineup was called Airigh Nam Beist. It was bottled for three years, 2006, 2007 and 2008, but all of them were vintage dated to 1990. Many people viewed Airigh Nam Beist as a replacement for the iconic 17 year.

At some point in 2008 the flagship 10 year old was transitioned from distillate produced between 1989 and 1996 to distillate produced from 1997 onward. There was a change in the label design mid way through 2008 that is generally considered to indicate when the transition took place, but some people claim to have tasted the change in the flavor profile several months before the labels were modified.

Another lightly peated release called Blasda was bottled in 2008, 2009 and 2010. It was non-age stated, but said to be about 7 years old.

Corryvreckan was the next addition to the lineup, arriving in 2009. Upon its introduction it was touted as the replacement for Airigh Nam Beist. This bottling is aged in a combination of French oak and American oak ex-bourbon barrels. It is non-age stated but said to be in the 10 to 12 year range (making it all from post-1997 distillate).

There seems to be some conflicting information about the French oak aged portion of Corryvreckan. It was actually first seen as an Ardbeg Committee bottling in 2008 using first-fill French oak casks (either Burgundy or Bordeaux casks, I’ve seen mentions of both). Most reputable sources now state that Corryvreckan uses new French oak rather than first-fill French oak (along with the bourbon aged component). I’m not sure if this was a gradual transition over a few years or a sudden change when it became part of the regular lineup, but the bottle of Corryvreckan I have from 2009 seems to show little if any wine cask influence. While I was at the distillery in 2012 I was told that Corryvreckan was aged in toasted new French oak.

The limited releases have continued from Ardbeg as well. There was the more heavily peated (100+ ppm) Supernova in 2009 and 2010. Also released in 2010 was Rollercoaster; a vatting of the first ten years (1997-2006) of the new owners’ production. Next, in 2011, was Alligator; a vatting of ex-bourbon barrels and heavily charred, new American oak barrels. 2012 saw the release of Galileo, which was distilled in 1999 and aged in a combination of bourbon and Marsala casks.

Recent years have also seen wider releases of the annual festival bottlings from Ardbeg; Day (2012), Ardbog (2013), Auriverdes (2014) and Perpetuum (2015).

With all of these limited releases and changes to the core lineup, it can be pretty tough to keep track of what was bottled when at Ardbeg. And that has led to the Ardbeg Project. This privately run website attempts to catalog all official Ardbeg releases by their corresponding bottle codes and provide additional information when possible.

In the case of Uigeadail, the Ardbeg Project is particularly helpful. When it was first bottled in 2003, the sherry cask component of Uigeadail was distilled in the 1970’s and aged to about 25 years. I haven’t seen any information about the age of the bourbon barrel component of the early bottlings of Uigeadail, other than the generalization of it as being “young”. At that time though, most of the limited production from the 1989-1996 period was probably being used for the 10 year old, so it stands to reason that the bourbon barrel aged whisky in Uigeadail would have been distilled after the facility was restarted in mid 1997 and at about 6 years old.

Of course, with limited stocks of whisky from the 1970’s which were becoming increasingly more valuable as time marched on, it was inevitable that the recipe for Uigeadail would change. Perpetually on a quest to taste new whisky, it’s rare for me to purchase the same bottle twice. But I was so enamored by my early bottle of Uigeadail that I picked up another a few years later and would recommend it to anyone who asked. At the time all of this history was unknown to me, but the second bottle I had, while genuinely impressive, didn’t seem quite so magical as the first. I originally heard about the change in recipe from a tour guide while visiting the distillery in 2012.

More detail of Uigeadail’s changing formula can be found in this 2013 interview with Dr. Bill Lumsden (at the 23 minute mark), where he states “I’ve tried to gradually drift the recipe to a more appropriate age profile”. I’ve read on other blogs that the most highly regarded bottlings came from 2003 through 2009, and the most noticeable change happened across 2010, 2011 and 2012. I pulled the bottle from my shelf which had just enough liquid left in it for a few drinks, and checked the code against the information on the Ardbeg Project. Much to my surprise, it had been bottled on March 23, 2008. Next it was up to the attic to find my first (and long empty) bottle of Uigeadail. That one dated to April 20, 2005.

I’ve read that the sherry cask component accounts for 35% to 45% of Uigeadail, and that the percentage hasn’t really changed over the years. Having youthful, bourbon barrel aged whisky in the mix is part of what makes this bottling what it is, so I suspect that component has remained around the 6 year mark. I’m also speculating a bit on the sherry cask portion, but I feel like they must have started transitioning it away from late 70’s / early 80’s distillate pretty early on. Remember though,
during the 1989-1996 period production was limited and intended for blending, so there was probably little if any sherry cask whisky from that time available for Uigeadail. The biggest shift of the sherry cask component to 1997 and newer distillate likely took place across 2010, 2011 and 2012.

If I’m correct, over the course of 10 years the sherry matured component of Uigeadail has drifted down in age from roughly 25 years to around 15 years. Not only that, but it has also transitioned across three distinct periods of Ardbeg’s history, each with its own style of distillate.

I feel quite fortunate to have some 2008 Uigeadail left to taste today:
The nose is sharp and biting. It almost seems astringent at first but shows its true nature upon more cautious inspection; dense, chewy peat smoke aromas are intertwined with dry, nutty, oxidized sherry notes. The palate shows incredible depth and complexity. While the peat smoke is the most obvious element, there’s so much more going on along with it. There’s a gingerbread-like maltiness, mint and wide range of spice notes. The sherry fruit character is dark and moderately dry, with a hint of nuttiness. A touch of brine rounds out the flavor profile. The lengthy finish evolves without losing balance and maintains a good level of grip even as it fades.

Corryvreckan (bottled 2009):
There are some nice aromas on the nose, but a healthy dose of alcohol riding along with them. The peat smoke is somewhat light and floral in character and is accompanied by some subtle tree fruit and tropical fruit notes. There is less heat and aggressiveness on the palate than expected considering its nature on the nose. Notes of dry spice and leather come to the fore and add complexity to the smoke of driftwood burning on a beach. A bit of earthiness and a subtle stone fruit element come into play as well. The finish is long and warming, with a building spice element and lingering peat notes.

Comparing Uigeadail and Corryvreckan to the 10 year, its peat smoke stands out more on the palate. But that is, in my opinion, a matter of the other two having wider ranges of accompanying flavor elements. And while the Corryvreckan stands nicely on its own, it simply pales in comparison to the Uigeadail.