Cadenhead’s Benrinnes 18yo
Small Batch Collection | 55.7% ABV
Score: 8/10
Something special.
TL;DR
Weighty Benrinnes aged to near-perfection
Let’s talk about distillation
I’ve written previously about many facets of whisky production. From the influence of varying peat sources to copper-catalysed transesterification, nothing has been off limits. So why haven’t I discussed distillation - the basis for our beloved beverage?
It’s because it’s complicated. No seriously, really complicated. It’s also stunningly, jaw-droppingly and beautifully simple: ferment something to produce booze and congeners, heat it up, condense the vapour and voila… spirit. Use grains, you get whisky of some kind or another. What more could you want?
Well of course that’s all true, but trying to understand the how and why of it all is enough to drive one to drink - at least we’re in the right business, right? I’ll start by outlining some principles we need to grasp at least fundamentally before we can have any kind of serious chat on this topic. Much of this will be familiar to many of you, but for the sake of some semblance of completeness please indulge me. I’ll limit this article to simple pot stills for the time being, given there are many other potential configurations (hybrid alembics, column stills, pots with doubler configurations etc) which complicate things immensely even before considering their combinations and permutations.
Fractionation
Basically, this is what distillation is all about; different volatile compounds have various boiling points, so by gradually changing the temperature of a mixture we see different vapours leave the still at different points. Ethanol has a lower boiling point (roughly 78 degrees Celsius at sea level) than water, so by heating a still to less than the boiling point of water, we can extract a relatively high concentration of ethanol. Simple, right? Well yes and no, because of pesky physics/chemistry there are complications. More on this later.
Evaporation
First, thermodynamics dictate that liquid systems which experience change in temperature without pressurisation begin to evaporate before their boiling point. This is the reason sweat evaporates off our skin and leaves us feeling cooler. In a liquid with some given temperature, which we can think of as an average energy of the constituent molecules, some molecules with higher energies leave the system as vapour. When this happens, the average energy of the liquid system is reduced and so the temperature decreases. The rate that this occurs is obviously temperature dependent; at a chemical’s boiling temperature, all of that chemical will transition from the liquid phase to gas.
Intermolecular attractions
Because of intermolecular attractions (Hydrogen bonding and Van der Waals forces - read more here) many volatile compounds are difficult to fully separate from one another by simple pot distillation. Many varying alcohols observe this behaviour, including ethanol and methanol. There’s a long-standing myth, especially in home distilling, that it’s easy to separate ethanol and methanol by distillation, ie simply discard the foreshots where the more volatile methanol ought to appear due to its lower boiling point. The reality is that intermolecular attractions between the two make it frighteningly more difficult to achieve this than one might hope.
Luckily, cereal grains contain very little to no pectin, and pectin fermentation (hydrolysis by esterases) is the primary producer of methanol in alcoholic beverages such as wine. Anyway, we’re getting sidetracked; the point is mixtures of chemicals have different boiling points than their individual components. For instance, the boiling point of a 50% ethanol-water mixture is approximately 82 degrees Celsius.
It’s also worth mentioning that depending on concentrations of certain compounds, intermolecular attractions and repulsions can cause density distributions at liquid-gas surface interfaces due to enthalpy/entropy arguments (minimising enthalpy). As I referenced in my Laphroaig 10 Cask Strength review, guaiacol is a prime example of this; in low/no ethanol concentration environments (water, wine) guaiacol can contribute to orthonasal character, but in higher concentrations such as those found in whisky the degree of intermolecular attraction prevents significant evaporation, at least under standard conditions. This accumulation at the gas-liquid interface may also impact the tendency for other molecules to volatilise depending on their alterations to surface tension arguments.
Azeotropes
An azeotrope is a mixture of two compounds which, at some given concentration and pressure, have the same ratio of components both as a liquid and vapour. That is to say, there is the same concentration of each chemical in liquid and gas form. For instance, water and ethanol form an azeotrope at sea-level pressure at a little over 96% ABV. Azeotropes tend to be of much lower concern in whisky distilling than intermolecular attractions, however it’s worth noting to highlight how simple distillation is not an exact means of separating mixed components, particularly regarding pot still distillation.
That should do for now, let’s proceed with a horrendously compressed and oversimplified summary of distillation.
So, what specific bearings do the above principles have on whisky distillation? Well, it basically means all the compounds which can be extracted from wash come through from the still not discreetly as a function of temperature, but rather in a messily continuous fashion with sliding scale concentrations dependent on many interrelated factors. Modelling these mathematically requires an extremely convoluted series of partial differential equations with a massive number of initial conditions to consider; it has been done for a limited number of compounds (such as by Ikari & Kubo, 1975) but it is monstrously complicated and like all modelling only approximates experimental observations with so much accuracy, so we’ll go for a vastly more generalised investigation.
There are a huge number of factors to consider, but it all basically goes back to fractionation. The height of the still, specifically the vertical distance between the wash’s liquid-gas surface interface and the lyne arm, dictates a large aspect of fractionation. The circumference and rate of inward taper of the neck as a function of height also impacts fractionation. The higher the surface area to volume ratio of the copper surface (which conducts heat away from the interior gas environment to the cooler external atmosphere) the greater the volume of vapour which will condense near or on the copper surface and cause reflux.
This effect is exacerbated by stills with water-cooled jackets, such as those at Dalmore. Similarly reflux bulbs increase fractionation by principle of gases cooling as they’re allowed to expand; by forcing vapours through a bottleneck out into a greater volume space, the vapours cool and certain less volatile component gases will condense back into liquids.
More volatile compounds will remain a vapour for longer as they cool while travelling up toward and through the lyne arm, whereas less volatile components will condense back into a liquid and flow back into the still more readily. This is particularly true for distilleries with lyne arms with an upward angle from the neck to the condenser, the degree of such extending with increasing lyne arm length.
Purifiers compound this post-still fractionation with various designs by forcing the vapours through a high surface area system (i.e. jacketed piping or a plate) as a kind of intermediate partial condenser. Some are water cooled, such as those used at Glen Grant, Edradour, Strathmill and Glen Spey, while those at Ardbeg, Glenlossie, Tormore, Scapa and Talisker are not. The components of the vapour which are less volatile have a higher tendency to condense back into a liquid, and these are fed out from the bottom of the condenser directly back into the pot for redistillation.
These reflux variables naturally correlate with the degree of copper-vapour reactions which occur during distillation, particularly sulfur conversion/removal and transesterification in the spirit, as has been touched on in my Springbank 12 vertical and SMWS Bowmore reviews respectively.
So, first thing’s first - what is standard procedure for distilling whisky? Well in the majority of cases, a distillery will run the wash through their wash still in what’s varyingly known as a stripping run or blank run to extract as much ethanol as possible from the wash - as well as to collect as many of the volatile congeners as possible. During the stripping run there are generally no “cuts” (switches between receiving tanks) made - everything is collected as what is now the low wines. The second distillation is where the low wines are converted into new-make, and it’s generally only this second distillation where two cuts are made; one to go from the heads (highly volatile congeners mixed with the highest concentration of alcohol) to hearts (the desired components collected during the run) - and another to go from the hearts to the tails (the lower volatility compounds with a lower concentration of ethanol than the hearts or heads).
To the best of my knowledge, all well-established or traditional malt distilleries recycle the heads and tails (collectively known as feints) into subsequent distillations - more on this later in the article. So how do we measure the progress of a distillation and decide where these cuts should be made? There are a few ways to track it which can be used in parallel to one another. First, temperature.
The temperature of the still’s charge, the vapour temperature in the neck and/or lyne arm and the temperature of the liquid exiting the condenser all give information. The still’s charge temperature indicates the input energy driving volatilisation and indicates the proportion of given congeners which will favour entering the vapour phase. The vapour temperature in the neck/lyne arm indicates the distribution and energy level of congeners which have remained in the vapour phase to that point.
The two of these can be monitored somewhat using the chart below - the blue line is the bubble curve, which indicates the temperature at which a given concentration ethanol-water mixture will begin to produce vapour or “bubble”. Everything below the bubble curve indicates a single liquid phase of ethanol-water, while everything between the blue and red line indicates a two-phase state of liquid and vapour.
The red line is called the dew curve and represents the temperature at which a given concentration ethanol-water mixture will begin to condense or “dew” into a liquid. When a mixture of some given concentration is heated to a temperature which falls in the boundary between these two curves, the resulting liquid and gas phases will equilibrate to their respective points on the bubble and dew curves horizontally, ie. if a 25% low wine is heated to 89 degrees Celsius and the resultant mixture of liquid and vapour is allowed to come to equilibrium in a closed system, the resultant vapour component should have an ABV of roughly 67% ABV and the liquid component should be roughly 18%.
Of course the liquid and vapour don’t reach equilibrium since the vapour passes through the condenser and out of the system, cooling gradually on its way, thus the vapour temperature should always read lower than the liquid temperature. The disparity between the two is an indication of the degree of fractionation occurring in the still; the greater the difference, the more fractionation that is occurring. The cooler the vapour is (over 78 degrees of course) the higher its ABV. If the vapour temperature increases too quickly and the difference to the liquid temperature decreases then usually that’s an indication the still is being run too quickly and resultantly there is poor fractionation.
This is just one method by which these temperatures can be used to get an idea of what is going on throughout the course of a distillation. The temperature of the liquid exiting the condenser also gives an idea of how efficiently the condenser is being operated - more on this later.
The second major test distillers use for measuring distillation is time - once a plant has its SOPs (standard operating procedures) well established and the input variables are fairly stable, this is the most convenient shorthand for operating a distillation. A highly generalised exemplar model for the distribution of ethanol and congener classifications in time is given in the below graph (Panek & Boucher, 1989). In reality this time scale for each of the components can be condensed or dilated significantly depending on many factors, but primarily by the amount of energy added to the system via still heating.
The below table gives a good example of what time progression in early distillation looks like regarding some of the Islay distilleries, data taken circa 2004.
*We can probably extrapolate this to similar values as the others, although due to flow rates from the condenser being discontinuous we can’t do such with much accuracy.
** Variation in data due to multiple peating levels in distillate, thus requires different distillation.
The table above is also handy for seeing the relationship between time progression and strength of collected liquid in the heads from varying still operations. This also leads us into the most commonly used of the distillation measurements; ABV. The concentration of ethanol is a good indicator of which other chemicals are present at varying stages of the distillation. This is primarily because ABV is an indication of the temperature of vapour directly pre-condenser, thus approximately represents the distribution of congeners present at that point due to relative volatilities. It’s also partially due again to intermolecular forces, with varying ethanol concentrations favouring the transfer of different congeners depending on their molecular mass, functional groups and structures (ie hydrophobic, amphiphilic, hydrophilic etc).
At the end of the day, none of these measures individually gives the full picture - varying temperatures in different apparatus, the length of time the distillation has progressed and collected ABV when combined give a much fuller perspective of a given distillation.
Another progress check (although admittedly rather old-school) is the demisting test, which I’ve also touched on briefly in my Bowmore article. Essentially it relies on phase separation of certain congeners as a function of dilution with water; if a new make sample mists upon dilution, it still contains significant fusel oils and long chain fatty acids. Usually this test is used to help determine when the heads in a run have finished flushing out the tails in the lyne arm and condenser from a previous spirit run, but the principle can be used anywhere that a determination of these congeners’ concentrations is desired. It’s unlikely that many of the large and efficiently run distilleries use this test anymore - modern equipment and chemistry allows for much better analysis on the fly, and many are probably run strictly off time/temperature regiments established from deeply investigated data; the ability to send new make samples for GCMS analysis for instance.
There are of course many other factors to consider aside from “simple” fractionation considerations. The wash distillation for instance is usually conducted as quickly as is feasible - the primary objective being ethanol and broad spectrum congener acquisition for later sorting via the spirit distillation. Depending on how the wash still is heated (ie. the temperatures achieved for surfaces in contact with the liquid) various levels of furfural are produced at different rates from Maillard reactions - direct fired stills generate the most by far. Distillers must keep several factors under control during this time, the most important of which is foam over.
There is still some carbon dioxide dissolved in the wash when it’s transferred to the wash still - more for shorter fermentation times and/or wash from a clear wort, less for longer times and/or wash from a cloudy wort (fine particulates in trub act as nucleation sites). The wash has abundant grain and yeast derived short chain proteins and peptides which, combined with the residual dissolved C02, favour foaming. If left unchecked during wash distillation, this foaming can climb all the way up the neck of the still and flow over the lyne arm down into the condenser. This can also occur as a result of so-called “Chugging and puking”; if a condenser, particularly the shell and tube variety, experiences significant congestion due to high volumes of rapid vapour condensation in the narrow tubes, a cyclic pressure buildup/release can occur in the still. If the still walls have thinned with extended use between repairs/replacement, the pressure buildup can cause a visible expansion of the still’s body. When the pressure overcomes the vapour flow impedance and clears the tubes, the still contracts as pressure is relieved - allegedly the process appears similar to a panting dog’s chest. This is known as “chugging”. The chugging action with pressure shifts can cause more foaming in the still, and as the still depressurises can stimulate a foam-over action; the so-called “puking”. Regardless of the cause, it is of paramount importance to avoid foam over- the transfer of significant amounts of nonvolatile, unfractionated material through the condenser and into the receiver taints the entire low wines. Usually this results in the low wines needing re-distillation with a subsequent batch of wash and the wash still, lyne arm and condenser getting cleaned out.
A much more moderate and acceptable form of this non-volatile transfer action is misting- completely separate from, and not to be confused with the demisting test. Misting refers to the transfer of particles through the lyne arm and condenser not by fractionation arguments but by kinetic transfer; ever felt the spray from a bubble bursting near you? When surface tension is overcome by internal pressure, material can be sent by these bursting bubbles carried further by vapour streams up and over the neck/lyne arm and through the condenser. Ohtake et al established the transfer by this process to be small, about 0.027%-0.076% of the wash mass depending on process and design. The impact on organoleptics was noted to be important though, with the lower bound example’s new make characterised as light and fragrant and the higher bound as being full, yeasty and like cooked mash. The use of surfactants to decrease foaming- once upon a time simple soap, nowadays usually a simethicone defoamer- increases the degree of bubble formation and mist transfer. Increasing the vertical distance between the liquid-gas surface interface and the lyne arm decreases mist transfer.
So, having established a few of the individual parameters and considerations of a single distillation, we should consider the impact of multiple distillations. The composition of the low wines gathered from the wash still’s distillation impacts the starting point for the spirit distillation, both in terms of alcohol concentration and certain congeners. For instance, furfural and other Maillard products can’t be produced in the spirit distillation since there are no proteins and sugars to react along those Maillard pathways, so the level of furfural developed in the wash distillation determines a significant component of the new make character.
Also, it is usual in Scottish distilleries (among many others around the world) to recycle the feints from previous distillation runs of same recipe styles (ie peated feints are usually kept separate from unpeated feints, each type being recycled only in subsequent homogeneous low wines) by blending them either with subsequent washes before wash distillation or blending them with the low wines before the spirit distillation. Blending them with a subsequent wash means that the feints have to pass through two distillations before they might contribute some impact to the new make spirit (ie higher degree of fractionation/separation) whereas blending them with subsequent low wines means they only have to pass through one distillation before imparting some character to the new make, and hence there is less separation. This is probably what’s done for many Islay distilleries for instance, so that fewer phenols are lost in the pot ale waste stream.
If a distillery has been in operation for some time then usually the feints have been recycled through enough distillery regime repetitions that they reach a steady state where the concentration of congeners in the heads and tails remains approximately the same. Obviously small differences may occur depending on other input parameters such as from fermentation of wash or the temperature of the condensers- especially if they rely on cooling water drawn from a source exposed to the external environment where the time of year and resultant weather may have an impact.
On a minor tangent, some distilleries have a slightly less straightforward process for running stills, including their feints. Mortlach, for instance, has a mesmerizingly convoluted process which is described along with the other two Scotch distilleries utilising partial triple distillation in this article from Dave’s Whisky Reviews. Indeed the unique distillation practices of Benrinnes were the inspiration for today’s focus on the topic. On that note, shall we chat about the distillery?
Review
Cadenhead’s Benrinnes 18yo, Small Batch Release, 2000-2019, 55.7% ABV
AUD$200 (AUD$165 paid)
To begin exploring Benrinnes’ fairly unique profile, let’s discuss the potential merits of triple distillation. As with most of the rest of this article, it comes down to fractionation to control the composition of new make congeners. For instance, one may wish to split the feints into distinct fractions of heads and tails for separate blending. A distillery which desires a greater fruity ester component might blend the heads from one spirit run with the low wines from a subsequent run - this would increase the concentration of volatile short chain acetate esters in the next spirit run. The tails from the first spirit run might get blended with a subsequent wash prior to a low wines run to decrease the concentration of lower volatility compounds- long chain fusel oils and high molecular mass thiols for instance. A hypothetical exemplar distillery that might suit this operation would be along the lines of Glenmorangie or Glen Grant - light-bodied, fruity and pretty universally inoffensive.
The advantages of running split distillation regimes such as those at Benrinnes are far more speculative without supporting data, but basically comes down to the same concepts. By taking only the first of the two fractions from the wash and low wines stills into the spirit still, there ought to be a higher proportion of higher volatility compounds and a lower proportion of lower volatility compounds, as well as a higher ethanol concentration. The splitting of the tails into strong and weak fractions (the stronger portion being directly recycled into the spirit still with the next batch of low wines, the weaker portion being recycled to the low wines still for further rectification) might indicate a desire to avoid the least volatile compounds in the tails from accumulating in the spirit feints. This is somewhat juxtaposed by the new make strength which, according to scotchwhisky.com, sits at a mere 67%. This lower strength would usually indicate that the tails cut extends further than most (especially given the extra distillation steps) given that ABV decreases monotonically throughout a standard distillation’s progress. This may be a reflection of the new (post 2007) double distillation regime though, so again we must speculate.
According to the 2023 edition of the Malt Whisky Yearbook the distillery utilises fermentations between 65-100 hours, so certainly on the longer side. Combine this with the extra distillation processes and a propensity for collecting higher volatility components and we have a recipe for some good ester content. As a final word on the distillery, we should note the usually shocking lack of availability from Diageo who favour its use in blends. The distillery has been handled much the same as Mortlach before its critical acclaim brought about by independent bottlers, which caused Diageo to rebrand and market the living daylights out of it. Take from this what you will…
It’s worth noting that for all three distilleries in Dave’s article there is a common factor in the use of worm tub condensers. Admittedly Springbank only uses a worm tub for their wash still, the rest being modern shell and tube design, but the point stands. This means that despite the varying degrees of increased fractionation between the distilleries, there is a significantly reduced capacity for the removal and conversion of sulfur compounds during condensation. Again, as per my Springbank article, worm tubs have a significantly reduced copper-vapour interaction; this is due to a lower copper surface area average-enthalpy product per unit flux of vapour mass through the condenser. This means that there are fewer collisions between various vapour and copper molecules resulting in a reaction, which thus results in a higher retention of sulfurous compounds in the spirit. The practical means by which this is measured is the concentration of copper found in a given distillate. Distilleries- for example, Glenfiddich, as confirmed by their former head of production- have measured these copper concentrations as a function of the temperature at which the condensers operate; the hotter the condensers are run whilst still achieving conversion from vapour to liquid, the more dissolved copper and thus the greater the reduction of sulfur on organoleptics. Naturally this doesn’t apply to distilleries that aren’t using copper condensers in the first place, though those are very few and far between.
Right, I think that’s enough about distillation for one review - if there are any queries or topics not covered here that any of you would like addressed, please leave a comment under this article. I’ll do my best to further research any of them and compile another article addressing distillation outside of the relatively simple constructs that guide single malt scotch.
Nose
Without wanting to seem biassed by my own research or popular sentiment, there is indeed some distillate sulfur, fruity esters and a weighty fullness. Warm silage, rich malt, damp earth with a pleasingly sweet faecal dimension and slightly burnt hair- all very Campbeltown-like. Intermittently fleeting streaks of musk sticks with some restrained rummy tropical tones, particularly piques of mango, then waxy lemon peels. The oak is plentiful but not overbearing with the typical bourbon cask baking spices and demerara, then accents of tobacco and polished wooden antiques. Water gives a pleasing accentuation to the fruit and sulfur.
Palate
This is where the fruits really shine - we’re not quite in Clynelish territory but the oily texture certainly helps to point in that direction. Indeed more mango, but now also accompanied by pineapple and paw paw, thioesters geared to passionfruit and guava, a touch of meatiness, moderate menthol and more burly citruses. The oak has the decency to accent rather than dominate until the finish and retronasal where we see more tobacco, butter and coconut plus an almost Armagnac style old-oak spice. Dilution (no more than 5% ABV) brings extra texture, softens the oak and brings the citrus forward. I almost prefer it at cask strength, but each version has its merits depending on palate and mood.
The Dregs
Another classy Cadenhead’s bottling with a bit of something for everyone. I first tried this in a whisky club lineup against an SMWS 17YO Clynelish (it’s improving with headspace), an 8YO SMWS Clynelish and Glenmorangie’s Signet- stiff competition for most, but I must say this came out on top. I’m yet to try any post-2007 style Benrinnes, but quickly finding this older style distillate to be very much my niche. The shame of it is we will eventually run out of this profile at an affordable price as the years tick on. Even if Diageo had the sudden epiphany to restrict the number of these casks going to blends, their behaviour regarding other malt distilleries of enthusiast interest would indicate a punitive pricing geared to premiumisation. Grr. In the meantime, all I might suggest is keep the eyes peeled for similarly well presented indie bottlings at a reasonable price and keep the faith. Good whisky will find us- particularly when the international community helps guide us along the way.
Score: 8/10
Tried this? Share your thoughts in the comments below. TK
-
Dramface is free.
Its fierce independence and community-focused content is funded by that same community. We don’t do ads, sponsorships or paid-for content. If you like what we do you can support us by becoming a Dramface member for the price of a magazine.
However, if you’ve found a particular article valuable, you also have the option to make a direct donation to the writer, here: buy me a dram - you’d make their day. Thank you.
For more on Dramface and our funding read our about page here.
