Introduction

At Retribution we’re big believers that the future of modern day whisky lies both in the past and within the place that we are located, and this project stemmed from a combination of these ideas.

Craft and macro whisky distilleries are pretty much using one of three yeast strains; DCL M strain, SafSpirit M1 or Pinnacle MG+, they have many other names depending on who you buy your yeast from, but when you ask a few questions they always tend to lead to the same strains.

The original DCL M strain was introduced in around 1952 and is rumored to be a high attenuating brewing strain that may have come from the Nottingham area selected for its (at the time) high yield and flavour profile.

The SafSpirit M1 yeast strain is a modern-day version (introduced in 1960), that is used for its high (higher than M strain) alcohol yield and balanced flavour profile and is now widely used across the industry.

The origin of SafSpirit M1 yeast is only really known to Fermentis but is widely believed to originate from Mclays brewery and happened to be sample 1 hence the M-1 name.

Pinnacle MG+ is a later development and has been designed for high gravity, high yield and ability to work at higher temperatures, and produces a light and fruity new make and could be described as an industrial distiller’s yeast.

From the above timeline and descriptions, its fairly obvious the whisky industry has moved from using brewing yeast strains which resulted lower yields but higher levels of flavour to yeast that have prioritised yield over flavour. 

We talk about this trend a lot and its best represent by a simple graph – it’s a little extreme but makes the point well.

This trend is not confined to yeast, the same story is true for grain as the industry has moved from locally grown landrace varieties (lower yielding) to varieties that have been bred for yield and ease of processing.

One thing we have done at Retribution from the very first fermentation is to co-pitch brewing strains alongside distilling strains (predominantly M1/Ares WHC Labs version of M1), the idea being you get the flavour profile from the brewing strain and then the yield is boosted by the distilling strain. More recently we have been using 100% brewing strains which we have sourced from WHC Lab, these have been dry yeasts and the yeast strain we have had the best results from is Saturated which is an English Ale strain, which has been very popular of late in the craft brewing world for hazy IPA’s.

The next logical step from this was to move to wet yeasts sourced from breweries relatively local to where the distillery is based and also ideally historic family owned/independently owned breweries. These also had to be breweries that would be happy to talk to us and supply us with yeast. What we did not want to do was use yeast from a brewery that is using dried yeast, using cropped dried yeast from the point of view of this project would be pointless.

Luckily our founder is a member of the Chartered Institute of Brewers and Distillers (CIBD) this was pivotal in making this project come alive as it gave us the ability to talk to a good number of breweries in the south of England in the hope a few would agree to supply yeast for this project. This combined with a bit of networking at a few events meant we were able to source yeast from some amazing breweries, some local and a couple a little further afield. This resulted in the following breweries agreeing to supply us yeast for the project: Harvey’s, Fullers, Wadworth and Butcombe.

It is also worth noting that distilleries when using brewing yeast tend to refer to it as spent brewing yeast or brewery barm, this historically is yeast that is no longer fit to make beer with (possibly infected/low viability) or simply surplus to requirements. With modern brewing practices - yeast propagation/yeast health/use of nutrients and oxygen the viability of yeast is likely to be significantly higher than yeast from 100 years ago. As a result, spent brewing yeast no longer really exists as did 100 years ago and transport times between the brewery and distillery is now hours/maybe 24 hours is using a courier, and not days or weeks which it may have been for Island distilleries relying on ferries to transport yeast. Imagen a wooden cask sat on a quayside/dock in direct sunlight for a week getting up to 30degC yeast bubbling out the bunghole, drying out and turning black with mold – then being used for a fermentation!

If looking to produce a historically accurate brewing yeast driven fermentation the yeast would need to be abused for a period of time before use, ideally this would allow the yeasts to get infected with bacteria, viability of the yeast to be reduced – this yeast would then need to be revived at the distillery using what was known as a bub (a yeast starter). A bub is a mixture of a highly fermentable liquid made up of wort, grist and yeast the idea being to get a highly active yeast starter which is also laced with lactic acid bacteria which can then be pitched into a wash back – a project for the future!

Project Aim

Can flavour/aroma differences be observed in new make spirit when brewer’s yeast from different breweries is used. How does any flavour/aroma differences change over time once in cask – can yeast impact flavour after years of aging.

Method

Use our distillery standard mashing, fermentation and distillation protocols to produce yeast specific new make spirit, with maturation taking place in identical casks for a period of approximately 4 years.

Yeast

Liquid brewery yeast collected from brewery on the Monday, with mashing on Tuesday and Wednesday. Yeast was stored in 4 degC until used. Enough slurry was collected from each brewery to allow approx. 1 million cells of yeast to be pitched per ml per degree plato – this equates to a professional pitching rate for a high gravity ale.

Fullers yeast – taken from brewery propagator, cell count approx. 2 billion cells per ml, viability 95%+. 40kg of yeast slurry collected. A truly historic yeast that is used for 10-12 generations before being grown up from a fresh culture to maintain consistency.

Harvey’s yeast – top cropped from open fermenters, historic cropping method, cell count minimum 1billion cells per ml and viability 90%+. 55kg of yeast slurry collected. The yeast has continually been cropped and re pitched since 1954!

Wadworth yeast – cropped from the bottom of an 80HL DPV (dual purpose vessel) aka conical fermenter, cell count 1 billion cells per ml and viability 80%+ (I have a feeling this would be closer to 90% in reality). 55kg of yeast slurry collected. Head brewer noted the need to oxygenate the wort to get the yeast to fully ferment out.

Butcombe yeast – Cropped from the bottom of a conical fermenter, cell count 0.5 billion cells per ml and viability of 90%+ (0.5 billion cells was probably on the low side as the slurry was a lovely thick  cream probably closer to 1 billion cells per ml used 0.75 billion for pitching rate calculation). 75kg of yeast slurry collected, brewer commented on ideally the need for oxygenation.

Mashing

We source all our malted barley from Warminster Maltings and for this project we decided to use floor malted Maris Otter. We decided on two mashes/fermentations for each yeast strain with 500kg of malt being uses for each mash. A total of 4000kg of malted barley being used for the project. All grain was milled on site using our own mill.

We aimed for a mash temp for 64 degC and using a liquor to grain ratio of 2.8L/1kg  which fills our mash tun to the brim (if we had a bigger door to the distillery, we would have a bigger mash tun!). This equates to mashing in 500kg of malt, 1400L of water at 73 degC, all 8 mashes hit the 64 degC temperature to within 0.5 degC. All mashes had a duration of 90, following the mash rest we start the run-off and start sparging with 85 degC water total sparge volume being 900L with the aim of collecting 1800L in the fermenter

On the way to the fermenter the wort was cooled to 19-20degC, which is our standard yeast pitching temperature.

Fermentation

Fermentation was where things got interesting but more on this later in the results section. As mentioned above the yeasts were all pitched at the same temperature 19-20degC, we made efforts to ensure the pitching rates were similar/in the same ballpark. All fermentations were allowed to free rise up to 28degC before attemperation to maintain this maximum temperature, once the fermentation slowed and the temperature started to fall the fermenter cooling was turned off and heating turned on to maintain the fermenter at 28degC for the full planned 7 day fermentation period.

Gravity measurements are taken daily and recorded to track fermentation progress once FG has been reached we calculate and record the liters pure alcohol (LPA) in the fermentation, this then give us our LPA per ton of malt figure.

Distillation

Distillation is broken into 2 parts, wash and spirit runs.

Each fermentation requires one distillation, so 2 wash runs per yeast strain; in this distillation we are simply looking to get all the alcohol out of the wash. The wash still is filled the day prior to the distillation run, it is preheated overnight to 65degC using electrical elements, in the morning the elements are turned off and the gas burns turned on. Low wines start running off the still when the pot contents hits 92 degC and we run the burners till the pot contents is 99.9-100degC. The Low wines are collected in a 1100L tank, volume and abv are measured, recorded and again we look at the LPA figure to track alcohol losses and efficiencies of process.

For this project the low wines from the 2 wash runs were loaded into the spirit still in one go, the spirit still is heated rapidly to 80degC, the burners are then reduced to approx 30% output for the start of heads collection at approx. 83-84 degC. The heads are collected into our heads and tails tank. At 74-76% abv we typically switch to hearts collection, for this project we have decided to go at the lower end of this range to ensure clean spirit is collected and continue with hearts collection down to 64% abv before swapping back to tails collection. The hearts volume and density/abv are then measured and recorded along with the heads/tails volume and density/abv. Again, we track LPA’s and losses.

Dilution and racking to cask

We typically fill bourbon casks at 65% abv, new make typically comes of the still with an average abv of 71-72%, we then dilute it with RO water to the desired cask fill abv.

For this project we have sourced ex-Woodford Reserve casks, that were part of the same batch and from the same production cooperage, this means the casks are as identical as we could find.

Casks will all be stored in the same warehouse next to each other.  

Yearly samples will be taken to track the liquids process

Assessing the new make spirits

This is where you come in, with the new make samples you will find a flavour card, it would be great if you can fill this card in, and the take a photo of your results and email the images of the card back to us to compile the tasting notes/observations.

We have also arranged to get the new makes analysed via 2D Mass Spectrometry at the University of Southampton.

We will look to compile tasting notes and volatile data from the MS analysis to pull out key flavours components and look to see if the tasting notes match the analysis.

A limited number of samples will be drawn at the 2 year point and we expect to empty the casks at around the 4-5 year mark.

Results

Mashing

All mashes progressed as planned with typical levels of extraction from the malted barley, below is a table showing Volume transferred to fermenter, OG and mash efficiency.

Fermentation

Below is a plot of fermentation progression showing day vers gravity, all fermentations started at an original gravity (OG) of 1.070, as fresh brewery yeast has been used there is no visible lag phase, its straight into the primary fermentation phase, exponential yeast growth and rapid decrease in gravity before moving into the secondary phase once all available nutrients and fermentable sugars have been used up.

From looking at the plot several things are very noticeable:

- As you would expect the fermentations using the same brewery yeasts are paired together

- All fermentations are complete pretty much within 3 days

- There is a large spread of final gravities (FG) from 0.998 to 1.019

The high final FG’s of the Wadworth and Butcombe yeast has an impact on yield, reducing yield by 60LPA and 80LPA per ton respectively when compared to the Fuller's fermentation, this is shown in the table below, its also shown by the apparent attenuation figures for each fermentation yeast.

Distillation

Wash distillation runs were completed as expected, the table below show the LPA recovered per distillation and combined per brewery yeast.

It’s interesting to note that for a couple of the distillation runs we distilled in the wash runs more alcohol than was calculated from the fermentation data – I take this to mean that we got pretty much everything out of the wash!

A spirt run followed each of the two wash runs ensuring the low wines were kept separate and not mixed between different yeast strains. Below is a table showing the recovered alcohol (hearts and heads/tails) from the spirit runs.

As expected, the largest volume of hearts was collected from the Fuller’s spirt run, this is pretty much as expected, as it was the largest volume of low wines by a considerable distance. The volume of hearts collected then decreased as still low wines charged decreased.

However, even though the two Fuller’s fermentations yielded 340LPA, which for 1 ton of malt is very good for a distillery of Retributions size and on par with what we would expect to achieve using a blend of brewing and distilling yeast the final hearts volume was smaller than expected, we would expect on that size of spirit still charge to get closer to 180LPA of hearts not the 148LPA we actually achieved. However, the 32LPA that went to the heads/tails tank is not lost, it would be recycled round. So yes fermented alcohol yield was high but due to its high flavour level a lower yield was seen in the final hearts cut.

This lower heart volume can be seen across all yeasts, i attribute it to a more flavoursome alcohol being created resulting in a larger loss to heads/tails.

Dilution and racking to cask

The new makes were diluted to 65% abv and fill into cask in December 2025, full details are given in the table below.

An amount of new make was kept back for the project/tasting samples, maybe even a competition or 2.

Assessing the new make spirit

This part of the project is still to come, and we need you to assess the new make spirits produced – a 2^nd blog post will be written around march time.

Conclusion

Please note that the vagueness about the flavour of the new make spirits is on purpose as we do not want influence your thoughts on the samples.

From a production process point of view this project progressed without issue.

All four of the yeasts collected from the breweries fermented well although the FG’s did vary, but this variation was to be expected.

From simply drinking Butcombe and Wadworth beers I expected the FG’s to be a little higher due to the residual sweetness you tend to find in the beer especially when compared to Fuller’s beers.  Both their head brewers (Butcombe and Wadworth) did also comment on the need for oxygenation of the wort prior to pitching the yeast which is something we are not set up for at the distillery.

Oxygenation of the wort helps the yeast in the growth phase, and if a yeast strain is oxygen level sensitive it can result in a sluggish fermentation and lower attenuation, we did not see sluggish fermentations but the FG’s were a little high. However, this could prove very interesting in the new makes as residual sugars left in the wash will mean increased mallard reactions and caramelization during the wash distillation.

The Harvey’s yeast performed very well, their head brewer was not convinced it would ferment out that well, but that was not an issue.

The Fuller’s yeast ferment unbelievably well, on par with distilling yeasts we have used! But as mentioned did result in smaller hearts cut, this we believe is due to a more flavorsome alcohol being produced resulting in more alcohol going to heads and tails.

The next stage is to collate tasting notes and MS analysis to assess how the flavor changes between new make samples and we cant wait to get your feed back on the different samples.

Part 2 to follow on a couple of months!

Thank you

Rich

Head Cleaner - Retribution Distilling