I’ve had the benefit now of discussing with two experienced distillers my recent posts on 19th century American whiskey fermentation achieved solely with backset (no addition of yeast).
One works for an American craft distillery, the other is a retired distiller who had worked in a number of large distilleries internationally including Canada.
First, just to resume a bit: the various sources I quoted in these posts make it clear a fermentation was achieved in three to five days using just backset to scald (heat) the mash in small tubs.
This was denominated a sour mash fermentation. Today, sour mash means using backset in the mashing liquid, often one part backset to three of water, but fresh yeast is always added whether jugged or a distiller’s dried yeast. In the 1800s, sour mash as developed by bourbon legends James Crow and E.H. Taylor involved no addition of yeast to the mash.
To be sure, there were variant ways to sour mash in the 1800s. Some descriptions call for yeast to be added, either fresh or collected from a prior ferment. But it is very notable that the Crow-Taylor sour mash whisky used no added yeast. This small tub method, also validated by C.K. Gallagher’s account of small tub production which I referred to earlier, was a low-yielding method, requiring longer aging than sweet mash whisky but prized for its fine flavour.
Modern bourbon writing has not referenced this aspect of a wild fermentation except for Gerald Carson’s 1963 The Social History Of Bourbon. Even then the reference was lapidary, almost in code unless one understood the background. At Page 88:
His beer was a creamy liquid, rich in yeasting power. His fermentation was faultless.
Hence my interest to point it out and parse the details.
A question that arises when drilling down is, how did the backset actually do its work to ferment the mash, particularly as boiled spent beer can be presumed to be sterile?
In my discussions with the distilling experts, it became clear to me that in all likelihood, backset alone did not do the job. While it may have contained wild yeast in process of fermenting residual sugars in the spent beer, other sources of yeast – ambient air, wood vessels – probably assisted a natural fermentation. In addition, backset contributes certain bacteria that work with wild yeast to form alcohol. I wasn’t aware earlier that some bacteria can make ethanol. I knew of course they can produce lactic or acetic acid.
But some forms of bacteria are capable of converting simple sugars into ethanol. Zymomonas mobilis is one. (Some of these bacteria have been engineered in fact to boost ethanol production for industrial purposes). Dunder was a source of such bacteria, one distiller told me, and possibly the Crow-Taylor backset was too. Some accounts state that bananas and other things were thrown in the dunder to create sugars that would be fermented along with the converted grain starches. Since there is no evidence starch or sugar was added to cooled backset, the question arises if backset had any function as a yeasting agent or inoculum. Even if it didn’t, its role is still important as a source of the bacteria mentioned and to acidify the yeast.
But I think likely Crow’s and Taylor’s spent beer was in active fermentation when added to the mash. This would explain Carson’s use of “creamy” to describe it and its potent fermentative power he noted. Even dunder with yeast added can take eight weeks to ferment. If, as we know, the Crown-Taylor sour mash took three to five days,to my mind that could not have been achieved merely with atmospheric flora: the backset itself was probably an inoculum, as Carson’s account suggests.
All spent beer, even where you try to get every gram of sugar converted to alcohol, contains some residual sugar. One study I found shows that a mash for ethanol production – feedstock not clear, I think sugar beet – produced about 1.5 grams simple sugars per litre of spent beer. One can imagine the sugar content was much higher in Crow and Taylor’s spent beer since its fermentation was known to be under-attenuated, leaving in other words a rather sweet wort. Taylor specifically stated in the Bonfort’s article I cited earlier that his mashes contained considerable unfermented sugar. Needless to say, that sugar stayed in the mash after the boil in the still – it was a solid that couldn’t vaporize.
Let’s now think about why the Crow-Taylor method employed a series of small tubs. I think the reason may have been to maximize the wood-to-mash contact. The more wood in which lurked microbiota of the type we are discussing, the more likely a successful fermentation would occur. With a sweet mash for which only one large tub was used, it didn’t matter what wild yeast contribution there was since ample purpose-added yeast would ensure a good ferment.
Therefore, the main point I was concerned to show stands – Crow’s and Taylor’s classic sour mash bourbon used no added yeast in the process, which is significant unto itself. But it is also important to state that in all likelihood, the backset alone, and the same applies for dunder in heavy rum production, did not account for the fermentation. Rather, their acids and bacteria worked hand-in-glove with wild yeasts resident in the environment of these non-sterile distilleries.
One of the distillers told me yeast on malted barley in the mash might have been part of this yeast cocktail. He noted that unlike for beverage beer, the malt in a whiskey mash is not heated to a high temperature (vs. the corn and rye which need to be cooked)* and its resident wild yeast would be alive when the fermentation started. I pointed out that barley malt in a bourbon mash is today not more than approximately 5% of the mash. Even if it was more in the 1800s, and it was, generally, that source alone surely would not account for a vigorous – three to five day – wild fermentation. But it may have been an element in the total picture.
My conclusions are, the cooled backset:
i) probably contained fermentable sugar due to the low attenuation of the ferments;
ii) may have absorbed wild yeast from the environment which started to feed on that sugar and the dead yeast cells**;
iii) worked in tandem with natural yeast influence from wood vessels, ambient air, and possibly the barley malt in the mash; and
iv) provided some bacterial action to make alcohol,
which in sum achieved the fermentation in question, one which occurred naturally without addition of yeast by the distiller.
Note re image: the image shown above is from this fine early American antique bottle collection site. Image is believed available for educational and historical purposes. All trademarks shown belong to their owners or duly authorized licensees. All feedback welcomed.
*I want to clarify that the distiller I spoke to actually said, the corn must be cooked, the rye may be, and the malt must not be.
** A distiller reading these notes commented that if, as the historical literature states, hot slop was used to scald the mash, the heating of the slop to boiling temperatures would kill any yeast activity that had developed, therefore can we view the backset, assuming it had undergone fermentative activity, as an inoculum? Good point. My response would be, first, in the 24-48 hour or more cooling down phase (see C.K. Gallagher and other sources I’ve cited), the temperature would drop enough to permit inoculation by ambient wild yeasts which would feed on the nutrients in the slop. Second, even some modern sources, see e.g., Waymack and Harris’s The Book of Classic American Whiskeys, state that backset is sometimes added to the fermenter. It could not be added hot for this purpose, readers will appreciate. (E.H. Taylor did not do this apparently, but didnt’t need to, one could infer). Also, Gallagher seems to suggest slop was also used after the tubs had started setting to cool them down. Again this must have been stored or at any rate cooled slop. Therefore the point is taken but in light of numerous statements in 1800s literature that backset takes the place of yeast in sour mash distilling, and given the various times in the process it could be added, one of these explanations must I believe explain why backset was viewed, by Taylor and numerous others, as having fermentative power.