Follow along: IG @birectifier
I had collected this paper many years ago because it was part of the curriculum at the University of the West Indies in Jamaica. A blog reader was looking for it so I decided it was not highlighted prominently enough.
Its Manufacture And Quality
D.H. West – Sugar Technology Research Unit, Barbados
Rhett Harris – West India Rum Refinery, Barbados
I also just learned that it had become a chapter in a book a year later. That (identical) version is available if you are a member of the Sci-Hub Book Club… I would dive into the original which is requisite reading for any industry professional, but also feel free to consider my notes upon re-reading:
Something to keep in mind is that a major reason this document exists is because producers faced a collective problem as their molasses was changing. They were facing adverse changes in their rums that needed resources beyond a single firm’s ability to solve. Hence we have the teaming up of two great scientists trying to figure out exactly why amino nitrogren was rising and what to do about it because of the consequent increase of n-propyl alcohol, a higher alcohol which is part of the fusel oil.
The industry still faces both problems and now opportunities beyond any one firm’s ability to solve or capitalize on and would benefit from collective research projects creating a high tide that lifts all boats.
- Clutton, D.W. (1974) Flavour Industry 5, 286.
- Campbell, P.F. (1985) The Story of Barbados Rum. Barbados Rum Book.
- Private Communication West India Rum and Spirit Producers Association.
- Draft Barbadian and Jamaican Standards on Rum.
- I’Anson, J.A.P. Rum Manufacture, Process Biochemistry, July 1971.
- Arroyo Rafael, Manufacture of Rum, Sugar, December 1941.
- Kampen, W.H., Sugar y Azucar, July 1975.
- Arroyo Rafael, Manufacture of Rum, Sugar, January 1942.
- Parfait, A. & Jouret, C. Formation of higher alcohols in rum.
- Unger, E.D. & Coffey, T.R. Production of light bodied rum by an extraction distillation process.
- International Sugar Journal, July 1934, page 272
- Geerlings, Cane Sugar 2nd Edition page 131.
- Hance, F.E., Proceedings Hawaiian Sugar Planters Association Report 37 (1932) page 27.
One thing to note is that the average pH of Caribbean molasses is 5.5. Sourcing small scale amounts of molasses in the U.S. has been a challenge and often I’ve had to treat my molasses with lime to hit 5.5 or 5.8. The authors immediately references Arroyo on molasses quality.
The costing data on page 8 makes it seem like the authors would consider using fresh sugar cane juice if it was viable. Viability appears to be more important than tradition in rum making. A G.I. for rum becomes very different than a G.I. for a whiskey producer or fruit spirit because molasses is a byproduct and its fate is tied to another market and producers even face major changes to their substrate beyond their control such as the consequences of upgrades to refining technology. Creating standards of identity on shakier ground than anyone else is very hard.
I have never tried the alumina and calcium phosphate method for molasses treatment described on page 9. The authors are wary of both capital costs and sugar losses for molasses treatments.
In general it has been found that molasses pre-treatment promises a lot more than it actually delivers and by and large the major producers have avoided this procedure.
I’ve heard a lot of complaints about current molasses quality and I wonder if this will ever be reconsidered due to new circumstances. There is evidence finer aroma can come from clarified molasses and evidence it may be a requisite for Arroyo’s high pH heavy rum process. Molasses clarification was pushed by the Rum Pilot Plant well after Arroyo and I’m aware that even Caroni was setup for clarification but abandoned it because they could not maintain the equipment after budget cuts.
I was not aware of the added scaling inhibitors or if they still have a place in contemporary practice. I have heard of some places still interrupting continuous still operation to chisel out encrustations.
We see the authors awareness of fission yeasts on page 10:
On the other hand the fission type schizosaccharomyces strains are more suitable for the production of the heavier types of rum where the primary objective is congener formation rather than alcohol yield.
The authors reference a 1940’s document from Arroyo’s Manufacture of Rum five part series published in the Journal, Sugar.
As a result of the international demand for lighter rum, the budding type yeasts predominate fermentation flora in Caribbean distilleries, but fission type yeasts are still used in the production of the highly flavoured heavy rums of Jamaica, Guyana, and the French islands.
You wonder what specifically the authors know and have personally observed. For Jamaica, the only two distilleries are likely Hampden and Long Pond. For Guyana it is DDL who has not presently acknowledged the use of fission yeasts. For French islands, it would likely only be Galion in Martinique and not necessarily anyone in Reunion. However, the situation is rapidly changing and Foursquare among others are getting into the game as they pursue heavy rum. I’m excited for all of them!
The authors mention growing their own yeasts and you wonder how prevalent it still was in 1987 before the success of Lallemand changed what the industry did in house.
One page 11, we see this extremely cool note:
For a slower fermentation the inoculant is added after the fermenter has been filled with wash. Slow fermentations are used for heavy rums.
We know that Arroyo championed the technique of inoculating while the vats were being filled to make rapid fermentation start faster (these are incredibly large vats that take a while to fill). Here we see them stylistically doing the opposite! No doubt they also seed at a different rate.
Advantageous or spontaneous fermentation is still practiced in a few Caribbean rum distilleries. The fermenters or mixing tanks are simply filled and left to ferment relying on airborne yeasts, naturally occurring yeasts in the molasses or yeasts from previous fermentations attached to the fermenter walls. These fermentations can take from two to four days.
One thing that I don’t think is realized here is that these spontaneous ferments rarely achieve ethanol yields that are over 5%. I’ve been told if you achieve 5%, you are doing good and that typically takes more than four days. I have trouble with their numbers. Double retorts which are essentially triple distillation are needed when the wash is so low ABV. You need that third multiplier.
A peculiarity to Jamaica and the French Islands is the practice of using “dunder” in the production of their high ester heavy rums (5,7). “Dunder” is the lees of previous distillations which have been allowed to age and ripen by bacterial action. The matured dunder is added to the molasses wash prior to fermentation. Symbiotic fermentation by the yeast and bacteria produce the percursors for the formation of high ester concentrations in subsequent distillation in pot stills. This symbiotic fermentation has been shown by Arroyo to reduce the time required to complete the fermentation for heavy rums.
The authors no doubt know Barbados inside and out and it appears that in 1987, no one in Barbados is using dunder which is surprising because we do know there were quite a few pot stills among producers as well as oddities like the Vulcan three chambered still. It is odd that they mention Arroyo whose writings were already forty years prior and the work had pretty much never been duplicated.
On page 12 there is mention of St. Lucia as being continuous still only, which may have been true in 1987, but these days they are on the map for glorious well aged pot still rum bottled Ed Hamilton. If there was a story about how they rediscovered pot distillation, it would be awesome to know!
On page 13, there is an interesting note in a paragraph relating to the pot stills of Jamaica and Guyana:
The desired desired strength of the rum is about 80% v.v. alcohol and to achieve this it is sometimes necessary to install a small rectifier on top of the high wines retort.
This begs a little math. What would the ABV of their ferment have to be if 80% could not be gotten from triple distillation with the natural reflux of the pot still? Were ferments under 5% being distilled? Worthy Park, evidenced in recent photos, still has a short very cool looking rectifier on top of their high wines retort. Does that imply they start with low ABV washes that likely use a fission yeast?
A unique note from page 13:
A lighter pot still rum is produced in Barbados. This rum is distilled from a mixture of fermented wash and a medium type continuous still rum.
This is quite interesting. They bump the ABV of their ferment with continuous column spirit. The strategy here may be to generate more esterification from a high acid but lower ABV ferment. Does anyone still practice this? Cognac double distillation typically only achieves about 70% ABV with a wine input that is about 7-10% ABV. To hit 80% ABV, they may possibly require an input in the wash still as high as 12%? I’m sure someone can give us a clearer number.
From page 16:
Barbados is peculiar in using both charred and de-charred barrels while all of the other major producers use charred barrels exclusively.
This is followed by the note on page 17:
It appears that a charred barrel is more active than an uncharred barrel and producers more congeners and a darker rum.
Does this imply that Barbados prefers de-charred barrels for their lighter rums?
What is odd, on page 18, is how the higher alcohols in Guyana rums are so much higher than Jamaica. We could speculate that it is the difference between using a budding yeast in many of the ferments versus a fission yeast. Fission yeasts are known to produce far less higher alcohols.
An interesting recommendation from page 20:
With respect to quality, the importance of organoleptic assessment must never be overlooked since some of the compounds which are analytically grouped may contribute good or bad characteristics to the rum.
Starting on page 20, the authors present research into changes they have been noticing in Barbados molasses with the main issue being an increase of amino nitrogen. It looks like the changes are being driven by adoption of mechanical harvesting coupled with more efficient methods in the sugar mill. The result changes nutrition presented to the yeast with the consequent increasing of higher alcohols.
I think the issue with n-propyl alcohol is the difficulty of separation if you are trying to produce a very neutral spirit:
Since n-propyl alcohol is so difficult to remove during distillation, it is therefore very important that its concentration in the fermented wash be kept within a tolerable range.
One page 23, we see the authors referencing a paper from Parfait and Jouret of the INRA, which I just translated!
The authors present their project over the course of 15 pages and it becomes quite technical. It is interesting to consider whether if you faced a challenge like they did during your career, how would you solve it? What they present is technical beyond the scope of those we even refer to as “master distillers”. Would this require a collective effort? Are there any collective efforts we already should be making? I can think of quite a few.