Conduct of the Alcoholic Fermentation Workshops of Molasses and Beet Molasses Products

Miniac (de) M., 1988. Conduite des ateliers de fermentation alcoolique de produits sucriers (mélasses et égouts). Industries alimentaires et Agricoles 105, 675-688.

This is a very cool paper. It is safe to say it is very significant. Not many new producers will set out to make a rum as described here, but the know-how revealed will help anyone better understand their production.

The big  idea here that will help a lot of people is that of Δ Acid. That is the change in acidity of the fermentation by yeasts and bacteria. Only so much can be expected from yeasts so changes of certain magnitudes can be attributed to bacteria which either represent contamination or intentional flavor production.

Astute connoisseurial consumers may be able to assess this number like ester counts. For rum to progress, enthusiasts have to get away from nerding about the distillation phase and start exploring fermentation. The first big variable is to Pombe or not, and the second big variable is Δ acidity.

Conduct of the Alcoholic Fermentation Workshops of Beet Molasses and Secondary Beet Molasses Products
by Michel de MINIAC*
*Research Fellow at the National Union of Alcohol Distillers Groups

[When this was first translated I may have receieved an incorrect translation of égouts. It was thought molasses implied cane molasses and égouts implied beet molasses. What is more likely is that everything is beet and égouts implies a secondary molasses type. It can be seen in this flow chart here (page 7). This was figured out by the amazing Yves Cosentino! Thanks!

I have yet to correct the paper, so keep that in mind. I’ll get to it soon.]

I. DESCRIPTION OF THE CIRCUITS AND TECHNOLOGICAL PERFORMANCE RECORDED
1. Process by mother pot (figure 1)
2. Yeast recovery method (figure 2)
3. Continuous and discontinuous (figures 3 and 4)

II. OPTIMIZATION OF THE MAIN PARAMETERS
1. Constitution of musts.
a. Nature of the product used
b. Origin of the water used
c. Addition of acid
d. Nutrient salts
e. Antiseptics
f. Defoamers
g. Industrial Implementation

2. The acidity of the must
a. Action of the acidity (figure 5)
b. Search for an optimal value of acidity (figure 6)
c. Choice of acidity, preferably at pH, for conducting a fermentation

3. The rate of non-sugar
a. Purity of the product used
b. Recycling of non-sugar by vinasses [stillage, dunder]

4. Must formulating conditions and alcoholic degree desired (figure 7)

5. Aeration
a. Action of the air On the yeast
b. Implementation of the aeration

6. Yeast recovery = washing and acid treatment
a. Description of the procedure (figure 8)
b. Conditions for a good functioning (figure 9)

III. Origin and prevent of fermentation accidents
1. Bacterial contamination
a. Origins of contamination
b. Fight against bacterial infection (figure 10, 11, and 12)

2. Yeast contamination
a. Recall of yeast metabolism: Brettanomyces
b. How to prevent the development of Brettanomyces

3. Various chemical toxicities
a. Sulphite
b. Nitrites
c. Organic acids
d. Various additives in sugar

IV. GENERAL CONCLUSION AND PROSPECTIVE RESEARCH
1. Facility design
2. Optimization of some parameters
a. The acidity of the must
b. Using an antiseptic
c. Aeration
d. Use of anti foamers
e. Condition of must formulating

Since 1984, four publications have been made in this journal, sometimes in collaboration with industrialists or research institutes, to determine, under what conditions, it is possible to improve the operation of the molasses alcoholic fermentation workshop. These works are referenced at the end of this talk. We have thought, however, to be useful to the profession, by synthesizing here, observations and measurements made on the sites, or laboratory tests to which they gave rise, in order to give useful advice to the sugar industry which wishes in the near future develop ethanol production.

Initially, the outline of the fermentation circuits currently observed in France will be briefly described, and their performance, advantages and disadvantages can be compared.

In a second step, the key parameters implemented will be reviewed, whatever the process used, in order to underline their action, their importance and the optimal values ​​necessary for the proper functioning of the fermentation.

We will finally see the main causes of fermentation accidents by trying to give precise instructions to prevent them, both in terms of the design of the circuits and their conduct.

I. DESCRIPTION OF THE CIRCUITS AND TECHNOLOGICAL PERFORMANCES RECOGNIZED

All fermenting tanks of molasses observed in France can be classified in two groups, according to the nature of the process used: mother tank or recovery of yeast.

In the first case, the yeast is constantly generated from a fermentor supplied with must and aerated. After fermentation, the yeast biomass is destroyed by distillation and is found in the vinasses.

In the second case, the yeast is concentrated in the form of a cream, before distillation, and is the source of biomass, which is returned to fermentation after washing and acid treatment. These two processes will be described in batch and then in continuous multistage. A comparative table of performance, advantages and disadvantages can then be drawn up.

1) Process by mother tank. (Figure no. 1)

It consists in developing constantly, from a mash lightly loaded with sugar, the yeast biomass which will then be used for the fermentation proper and not reused after. The must is of two kinds, depending on whether it feeds the mother vat (weak must) or the fermentation vat (strong must).

– The mother tank: Often made up of several tanks, which are able to produce, in 10 hours of residence time, 30 to 50% of the volume of the fermentation tanks.

In the mother tanks there is simultaneously the production of yeast biomass from a must, and fermentation of this must with a fermentation balance as good as in fermentation itself. The low wort that feeds these vats is low in sugar (70 g/l), so that it can not give a wine higher than 4° GL. The mother tanks are always ventilated. The fermented medium that they produce, will constitute the biomass of yeast which will be sent in the fermentation tank under the name of “pied de cuves” [footing vat].

As mentioned above, it will have to occupy 30 to 50% of the volume of the fermentation tank.

This initial biomass load is of the order of 60 × 10^6 seeds / ml or 3 to 4 g/l of yeast expressed as dry matter.

The whole fermentation goes off at 33°C.

–The fermentation tank: of volume generally larger than the mother tanks, the fermentation tanks are never ventilated. We will see later, that it is however desirable to do so, without risk for the fermentation balance as fear some. These vats therefore receive, at regular intervals, a “vat” from the mother vats. The 50 to 70% of the unoccupied volume is therefore slowly filled with must rich in sugar (= 200 g/l) called “strong must”. This progressive filling is called “pouring“. It is usually done with a colder must (= 15°C) so as to absorb the calories from the fermentation which will keep the mixture at about 33°C throughout the fermentation. A cooling device, (exchangers, or water flow on the walls) is nevertheless put in place.

This pouring of the must that will fill the tank is done in 10 to 15 hours. When the tank is full, the fermentation will be completed at 3/4 (6 to 7 °GL). The last degrees will be completed slowly for ten hours, usually called “fall” because it is followed by measurements of densities that decrease due to the transformation of sugar into alcohol. The total fermentation time from the “pied de cuve” is therefore around 25 hours for a wine at 8.3°GL from a total must of 14% sugar. To compensate for the low load of the low must, the wort, called “strong”, poured in fermentation vat should therefore titrate to about 200 g/l of sugar, depending on the volume occupied by the “pied de cuve”.

During the pouring, for reasons that we will see later, the biomass varies little, especially in the absence of aeration.

Although having low productivity, this process is still often used continuously. It has the advantage of being relatively regular insofar as the production of yeast by the mother tanks remains stable. These constitute the heart of the process and must be monitored with the greatest attention (correct pouring, absence of bacterial contamination).

2) Yeast recovery method. (Figure no. 2)

The “pied de cuve” here consists solely of the yeast cream obtained by centrifugation of the wine before distillation. For reasons that we will see later, the yeast is washed, treated with acid and then regenerated in a small tank in the presence of must and air. This regeneration tank produces, after 3 hours of residence time, a “pied de cuve” of a smaller volume than in the process by mother tank, but much more rich in biomass of yeast: of the order of 3 times more. Unfortunately, the productivity is not increased in the same proportions since for 8.3 °GL the fermentation time descends to around 15 hours (instead of 25). Again, there is no air in the fermentation tank and the conditions of casting and temperature are identical.



This process has the advantage of multiplying by 1.7 to 2 the productivity of the vat room, but the process of washing and treating yeasts is often a source of trouble if it is not very well conducted. We will see in a next paragraph how to optimize the parameters.

3) Continuous and discontinuous

For the sake of clarity, we have just described two batch fermentation processes. It goes without saying that continuous fermentation can be applied in both cases. Figure no. 3 gives an overview of a continuous multi-stage fermentation. Five fermentation tanks (sometimes more), are put in series. The first tank is fed with yeast biomass and must. The yeast can then come from either one or more mother tanks, or a yeast cream regeneration tank in the case of yeast recovery. For tanks of identical dimensions the alcoholic degree ranges between 4° and 8°GL, from the first to the last tank. The fermentation times are about the same as those observed in batch.

The two advantages of continuous fermentation are:

–Ease of driving, since all the maneuvers of filling and emptying the tanks are eliminated. This of course implies a greater ease of possible automation.

– A better and complete use of the volumes since all the tanks are full and not in filling as it is the case in batch (empty volumes estimated at 20%).

The major, and often feared, drawback of continuous fermentation is the risk of bacterial infection and deposits in tanks that are never emptied and cleaned.

There is a particular case of continuous fermentation with recovery of yeast. This is the BIOSTIL process marketed by the ALFA-LAVAL Company.

There is not here multi-stage, but a single fermenter, and moreover, the yeast is recycled without treatment or washing. For reasons of growth of the yeast, the alcoholic degree can not, under these conditions exceed 5 to 6°GL. The process is, on the other hand, directed towards an increase of the non-sugar dry matter in fermentation by intensive recycling of the vinasses (12%). The fermentation time is very short: 6 hours, which gives a high productivity.

To conclude this chapter, here is, in Figure no. 4 a comparative table which recapitulates all the described processes as well as the advantages and disadvantages of each of them.

II. OPTIMIZATION OF THE MAIN PARAMETERS

The realization of an alcoholic fermentation requires first a good growth of the yeast, then the maintenance in activity of this biomass until the end of the fermentation and even beyond if we plan to recycle this yeast for subsequent fermentations.

The growth of the yeast depends first of all on the constitution of the must then on its implementation, that is to say the pouring; the concentration of sugar and non-sugar will also strongly influence the behavior of the yeast. It should be acidified to protect the fermentation against bacterial contamination. Finally, the yeast recovery method with washing and acid treatment that must be conducted under well-defined conditions, also conditions for the smooth progress of the fermentation. We will review all these parameters by trying to explain their role, to specify their optimal values and the conditions of their implementation.

1) Constitution of musts

The term must is used for any sweet solution ready to be fermented by the yeast, and which will give a wine after fermentation. Its constitution depends essentially on the nature of the products used (here: molasses and beet molasses) and the various additives, often of nutritive nature which are added to the basic product. We will first see the various constituents used and then, by what circuits it is desirable to perform this mixture industrially.

a) Nature of the product used

Regarding the sugar industry, all the products resulting from crystallization are currently used in alcoholic fermentation. Depending on the level of sampling, their purity (sugar / dry matter) varies from 92% to 60%. The rate of non-sugar substances in the wort will therefore vary in the opposite direction of this purity, as shown in the following table, for the different products of crystallization.

These “non-sugars” have a decisive action on the growth of the yeast, but their nutritional quality is however restricted because the calco-carbon purification that precedes their production, has largely eliminated yeast growth factors, (acid amines, proteins and various macromolecules).

We have shown (4) that a minimum of 2% of non-sugar was needed in the medium to ensure sufficient yeast biomass. This shows that an EP [Egoût pauvre, I think this implies any time of molasses] is just valid but that at the syrup level, there is nutritional deficiency. This is however filled by a recycling of vinasse. We will come back in detail about these non-sugar problems in another chapter.

b) Origin of the water used

The sugar product is therefore diluted in water to produce a solution of 14 to 16% of sugar, depending on the desired degree of alcoholic fermentation. Water is rarely an urban water, but rather of natural origin (drilling, surface water) which has the disadvantage of bringing bacterial germs or sometimes some chemical toxins. Good filtration is recommended, in the absence of too expensive pasteurization in energy.

Condensed waters are used: Various evaporation condensates: vinasse, evaporation of juice, distillation, etc.

By their heat treatment, they contain few bacteria, which is a good thing, but on the other hand, they are devoid of mineral elements (obligo or macro-elements) which are sometimes indispensable for the metabolism of the yeast (Ca ++, Mg ++, Co ++, B ++, Mo ++) and found in river waters.

c) Addition of acid

The solution thus obtained has a pH most often greater than 7, or even 8 or 9. It is therefore necessary to add a strong acid to bring the must to pH ranging between 3 and 5 depending on the product used. Sulfuric or hydrochloric acids are most often used. The first is sometimes preferred to the second for reasons of resistance of the materials. Biologically, both acids are tolerated by yeast at things from 2 to 3 g/l wort (expressed in g/l of sulfuric acid). We will return in detail on the role of this acidification which is essentially a protection against bacterial contamination, as well as the optimization and regulation of this parameter which must be done not by the pH but by a measurement of the acidity, expressed in g/l of sulfuric acid. [this is a very key point and part of the big takeaway.

d) Nutritious salts

The most commonly added nutrient salt is di-ammonia phosphate. Laboratory tests have shown that its action on yeast was visible up to the dose of 0.4 g/l of must. It is therefore useless to put in higher doses. We have shown that in this case it is the phosphate ion that is active and not the ammonium. It may be thought that the product is largely supplied in nitrogenous substances, but lacks phosphate to ensure the bio-energetic reactions of the yeast.

Sulphate ions are also necessary for the synthesis of sulfur-containing amino acids, but at a lower dose. Experience has shown that an optimum yeast growth was achieved well before the dose of 0.1 g/l of sulfuric acid.

Sulphate ions can therefore be added at this concentration if the sulfuric acid is not itself used for the acidification of the must.

Other cations are also necessary for the metabolism of yeast, but are often brought by water, as we have said above (eg. Ca + + and various obligo-elements).

Magnesium can be added advantageously in the form of magnesium chloride which is used in sugar refinery to regenerate the resins of the Quentin process. A dose of 5 to 20 millimoles of Mg ++ is sufficient, which corresponds to 0.5 g/l of magnesium chloride. It goes without saying that we must abstain from adding Magnesium in a fermentation of Quentin molasses which already contains a lot of them.

e) Antiseptics

We will come back in detail on this subject in the context of the use of acid and the fight against bacterial contamination. Note briefly that the two products used in molasses fermentation are: sodium fluoride (or ammonium, more soluble) and penicillin.

Fluoride is often used continuously at the soda level of 10 to 20 g/l of must if the pH is close to 4.5 to 5. Its action is effective and keeps the bacterial flora at a low level. It can not be used in the case of yeast recovery because the acid treatment at pH: 2 makes it very harmful for yeasts. It does not seem essential to use yeasts “acclimated” to fluoride, because this acclimatization can be done on the site in a few days.

Penicillin was proposed a few years ago at our initiative. Unstable at pH <4, it is best to use it at pH neighbors of 5, for maximum efficiency.

It is strongly recommended not to use it permanently to avoid inducing the production of penicillo-resistant germs. Its action, in case of severe contamination, is effective in 2 days at a dose of 0.3 mg/l of must.

f) Antifoaming

The gaseous releases in these environments very rich in organic materials give rise to the formation of sometimes very important foams which considerably reduce the usable volume of the vat room. Defoamers are therefore added, most often directly in the fermentation tanks, and not in the constitution of the must. These are most often oils of plant or animal origin (fish oil) that it is good to test beforehand to see if they do not contain inhibitors of growth of yeasts. We will see later that their action can be in some cases stimulating for yeast by participating in the biosynthesis of cell members.

g) Industrial implementation

All these compounds constituting the must must be mixed as quickly as possible to minimize bacterial contamination. The basic sugar product (molasses or beet molasses) is practically sterile (a few germs per gram) in view of the heat treatments undergone in crystallization and the concentration in dry matter obtained (close to 80%). On the other hand, its dilution at the level of the must with all the nutrition additives that are added to it, often allows a very rapid multiplication of the bacterial flora.

It is therefore necessary to avoid what is still practiced in the old facilities, namely wort tanks that are alternately filled and emptied. It is not uncommon to obtain, by this process, musts containing 10^6 bacterial germs per milliliter.

We recommend inline dilution in a baffled mixer (type: Sultzer) of all these products which will be mixed in a very short time and immediately sent to the winery. For the future, in the case of correctly homogenized fermenters, we would propose the pouring of the product without dilution, directly in the tanks. It will be necessary to be able to bring the mixture water, acid, salts by another way and in a very precise way, according to a flow rigorously enslaved to that of the product. This mixture having a pH close to 1, it would be protected from bacterial contamination.

2) The acidity of the must

It is an essential parameter, perhaps the most important, in the conduct of an alcoholic fermentation workshop. While in most fermentations the pH is usually used as a control parameter, it is here that the acidity is measured. We will see successively its action on yeasts and bacteria, what optimal value to give it and why it was chosen in preference to the pH.

a) Action of acidity

It is measured in grams per liter of wort, expressed as sulfuric acid, which is also unusual since it is often used in milli-equivalents. Its action is first of all at the level of the yeast whose growth it slows down. Figure no. 5 shows the evolution of the yeast growth rate (μ) in a molasses must whose acidity varies from 1 to 5 g/l. The decrease of μ is almost linear depending on the acidity of the must. The corresponding pH varies from 5.4 to 4. This fall in the yeast growth rate has the industrial consequence of an increase in the fermentation time and therefore a decrease in the productivity of the vat room. It is therefore not possible to conduct a molasses fermentation at an acidity equal to or greater than 5 g/l, although the pH of 4 is not, a priori, incompatible with the development of the yeast. We will see further the reasons for the choice between pH and acidity. The acidity also has a bacteriostatic action which is very important in a non-sterile environment which will see that bacterial contamination remains the permanent danger. more precise curves of this action will be seen in a later paragraph. Let’s say for the moment that the choice of an optimal value of the acidity remains a compromise between two figures: one too low, favoring the development of the yeasts but also of the bacteria, the other one, higher which must be bacteriostatic without too much hamper the production of yeast biomass.

b) Search for an optimal value of acidity

On molasses, the metabolism of yeasts during fermentation gives rise to some organic acids (including succinic acid or acetic acid) which are responsible for a slight increase in the acidity of the wine compared to the initial must. The difference observed is between 0.5 and 0.8 g/l of sulfuric acid. Let’s call “Δ Acid” this difference between the acidity of the wine and that of the must. In the case of excessive bacterial contamination (> 10^6 germs / ml), the “Δ Acid” then becomes greater than that coming solely from the yeasts. If the acidity of the wort has an effective bacteriostatic action, the “Δ Acid” should remain around 0.8 g/l but not higher.

Figure no. 6 shows the acidity of the must and “Δ Acid” in a factory during a month of manufacture. We find that the “Δ Acid” (dotted curve) returns to the level of 0.5 to 0.8 g/l when the acidity of the wort is between 2 and 2.5 g/l. If the acidity of the must (curve in solid lines) goes down below 2 g/l, its bacteriostatic action becomes too weak and the “Δ Acid” takes values ​​between 1 and 1.5 g/l which means the development of bacterial contamination. We will see in a next chapter the evolution of bacterial contamination in relation with this “Δ Acid” and the action of possible antiseptics. The optimal value of the acidity of the must is therefore between 2 and 2.5 g/l. We see from figure no. 5 it is not optimal for yeast, but it allows a compromise between an average growth of yeasts and a bacteriostatic action limiting the bacterial flora to a non-producing level of organic acids troublesome for yeast.

c) Choice of acidity, preferably at pH, for conducting a fermentation

Some people are surprised at the use of acidity as a control parameter, while a pH measurement, more commonly used, is also easier to measure, regulate or even automate.

First of all, let’s say that the pH is not a safe enough guideline here, because it changes a lot depending on the product used, so the rate of non-sugar varies in a must with 14% sugar. The following table reproduces the pH of musts at 2.5 g/l of acidity according to the sweet product used.

Having noted in the previous paragraph, that a molasses must at pH: 4 could not develop yeast because the acidity of 5 g/l was opposed, we also observe that such a pH: 4 is ideal for fermentation of beet molasses EP¹, or EP² acidity close to 2 g/l. [seems like this would be a significant consideration for dunder rich musts]

The buffering effect imposed by the non-sugar rate thus varies the pH between 2 and 5 for the same acidity. However, the nonsugar is variable, first of all according to the products used, and then, as we will see later, according to the non-sugar rate recycled by the vinasse. It is therefore entirely preferable to use acidity as a regulating factor. In addition, it is more accurate at high levels of non-sugar (10 to 15%) because the pH then evolves very little depending on the acidity, due to the high buffering effect.

Another interest of acidity measurements: they make it possible to constantly carry out a balance of the acid production and to determine this “Δ Acid” which remains an excellent indicator of the bacteriostatic action of the medium and the possible progression of the bacterial contamination.

3). The rate of non-sugar

Non-sugar is essential for yeast growth. It constitutes by its mineral and organic elements an excellent nutritive medium for the yeast. We recently published (4) a very complete work on the action of nonsugar on the yeast metabolism, in relation with the selection of yeast strains resistant to an excess of non-sugar which develops in the medium an osmotic pressure too high. There are two main causes of this variation of non-sugar:
– the purity of the product used,
– the recycling rate of non-sugar by the vinasses.
We will examine them one after the other:

a) Purity of the product used

The non-sugar of the must varies from 1 to 10% depending on the product used. The circuits usually used for the fermentation of molasses show us that the yeast perfectly tolerates 9 to 10% of non-sugar brought by this product and the technological performances that we mentioned above come from such a fermentation.

The use of purer products such as beet molasses gives musts less rich in non-sugar. An increase in the growth of yeasts and fermentation is observed, which shows that, with these products, an optimum dose of nonsugar is approaching around 2 to 4%. With syrups, we are at a level close to 1% of non-sugar, which is quite insufficient to the growth of yeasts.

To summarize :
The non-sugar is in excess in the must of molasses, at an optimal dose with the beet molasses and, there is nutritional deficiency in the must of syrup.

b) Recycling of non-sugar by vinasses.

It is necessary, indeed indispensable, for several reasons. First, in the case of syrups, it is an essential nutritional supplement for yeast. Then, in order to concentrate the effluents and save energy on the vinasse concentration workshop, it seemed desirable to increase the non-sugar fermentation rate up to the dose usually tolerated by the yeast, namely 9 or 10% of the molasses must.

In one of our first works (1) we have shown that at equal concentration the non-sugar of vinasse is more active on the yeast than that coming from the molasses. This is because the vinasse contains yeast autolysate proteins which are much more active growth factors on yeast than beet proteins contained in molasses. In the absence of bacterial contamination, the non-sugar of vinasse contains, in addition, no fermentation inhibitor.

If one recycles the vinasse in an Ep² beet molasses must to lower its purity (76%) until that of a molasses (60%), one observes a slowing down of fermentation. But this mixture remains of a better fermentation quality than molasses of the same purity. This has been demonstrated on laboratory tests (1) and verified in a distillery (2). At the same purity as a molasses, vinasse nonsugar produces an increase in yeast biomass, ethanol productivity and fermentation balance.

It has been desired, for reasons of energy saving, a maximum enrichment of the fermentative medium in non-sugar, at a level well above 10% of molasses must. This is possible with yeast strains selected for this purpose. This has been done in recent years as part of our work at the NATIONAL UNION OF ALCOHOL DISTILLER GROUPS. We will resume the main results already published (4). We currently have some strains of yeast capable of producing in 24 hours of fermentation, a molasses wine of 8°GL containing up to 20% of non-sugar. Under these conditions, the osmotic pressure of the medium becomes considerable and greatly disrupts the metabolism of the yeast on several points:
– Decrease of the growth rate, therefore of the biomass maintained in fermentation.
– Strong biosynthesis of glycerol (osmoregulatory of the cell) which results in a proportional decrease of the fermentation balance. It goes from 61 to 55 liters of alcohol per 100 kg of sucrose consumed when the nonsucre increases by 10 to 20% in the fermentation medium.
– Production of higher alcohols and organic acidity are also influenced by the strains.

For a wine at 8°GL, the nonsugar content in fermentation must therefore remain between 10 and 15%. Beyond these figures, the biological constraints become such that the energy saving sought at the level of the vinasse evaporation is largely absorbed by the decrease of the fermentary yield. It goes without saying that if we accept an alcoholic degree lower than 8°GL, the non-sugar will no longer have this harmful effect on the fermentation balance and it will again be possible to increase the nonsugar in fermentation. This is the choice that has been made in the BIOSTIL process (ALFA-LAVAL) which gives the predominance of non-sugar in relation to the alcoholic degree.

The recycling of vinasse does not therefore pose a major problem on the industrial level. The application that was made in a distillery (2) shows it well. However, the bacterial infection becomes more harmful since some of the organic acids produced by the bacteria are recycled into fermentation by the vinasses. It is therefore necessary to be particularly vigilant bacteriologically. Moreover, a large part of the mineral acidity of the must is recycled by the vinasse. It is not harmful, but it must be taken into account when calculating the acidity of the must. This is another reason to prefer acidity to pH in the regulation of alcoholic fermentations.

4) Pouring conditions and desired alcohol content

The decrease in the growth rate of yeasts under the influence of an excess of non-sugar is largely due to an excessive retention of ethanol in the cell under the action of a strong external osmotic pressure. Ethanol is, directly or indirectly, the main growth inhibitor of yeast. We reproduce in figure no. 7 an already published graph (4), which shows the evolution of yeast growth rate on molasses medium when its alcoholic degree varies between 0 and 8°GL during fermentation. If the growth is little affected by alcohol up to 2°GL, it drops considerably to 5°GL where it becomes almost zero. This is the reason why BIOSTIL process can not exceed 5 to 6°GL at 12% of non-sugar.

In fact, in the continuous fermentation of a single fermenter, the yeast is permanently maintained in a maximum alcoholic medium. If the degree approaches 8°GL, the growth rate becomes too low to regenerate the yeast biomass even in yeast recovery.

We therefore recommend pouring instructions which make it possible to maintain the yeast as long as possible in a weakly alcoholic medium (<5°GL), the excess sugar being not at all inconvenient for the growth of the yeast. We had developed this theme in a work (3) which highlighted how the recovery of yeast biomass in a sweet (and non-alcoholic) medium made it possible to increase the ethanol productivity as much as possible: 8°GL in 5 hours of fermentation. In industrial application two cases are envisaged:

– In discontinuity:
After pouring the stock (from a mother pot or yeast), the must must be poured as quickly as possible so as to dilute as much as possible the alcohol produced by the yeast in a large volume of sweet must. If the tank, once full, has a measure of less than 5°GL, we will have maintained better conditions for its growth than if, from the beginning of a casting too long, we allow it to reach 6 to 7°GL . This is unfortunately what happens in practice, because one usually uses a cold must (10 to 15°C) to “cash” the thermals released by the fermentation. If the must is poured quickly it will have to be warmer so as not to cool the winery, then the calories released later will have to be evacuated by exchangers at the bottom of fermenters.

– Continuously:
The dimensions of the tanks in a multi-stage system (SPEICHIM or VOGELBUSCH type) are made in such a way that the first tank has a residence time such that the alcoholic degree is already very high (close to 5°GL). We believe that lower volumes at the head would establish an alcohol gradient between 1 and 5°GL that would be more favorable for biomass production and therefore productivity. In any case, all the must and the biomass must be poured into the first tank and not staggered over several.

5) Aeration

Aeration has always been essential to the proper functioning of a molasses fermentation, unlike that of beet juice which can be content with dissolved oxygen during diffusion. In the present installations air is added, either in the tanks or in the regeneration tanks of the recycled yeast cream, but not in the fermentation tanks. We would like to evoke successively the physiological action of the air on the yeast and its industrial implementation.

a) Action of air on yeast

The metabolism of yeast can take two ways depending on the presence or absence of air.
– Without air, there is fermentation and production of biomass and ethanol.
– In the presence of air, there is respiration, and only biomass production.

These two paths are sometimes used simultaneously with more or less predominance of one over the other.

On the other hand, an excess of glucose in the fermentation medium blocks the enzymes of the respiratory chain and forces the yeast to follow the fermentative route. This is called “CRABTREE effect” or counter effect PASTEUR.

Without insisting on these problems of metabolism which are not at the heart of the subject, it should however be specified that, given the sugar concentrations found in our fermentations, it is impossible for the addition of air, even in high doses, to divert yeast metabolism to the respiratory tract. There is therefore no reason to fear, as some people think, that an excess of air can lower the fermentative balance in favor of yeast production. We are currently working on this subject, and some interesting results will be published soon in an in-depth study.

The air, at low dose, and always on the fermental path, is however indispensable to the yeast. Indeed, the membrane of the yeast cell is composed of sterols whose biosynthesis requires the presence of molecular oxygen.

Yeast can not reproduce in total anaerobiosis unless the medium contains certain sterols or their precursors which are unsaturated fatty acids. In this regard, some defoamers are probably very active, and a study is underway on this subject.

b) Implementation of aeration

From time immemorial, mother tanks have been ventilated without raising the question of the rate of aeration. This never affected the fermentation balance, which remained close to 62 liters of alcohol per 100 kg of sucrose. This is the proof of what we said at the previous chapter: the air can not be, on molasses medium, a factor of decline of the fermental balance. Our current tests confirm it. Precision although it is about molasses, because we had the example of a distillery which, on more favorable medium, (beet molasses), observed a clear reduction of fermental balance, with excessive production of biomass.

We recommend to ventilate the vats or the yeast recoveries at a rate close to 1 V.V.H. (Volume of air per volume of vat room and per hour). However, the study we are currently conducting shows that the air optimum is reached well before this value. Nevertheless, it remains indispensable to the yeast, even during the course of fermentation. In addition, the physiological constraints imposed by the non-sugar on the yeast require a correct membrane structure to remain active until the end of the fermentation and even beyond, in the case of the recovery of yeast. Air, at a very low dose, is here the necessary growth factor. It has, moreover, the advantage of maintaining the biomass in suspension until the end of the fermentation, when the flow of carbon dioxide is not enough anymore.

This permanent aeration, during all of the fermentation has been practiced for two years with profit in a molasses distillery working by the process of a mother tank. The fermentation balance has even improved, because we have shown that a lack of air can bring down this balance. There was also productivity gain.

In the future, the air will have to be measured, not in flow rate but in % dissolved oxygen, by using oxidation-reduction potential probes. The medium in fact, because of its organic matter load, can solubilize more or less the air sent to the vats. The transfer to the cells is then done by the dissolved part of this air.

6) Resumption of yeasts: acid treatment

We would like to recall first of all that the enrichment of the medium in biomass by the yeast recovery process has the only result, foreseeable indeed, to increase productivity. This increase does not occur in the same proportions as those of the biomass, but allows a production increase of the plant of 30 to 50% compared to the operation in the mother vat of the same volume of vat room. This is, of course, very appreciable. On the other hand, and contrary to the opinion of some, the fermentation balance remains unchanged, whatever the process used. This has been demonstrated in the laboratory and several times verified on industrial sites. This yeast recovery technology must, of course, be conducted in optimal conditions. Before going back over the details of its implementation, it is good to recall the objectives to be achieved on the biological level. The molasses wine contains, besides the yeast biomass, a bacterial flora (104 to 100 germ / m) which is not negligible and it is better not to recycle with the yeasts. The proposed method for reducing the level of this contamination is maintenance at very low pH (<2), by adding acid in the yeast cream. This acid treatment can only be acceptable if the quantities of acid that must be added do not exceed a dose compatible with a correct growth of the yeasts (2 to 3 g/l). We know that the buffering effect of non-sugar counteracts the lowering of pH for such doses of acidity. The acid treatment should therefore be preceded by a washing of the yeast cream with water so as to dilute the non-sugar which will be removed in the supernatant of a second centrifugation. This is the purpose of this treatment which has two phases: a first centrifugation followed by a washing with water. Then a second centrifugation preceding the acid treatment that will kill the bacteria without affecting the yeast biomass. The cream thus treated will be regenerated in the presence of aerated wort and will constitute the basis of a new fermentation (dis-continuous) will feed the first tank of a continuous fermentation multi-staged.

The often stated role of “disgorging” yeasts in water to remove toxins is to be taken with caution. We have ourselves carried out in the laboratory for several months (3), a discontinuous fermentation with recovery of yeast without any treatment of the yeast. In the absence of bacterial contamination, yeast performance remained unchanged.

Similarly, the BIOSTIL process recycles yeast without any treatment. We therefore believe that the purpose of yeast washing is only to get rid of the non-sugar to have a more effective acid treatment.

a) Description of the process

Figure no. 8 gives in detail the implementation of this technology, which has three parts: yeast washing, acid treatment and regeneration of yeasts to give the “pied de cuve”.

— First centrifugation and washing of the yeasts.

The wine from a tank at the end of fermentation is centrifuged a first time. The clarified wine is sent for distillation. The yeast cream obtained: primary cream, generally represents 10% of the volume of the wine and has a lodre [a typo might prevent that from translating] centrifugation pellet of 40%. This cream of yeast is mixed with once its volume of water, then agitated, usually by insufflation of air, for about an hour. This washing gives rise to a primary milk which is centrifuged a second time.

— Second centrigugation and acid treatment.

The second centriguation performed on this milk also aims to concentrate the yeast cream to the maximum (75% of pellet) so as to eliminate the maximum non-sugar substances of the supernatant.

This one also called “small waters” contain about 4° GL. It can not be rejected and must be recycled. Two solutions are possible: either it is sent for distillation, but then there is dilution of the wine and thus additional expenditure of energy at the level of the distillation, or it is put back in the must, which makes an additional source of bacterial contamination. The choice must be made according to the bacteriological quality of the circuits.

This centrifugation gives rise to a secondary cream which is mixed with four times its volume of water. Under these nonsugar dilution conditions, the dose of 2 g/l of acid is sufficient to lower the pH to around 2. The acid treatment of the secondary milk lasts 1 to 2 hours and the agitation is also carried out by air. As we said before, the action of the pH is decisive here to kill the bacteria. This is quite different from the bacteriostatic effect of the acidity of the must during the fermentative process. The acid treatment therefore has two components, pH and time, which must act only on the bacteria and not on the yeasts. This must be watched carefully.

— Regeneration of the yeast and formation of the “pied de cuve”.

It is desirable, after this intense chemical treatment of the yeast, to allow it to multiply rapidly before being put into fermentation, most often without air and alcoholic degree too high. The strong must is quickly poured on the “pied de cuve” which remains aerated, so that the yeasts can begin budding. This regeneration is maintained for 3 to 4 hours. The volume thus obtained is the “pied de cuve” which is sent to the fermentation tank.

b) Conditions for good functioning

We would like to mention here the essential points which condition the correct progress of this technology, namely the production of a biomass of yeast in large quantity, constant and partially cleared of its bacteria.

— Search for an optimal rate of acid treatment.

We observed in the laboratory the behavior of a yeast cream containing yeasts and bacteria, depending on the intensity of the acid treatment that was applied to it (pH, acidity and time). The results are shown in Figure no. 9. We see that the usual treatment (pH: 2) only decreases the bacteria by a power of ten but on the other hand, it is safe for yeasts whose population remains almost stable. More intense treatment may be considered (pH 1.5 or 1). We then observe a very strong bacterial decay. This is quite remarkable, especially since the yeast biomass evolves little during one hour.

If the washing has been effective, an acid treatment with a pH of about 1.5 can be envisaged over a very short time. Note the good resistance of yeasts to such a pH, but the disadvantage remains the high acidity (5 to 8 g/l) to be added for such a pH. Rapid dilution of the “pied de duve” will be necessary to lower this acidity, which is incompatible with good yeast growth.

— Recovery of a yeast biomass in good condition.

At the end of fermentation, under the action of the alcoholic degree and the non-sugar, the survival of the yeasts is rather short. This is accentuated by the deposit in the unstirred vats, the organic acidity coming from the bacteria and sometimes a too high temperature at the end of the fermentation. We therefore recommend some improvements:
– Good thermal regulation: 33°C.
– Agitation at the end of fermentation.
– Aeration can be a good way to agitate, because the air improves yeast survival by its action on cell membrane structures.
– Finally, the tanks whose fermentation has been completed must be distilled without delay.

Maintaining yeast biomass for a few hours in a molasses wine quickly causes yeast autolysis. In addition to this observed decrease, a bacterial flora develops rapidly in the yeast deposit and there is also an additional production of secondary compounds detrimental to a good quality of the alcohol (e.g. aldehydes). To minimize this expectation of “dropping” tanks, the distillation feed must be constantly adjusted to the flow of the vat room. This setting is not easy to monitor, but remains essential for a good regularity of a yeast recovery fermentation.

— Search for the most concentrated yeast cream possible.

This parameter is very important. It depends on the performance and implementation of the equipment used. Yeast pellets of 75 to 80% are currently common values in second centrifugation. The maximum concentration of yeast has two advantages:
– In first centrifugation, it limits the volume of wash water that must be recycled.
– In second centrifugation, it allows to minimize the buffer effect of non-sugar and to lower the pH value with the minimum added acid.

To conclude on this process of yeast recovery, it is undeniable that it allows a significant productivity gain compared to the process by mother tank. But this is often achieved at the cost of technological constraints often difficult to control, not to mention the significant investment in centrifugation equipment.

The process by mother tank, despite its low productivity, however, has some advantages of regularity of operation and simplicity of implementation. We are currently researching under what conditions it would be possible to increase this productivity in order to bring it closer to that of the yeast recovery process. Interesting results have been obtained on this subject and already implemented in a distillery. We will discuss this again in the general conclusion.

III – ORIGIN AND PREVENTION OF FERMENTATION ACCIDENTS

All the parameters we have just mentioned are globally responsible for the smooth running and especially the regularity of a fermentation workshop. A fermentation accident is never a sudden drop in productivity, but a slow decline whose causes are sometimes multiple and often difficult to elucidate. We would like to examine here, under what conditions these accidents occur and what are the means of avoiding them, both in terms of the conduct and the design of the workshop. We will start with the bacterial contamination which remains at the origin of the great majority of fermentation accidents. More recently, a few sites have developed varieties of yeasts producing acetic acid (Brettanomycès), which have seriously disrupted production. We will present the first results of work on this subject, in order to avoid the development of such a flora. Finally, we will see the various chemical toxicities from the products or their recycling, as well as the errors in the conduct of the fermentation workshop.

1) Bacterial Contamination

This remains the weak point of all molasses fermentation circuits. The high non-sugar maintains a pH close to 5, very favorable to the development of the lactic flora. Beyond 100 germs per milliliter, these bacteria produce in the medium an organic acidity which is responsible for the inhibition of yeast growth and therefore the productivity of ethanol. Its action is shown at doses of 0.5 to 1 g/l of acidity produced.

As it is excluded, for reasons of financial profitability, to practice the thermal sterilization of the circuits, we must do everything to ensure that the bacterial flora does not exceed 10^6 germs / ml.

a) Origins of contamination

The sweet product used (molasses or beet molasses) is very little contaminated, given the high thermal scales to which it was subjected during its manufacture. The high concentration of dry matter (80%) also prevents any bacterial growth. These products are however not completely sterile (some germs per range) and therefore provide the necessary seeding medium. At the time of the dilution of musts the bacterial development is very rapid and can reach commonly 10^4 to 10^5 germs / ml.

At this stage there is still no production of organic acid, but the fermentation must be able to take place in the presence of a bacteriostatic agent which prevents the bacterial flora from reaching or exceeding 10^6 germs / ml. The acidity of the must therefore plays this role, as we have already mentioned.

Apart from the must, the various parameters that accentuate the risks of bacterial contamination are as follows:
— The must dilution workshop.
The use of bins is to be avoided. They are difficult to clean and impose in the middle a residence time too long.

Inline dilution is preferable for its low residence time and the possibility of easier automation.

— The fermentation temperature.
It must not exceed 33° C. At higher temperatures (35 to 37 ° C) the growth of bacteria is favored at the expense of yeasts which become less resistant to alcohol and non-sugar.

— The residence time of the dropping tanks.
It promotes the deposition and autolysis of yeasts which is an excellent factor of bacterial contamination.

— The use of flat-bottomed tanks.
It promotes deposits, therefore acts in the same way as before and must be particularly canceled in the case of continuous fermentation.

— The storage of recycled light vinasse.
Non concentrated, vinasse from distillation are sometimes not sterile, especially in the case of vacuum distillation at low temperature. The buffer tank before recycling should be as small as possible, kept warm and often cleaned. It is also a source of contamination.

— The washing water of the fermentation gases.
The alcohol carried by the carbon dioxide is recovered by a washing column. The alcoholic water must preferably be sent for distillation because it contains a non-negligible bacterial flora. The same is true of yeast wash water (or “small water”).

b) Fight against bacterial infection

Apart from certain instructions for conducting or designing the installations, the fight against bacterial infection requires permanent monitoring of the bacterial flora at all levels of the installation. The main points to watch are:
• the dilution of musts,
• the tank at the end of fermentation,
• the mother-pot eventually
• acid-treated yeast cream (in case of yeast uptake).

Counting techniques are multiple. We use petri dish spreading with M.R.S. at 0.5 g/l of actidione. The counts can be done in 24 hours, if the incubation temperature is raised to 35-36° C. Recall the significant parameter of the bacterial infection: the “Δ Acid”. It is a fast, simple measurement, the result of which is immediate. It measures the difference between the acidity of the wine and that of the must during pouring. In the absence of bacterial infection, it is maintained between 0.5 and 0.8 g/l and above all it must remain constant.

Its increase of 0.5 g/l is significant of a contamination. We refer you to figure no. 6 which traces the evolution of “Δ Acid” according to the bacteriostatic action of the acidity of the must.

When this is insufficient and there is an increase in “Δ Acid” and bacterial flora, we can use some antiseptics.

We will mention two that are commonly used: sodium fluoride and penicillin (sodium G. de Rhone-Poulenc).

Figures no. 10-no. 11 and no. 12 show the comparative action of the acidity of the must, sodium fluoride and penicillin On:
• the growth rate of bacteria,
• the “Δ Acid” product,
• Yeast biomass.

As we have already seen, the acidity of the must, if it has an action on the growth of the bacteria, also has action on the yeasts. It is found that beyond 2 g/l its action is reflected on the yeasts. At lower values, the yeast is hampered by the acidity of the bacteria as shown by the “Δ Acid” curve (2-5 g/l). The use of a good antiseptic acting on bacteria and not on yeasts is here very appreciable.

Sodium fluoride (between 5 to 20 g / hl of must) has a weak action on the bacteria, but allows a good regeneration of the yeast biomass by a very correct attenuation of “Δ Acid”.

Penicillin is more effective on the bacterial flora. Its dose should be 0.3 mg/l (or, 16.5 million international units per gram). Beyond, Its action is not better, but it does not risk to hinder the yeast.

2) Yeast Contamination (Brettanomycés)

The fermentation of molasses is made by the yeast Saccharomycés cerevisiae. In most cases, bakers yeasts are used for their ease of use. It is possible that so-called “wild” yeasts develop in the medium since no sterilization is performed at the product level. If their growth rate is sufficient, they can maintain and produce alcohol without hindering the fermentation of the bakers strain initially implemented.

However, serious fermentation accidents have occurred, for 2 to 3 years, by the contamination of a yeast which eliminates the original strain and is nevertheless unable to perform the fermentation with a suitable productivity.

Three cases have so far been observed. The incriminated microorganism was in all cases: Brettanomyces intermedius.

a) Recall of yeast metabolism: Brettanomyces.

In 1940 M.Th. CUSTERS describes this yeast in his thesis, by its peculiarity of developing a negative PASTEUR effect, which was later called the CUSTERS effect.

Indeed, the alcoholic fermentation of this yeast is not inhibited (PASTEUR effect) but stimulated by the presence of air. In addition, its fermentation is accompanied by a high production of acetic acid.

We have recently shown that the acidity of the must, in values of 3 to 4 g/l, stimulates the growth of this yeast while it prevents the development of Saccharomyces cerevisiae.

b) How to prevent the development of Brettanomyces.

Such contamination is the consequence of two-level driving errors in the workshop.
– An excess of air,
– An excess of acidity of musts.

Under these conditions, the new yeast strain (Brettanomyces) implants in the medium, to the detriment of Saccharomyces cerevisiae whose growth rate is greatly reduced compared to that of the infecting strain. In addition, there is then biosynthesis of acetic acid by Brettanomyces which eventually is an inhibitor and the fermentation is considerably slowed or even incomplete. We are currently very poor in handly this type of contamination. In all cases, a complete liquidation of the vat room is necessary and the restarting can only be done after a thorough cleaning and disinfection of all the circuits.

The first accidents of this kind appeared on a BIOSTIL implementation site. Note that the ALFA-LAVAL Company recommended the use of Saccharomyces pombe instead of Saccharomyces cerevisiae for reasons, announced, for better resistance to the osmotic pressure of non-sugar, but which would require, to increase the low growth rate of Saccharomyces pombe, a very strong aeration. This, combined with a poorer control of the acidity of musts, is probably responsible for the accidents observed on the site and which were due to the invasion of the medium by Brettanomyces. [return of our hero Pombe in a high productivity environment]

Subsequently, in two other factories, a similar phenomenon appeared, and it seems, again, that this is the result of errors of conduct in the air or acidity. Laboratory tests clearly show that with a very low aeration (0.5 V.V.H.) and an acidity of between 2 and 2.5 g/l, Brettanomycès strains can not develop in molasses medium. It is therefore necessary to master these two parameters perfectly. We are currently working on aeration and will soon propose very specific instructions on this subject. As for the acidity of musts, it would be desirable to be able to control the bacterial contamination without resorting to an excess of acidity. Either by a better bacterial cleanliness of the medium, or by the use of a correct antiseptic allowing to work with a low level of acidity of the musts (1 to 1.5 g/l).

3) Various chemical toxicities

Apart from certain contaminations, the yeast can be hampered in its development by different products which have been introduced or have originated, in the manufacture of the sugar product.

a) Sulphites

Present in sugar products, its toxicity appears in fermentation from 1 g/l of must. It also varies according to the degree of binding of SO2 with other organic substances, and also according to the pH of use.

b) Nitrites

They are often formed in tower diffusers that facilitate the development of anaerobic nitrous fermentations. Inhibition of fermentation appears from 0.3 g/l of nitrates. On the site, the presence of red nitrous vapors is often visible in open tanks. It is therefore necessary to carefully monitor the bacterial contamination of diffusions that can develop an anaerobic flora of nitrous bacteria.

c) Organic acids

We have shown in molasses the presence of organic acids (type: lactic) (2) which are released at the time of the acidification of musts. These acids act on yeast, such as those we have already spoken of at length, which are produced by the bacterial contamination in fermentation. They come here from diffusion where an identical flora can also develop. The sugar industry is well aware of these problems, which it solves with antiseptics (formalin and others). It should be noted that increased monitoring at this level will be beneficial to the fermentative quality of the molasses type sweet product. Organic acids are indeed heavy products that concentrate in crystallization.

d) Various additives in sugar

Most of the additives used in sugar (biocides, defoamers, gutters, pressing additives, ect …) are found at a concentration multiplied by ten years in molasses. Some of these substances may be harmful to yeast. It is therefore necessary to make sure before use.

IV. GENERAL CONCLUSION AND PROSPECTIVE RESEARCH

As we said at the beginning of this work, we wanted to give an account of the conditions under which the various molasses fermentations workshops operated. We set ourselves the goal of proposing an optimization for each parameter and the overall design of the installations. We would like to return here, on a few key points that will determine our prospective research in the coming years at the NATIONAL UNION OF ALCOHOL DISTILLER GROUPS.

1) Facility design

These are designed according to two main principles vat-mother or recovery of yeast. In both cases the fermentation balance remains the same, only the productivity is increased by the yeast recovery process. This is often done at the cost of numerous technological constraints and often difficult to control.

For our part, we want to develop the process by mother tank by trying to increase its productivity by optimizing certain parameters stimulating the development of yeast biomass. In recent years, we have developed aeration throughout the fermentation process. We give as an example, the case of the ORIGNY-SAINTE-BENOTTE distiller which could thus significantly increase the productivity of its vat room, operating by mother tank.

The average figures for the last campaign are:
– Alcoholic degree of the wine: 10.6 g/l,
– Non-sugar wine dry matter: 10.5% (13.5% on vinasse),
– Fermentation time: 27 hours,
– Fermentation balance: 61.6 liters of alcohol per 100 kg of sucrose,
– Productivity of ethanol: g/l/h: 3.10.

The productivity is thus close to that obtained by yeast recovery, without having the constraints of this process.

2) Optimization of some parameters

a) The acidity of the must

This must remain constant in order to constantly appreciate the “Δ Acid”, a witness of bacterial contamination. We propose that it evolves between 2 and 2.5 g/l of must, in order to avoid contamination by Brettanomycès.

b) Using an antiseptic

Penicillin gives good results but can not be used permanently for fear of giving birth to penicillo-resistant germs. Sodium fluoride is used only in the mother vat, because the acid treatment process makes its action very harmful for yeasts. Traces of hydrofluoric acid destroy the yeast biomass. The search for an antiseptic that can supplement the bacteriostatic action of the must acidity would have many advantages.
— Gain of productivity by operation with weak acidity (1 to 1.5 g/l),

Non-contamination by Brettanomycès at low acidity.

Recycling vinasse facilitated by a weak organic acidity in the absence of bacterial infection.

We are in contact with antiseptic producers working on this subject in collaboration with us. The use of a new product at this level remains however subject to some conditions, besides its biological action:
–Does not interfere with yeast, in a wide range of use against bacteria.
– Does not give pollutants in the alcohol after the heat treatment of the distillation.
–Does not give toxic products in concentrated vinasse, often used for livestock feed.

An important but still useful work remains to be done in this field for the coming years, especially in the context of large production plants (5,000 hl of pure alcohol per day).

c) Aeration

We have recently shown in the laboratory that this ventilation remains indispensable, but could be reduced to a minimum. More precise figures will be given soon, as well as the methods of measurements on the sites by probes of potential of oxydo-reduction.

This decrease in air should also help to avoid contamination by Brettanomyces whose alcoholic fermentation is stimulated at high levels of air.

d) use of antifoam

According to researchers specializing in this field, some defoamers may contain unsaturated fatty acids that would enter the biosynthetic pathway of yeast membrane sterols. They would participate in the survival of yeasts under adverse conditions of alcohol and osmotic pressure. A study is underway on this subject and will lead to the action of defoamers used either in distillery or at different stages of the sugar industry and which are found more or less chemically transformed into molasses.

e) Terms of casting

Maintaining a low alcohol environment at the start of fermentation is also a yeast growth factor. It can be promoted especially continuously by sizing the heads of tanks imposing the fermentation medium very short residence times to obtain an alcohol gradient less than 5 ° GL.

We would not want to finish this work without telling the Industrialists that we have tried above all to inform them and help them. We hope that this work, which is necessarily incomplete and uncriticized, will elicit from them comments and information that may complement or even modify the broad lines of research that we will pursue over the next few years for them.

BIBLIOGRAPHIC REFERENCES

(1) M. de MINIAC, Fermentation alcoolique des sous-produits de sucrerie. Ind. Aliment. Agric. 1984, 101, 3, 123-135.

[Alcoholic fermentation of sugar by-products.]

(2) G. ALARD, M. de MINIAC. Recyclage des vinasses ou de leurs condensats d’évaporation en fermentation alcoolique des produits sucriers lourds (mélasses et égoûts). End. Aliment. Agric. 1985, 102, 9, 877-882.
[Recycling vinasses or their evaporation condensates in alcoholic fermentation of heavy sugar products (molasses and beet molasses)]

(3) M. NOMUS, M. de MINIAC. Gain de productivité d’éthanol en fermentation alcoolique des produits de Sucrerie (mélasses et égouts) Ind. Aliment. Agric. 1985, 102, 971-985.
[Productivity gain of ethanol in alcoholic fermentation of sugar products (molasses and beet molasses)]

(4) M. de MINIAC. Sélection de souches de levures pour la fermentation alcoolique de milieux mélassés enrichis en non-sucre de vinasse. Ind. Aliment, Agric. 1987, 104, 425-439,
[Selection of yeast strains for the alcoholic fermentation of molasses enriched in non-sugar of vinasse]