Carbonic Acid Content of Beer

MEETING OF THE BURTON-ON-TRENT SECTION HELD AT THE QUEEN’S HOTEL, BURTON-ON-TRENT, ON THURSDAY, MARCH 1st, 1923.

Mr. A. H. Yeomans in the Chair.

The following paper was read and discussed :—

CARBONIC ACID CONTENT OF BEER
By Julian L. Baker, F.I.C., and Henry F. E. Hulton, F.I.C.

Carbon dioxide is a normal constituent of the earth’s atmosphere to the extent of about 0·03 per cent, by volume. It is continually being removed from the air by green plants which, in the presence of sunlight, decompose it and retain the carbon, while it is again returned to the atmosphere as a product of combustion. Such combustion may be slow as in respiration, fermentation or putrefaction, or more rapid as in ordinary burning. The following diagram shows at a glance the cycle through which atmospheric C0₂ passes during these operations.

It will be seen that the C02 lost from fermenting vessels may eventually return to the mash-tun as carbon compounds in malt or to the copper as hops or sugar.

When sugar is fermented by yeast the reaction may be represented by the following equation due originally to Gay Lusac,

C₆H₁₂0₆ = 2CO₂ + 2C₂H₆0

This is only approximately correct since, as shown by Pasteur, about 5 per cent, of other by-products are formed (glycerol, higher alcohols, acids, etc.) and, of course, some of the carbon is retained to build up the cell substance of the new yeast. On the above basis 1 lb. of cane sugar, maltose or invert sugar would yield 0·49 lb. CO₂, and 0·51 1b. alcohol. This C0₂, if liberated entirely from solution, would occupy at normal temperature and barometric pressure 4·2 cu. ft. or nearly three-quarters of a barrel.

When a barrel of mild ale wort (O.G. 1043°) ferments, 3·66 per cent, of alcohol by weight is formed and being soluble remains in solution. The weight of CO₂ formed at the same time can be calculated from the equation and amounts to 3·5 per cent, by weight. This is equivalent to 18,714 c.c. C0₂ per litre of wort at 15·5° C. and 760 mm. All but about 1 litre of the 18·7 litres escapes leaving about 1 litre of CO₂ dissolved per litre of beer, which is only 5·35 per cent, of the total formed during fermentation. The escaping carbonic acid is equivalent to 102 cu. ft. weighing 12·6 lb. for every barrel of wort at 1043° fermented. Kerr-Thomas (this Journ., 1896, 2, 6) and Maynard {ibid., 1909, 15, 310) have elaborated this point and described means of collecting CO₂ in breweries.

It will be convenient at this point to refer to the solubility of C0₂ in water, alcohol, beer and other liquids. The quantity of gas which a liquid is capable of absorbing depends upon four factors: (1) Specific nature of the liquid; (2) nature of the gas; (3) temperature of the liquid; (4) the pressure. With regard to (1) it has been found that alcohol will dissolve at 0° C. about 2·4 times as much CO₂ as water, while the influence of the nature of the gas (2) is shown by the fact that ammonia gas is 638 times more soluble in water than is C0₂. Temperature (3) is an important factor; thus the volume of CO₂ dissolved by water at 20° C. is only half that dissolved at 0°. Langer (this Journ., 1904, 10, 365) states that between 0°and 5° C. a rise or fall in the temperature of beer of 1° C. causes a fall or rise of 0·1 per cent, in the weight of CO₂ dissolved or approximately 50 c.c. per litre. This we find is an amount in accordance with what can be deduced from a consideration of the solubility factor of CO₂ in beer, and Langer considers that this small difference exercises a great influence on flavour, sparkle and permanence of head, while he is also of opinion that a more stable and satisfactory absorption is achieved by lowering the temperature rather than by increasing the pressure. Finally, the influence of pressure (4) has to be considered. Langer {loc. cit.) found that some beer which had been bunged down for five days showed an increase in pressure of 144 mm. of mercury and an increase in soluble C0₂ of 233 c.c. per litre, which is equivalent to 50 c.c. per litre for an increase in pressure of 31 mm. of mercury. This volume of 50 c.c. per litre was also the difference he found for a variation of 1° C. (between 0 and 5). So it is seen that a pressure of 31 mm. of mercury or a change of 1° C. have the same effect on the C0₂ content. This pressure difference of 50 c.c. for 31 mm. mercury is in accordance with what would be anticipated from the ratio obtaining under Boyle and Henry’s law.

Fernbach (Ann. Brass, et Dist., 1922, 21, 17) states that the average pressure in storage casks at racking is 5 lb. per square inch, and the amount of gas held in solution in beer at that pressure varies according to the temperature from 3·5 to 4 grms. per litre (or 0·35 to 0·4 per cent.). This is equivalent to from 1325 to 2120 c.c. of C0₂ per litre, and, assuming the normal content of beer to be about 1000 c.c. of CO₂, per litre, a pressure of 5 lb. (one-third of an atmosphere) should bring this up to the lower of the two figures cited, namely, 1325 c.c.

The general law of Henry (1803) states that “the volume of a gas absorbed by a liquid is directly proportional to the pressure of the gas.” Since by Boyle’s law the volume of a gas is inversely as the pressure, Henry’s law may be re-stated in the form: “a given volume of a liquid will absorb the same volume of gas at all pressures.” There are two methods of recording this solubility. The first, introduced by Bunsen and known as the “Bunsen absorption coefficient,” signifies the volume of the gas (reduced to 0° C. and 760 mm.) taken up by a unit volume of the liquid when the pressure of the gas itself is 760 mm., or in other words, it is the volume of the gas measured at 0° C. and 760 mm. which is absorbed by 1 cu. cm. of a liquid at the same temperature and pressure and is, therefore, simply the volume representing the “solubility” of the gas reduced to 0° C. The more generally used expression is coefficient of solubility, or “The Ostwald Solubility Expression,” which represents the ratio of the volume (v) of gas absorbed at any temperature and pressure to the volume (V) of the absorbing liquid, that is v ÷V.  This expression differs from Bunsen’s absorption coefficient in that the volume (v) of the dissolved gas is not reduced to 0° C. and 760 mm. and is thus simply the volume of the gas dissolved by unit volume of the liquid at the temperature of the experiment. It is obvious that in cases where Henry’s law holds it will be found that under all pressures this ratio—the Ostwald Solubility Expression—is a constant. In the following table are recorded the Ostwald solubility values of CO₂, in water, beer, alcohol, etc., found by Findlay and others.

The data in this table make it obvious that the solubility of CO₂ in wort and beer is less than in water, and, moreover, if the solubility in beer is compared with a corresponding alcohol-water solution, that in beer is less, as is seen from the following table (Findlay and Shen, Journ. Chem. Soc., 1911, 99, 1317):—

That the solubility coefficient of C0₂ in beer diminishes more rapidly than that in the alcohol solution is to be accounted for by the fact that in the beers with the higher alcohol content there is also a larger amount of dissolved solids and these also exercise a lowering effect on the solubility.

Findlay and Shen in this paper (loc. cit.) then proceed to discuss the origin of the marked discrepancy between their results and those recorded by Langer and Schultze (Zeitschr. ges. Brauw., 1879, 2, 369; 1883, 6, 329). These workers found that beer absorbs or dissolves considerably more CO₂ than the corresponding water-alcohol solution, and even more than water itself, and attributed such increased absorption to the presence in beer, of positive colloids (dextrin and protein) and to the adsorption of C0₂ by these. Langer and Schultze attributed the apparent increase in absorption to these substances—results, however, entirely opposed to those of Findlay and Creighton (Table I, loc. cit.), who found that the presence of dextrin, so far from increasing the solubility of CO₂ in water, considerably diminished it. Findlay and Shen attributed this discrepancy to the method by which Langer and Schultze carbonated their beers, and give reasons for their belief that the solutions were not merely saturated but super-saturated (see also this Journ., 1921, 27, 121).

We considered this brief description of the facts relating to the solubility of CO₂ to be necessary before describing the method we adopted of estimating that gas in beer. Of the many methods which have been devised for this purpose, reference may be made to that of Windisch (Chemische Laboralorium des Brauers, 1902, p. 325 ; P. Parey, Berlin); N. Van Laer (this Journ., 1903, 9, 69), who described the titration method of C. J. Flamen, of Burton-on-Trent; and G. Bode {ibid., 1904, 10, 559). As we did not consider any of these were quite suitable, with the assistance of our late colleague, Mr. F. E. Day, we constructed the very simple apparatus figured in the accompanying block.

Its essential parts are a 50 c.c. burette A graduated to tenths of a c.c. fitted with a two-way tap at the upper end and connected by a short length of stout pressure tubing with a boiling vessel B, which conveniently can be made from a 100 c.c. pipette with the two ends shortened. This in its turn is connected with about 4 ft. of pressure tubing terminating in a 200 c.c. reservoir C fitted with a tap. All the connections should be strongly wired. The apparatus is filled with mercury, special care being taken to prevent the inclusion of air bubbles.

Method of Working.—About 12 to 15 c.c. of the beer under examination are introduced into A by blowing through the side tube fitted to the bottle. The beer is allowed to flow down into B and the contained gas liberated by shaking followed by gentle boiling. The contents of B are then cooled by a stream of water, the mercury levels adjusted and the volumes of gas and beer recorded. In this adjustment allowance must, of course, be made for the height of the column of beer resting on the mercury in A. This involves holding the level of the mercury in the reservoir above the level of that in A by an amount calculated from the expression:

Height of beer column in cms.
Sp. gr. of mercury (13·6)

The volume of CO₂ so found has now to be expressed as a volume at 15·5° C. and 760 mm. pressure, which may be done by multiplying the observed volume by the factor

This corrected volume of C0₂ is then calculated to c.c. per litre of beer. If this figure is required as a percentage by weight on the beer it must be multiplied by 0·000187. As regards published figures relating to the carbonic acid content of top fermentation beers, Fembach (Ann. Brass, et Dist., 1922, 21, 17) writes that he is unaware of any, and in discussing the effect of C0₂ on head formation in such beers he is thrown back on theoretical considerations. N. Van Laer (loc. cit.), however, in an interesting paper dealing with top fermentation beers, showed that ordinary cask beer of O.G. 1045° contained 956 c.c. per litre, and that this, on gyling at various rates, rose to 1316 c.c. and upwards to 2500 c.c. per litre. Ordinary carbonated beer in the trade contained 1421 c.c. CO₂ per litre. Ihnen (Brewers’ Journ., 1915, 51, 221) has recorded values for American beer as high as 2634 and 3170 c.c. per litre, and Lunger (this Journ., 1904, 10, 365) states that German beers containing 1650 c.c. CO₂ are satisfactory but not if less than this.

We have long been interested in the content of carbonic acid in beer and, as far back as 1909, before the chilling, filtering and carbonating system was used for bulk delivery, we made a number of determinations. In all cases the samples were mild ales collected in public-houses belonging to different London breweries.

It thus appears that 13 years ago draught mild ale was retailed with an average CO₂ content of under 800 c.c. per litre. It will be seen later that even under conditions of carbonation in bulk some of the mild ale of to-day contains about this same amount of gas. It may be of interest to record now some recent values for all classes of draught and bottled beers.

From these and many other analyses we have made it appears that bottled beers naturally conditioned show considerable variation in the gas they contain. In draught beers the deficiency of C0₂ is manifest, and we therefore considered it of more practical interest to pursue the investigation with this class of beer with a view of finding where the loss arose and how it might be remedied.

It will be helpful to have in mind some standard or datum line on which to base a judgment, and such a figure may usefully be taken as the point at which a beer is naturally saturated with gas at normal temperature and pressure and after time has been allowed for the escape of the excessive amount of gas retained in the beer (super-saturation) at the time of vigorous fermentation (see Graph I). Such conditions are met with in a beer at the end of its time in the collecting vessel after dropping from the fermenting vessel and just before the disturbance incidental to being passed to the racking tank. A number of beers were collected at this stage and their content of C0₂ determined.

Such beers as a rule taste rather flat and are deficient in the characteristic “grip” on the palate associated with a satisfactory gas content. It is therefore evident that more gas than this will be required for good condition, and such additional gas can only be obtained by introducing more carbonic acid and keeping it there by top pressure until it is released at the time of consumption. Two methods for obtaining .this result are available : (1) by allowing further fermentation to occur in a closed vessel, such as cask or bottle [gyling (Van Laer, he. cit., p. 434) or secondary fermentation] and (2) by forcing extraneous gas into the beer at the end of primary fermentation and keeping it there. The first method is that employed in so-called “naturally conditioned” beers, while the second is adopted in the chilling carbonating and filtering system. Since the solubility of carbonic acid in beer is greater the lower the temperature, such additional gas in the chilling, and filtering system is always introduced at a low temperature and, in addition, there is the advantage of its comparatively slow disengagement when the beer is again allowed to reach the atmospheric temperature.

As the system of treating mild ale with which we are specially familiar involves chilling, filtering and carbonating, we decided to trace the gas content of such an ale through all its stages in order to discover where gas losses mainly occur, since it is evident that beer which in cask or tank-wagon (Graph I) contained 1400/1500 c.c. per litre of gas should, if possible, be retailed on the counter with more than 900 c.c.—a loss of nearly 40 per cent. The accompanying graph shows at a glance the life-history of such a beer so far as its gas content is concerned. After chilling, carbonating and filtering this particular beer was racked into cask and connected with a beer engine as in public-house practice, where the fixed Doulton jar system is not in use. It was kept on ullage, 10 gallons being drawn from it on three successive days in order to simulate actual conditions. On the second and third days the beer was sampled both from the cask direct and through the beer engine in order to discover what loss of gas, if any, occurred owing to the use of the engine.

The graph is interesting in showing the rise and fall in the carbonic acid content of a mild ale. After carbonation the beer once more contains as much carbonic acid as it did after 15 hours’ fermentation. The loss between the cask and counter is considerable and becomes increasingly evident on ullage. The figures also tend to refute the loss which is sometimes attributed to the beer engine, since the gas in the beer, whether collected from the cask direct or by means of the engine, does not greatly differ on the second and third days.

Having thus analysed the loss of carbonic acid which occurs in beer whilst it is being drawn by means of the engine from the cask, it was decided—in order to make this enquiry complete—to determine the losses which take place during delivery by means of the tank-wagon to fixed vessels (Doulton jars) in the public-house cellar. It will be known to many that this latter method of retailing carbonated beer is now assuming large proportions in certain metropolitan districts (see Paul: this Journ., 1922,28, 794 ; Abbot, ibid, 1923, 8).

 A tank-wagon of 20 barrels total capacity was charged with 15 barrels of carbonated mild ale without employing top (counter) pressure and sent to replenish the Doulton vessels at four different houses. Samples, on which the CO₂ was determined, were collected from the wagon at the start of the journey and from the different Doulton vessels when filled. The time occupied in transit and delivery was 2 hours. The results are shown in Graph II.

The figures are of interest as showing the steady loss of CO₂ during bulk delivery to the retailer. The first house on the round received a beer with a high gas content whilst the fourth had lost nearly all it had taken up during carbonation. When these beers were tasted in the order of their gas content the relation existing between their degree of saturation and the effect on the palate was most striking. So far as head-retention was concerned there was little to choose between them.

Two methods suggest themselves by which these losses may be minimised. One is the use of a top pressure of carbonic acid maintained in the tank-wagon during the whole time of delivery, while another is to maintain a top pressure in the Doulton jar in the cellar. (This Journ., 1923,29,8.)

This enquiry, so far as it has been carried out, discloses in a quantitative manner what all along has been felt to be the weakest point in a method of delivery which in other respects has many points in its favour.

The importance of carbonic acid on the palate is so great that it will be evident that every effort should be made to ensure its retention, for when it is borne in mind how much care, attention and expense is devoted to carbonation it is unfortunate that so great a proportion of gas is in some cases lost before the beer is consumed.

DISCUSSION
The Chairman said that some 30 years ago a system of chilling and filtering beer was introduced by Mr. F. Faulkner. He (the speaker) was one of the first to install it. The object was to have bottled and cask beer brilliant without sediment and in good condition. As far as he remembered the procedure was to rack ordinary beer into a cask, prime it and roll it about for several days. The cask was then placed in cold store, when the beer absorbed the carbonic acid. It was then filtered and bottled cold, and when the normal temperature was reached the condition was good. The results were not, however, uniform; sometimes the beer was good, other times it was flat. Improvements were made in the process by chilling and filtering in cylinders under pressure. Carbonic acid was next collected from the fermenting squares and used for carbonating the beer and this was found to be better than using carbonic acid from other sources.

Dr. Slator remarked that the first diagram shown by Mr. Baker was of considerable general interest. It showed the vital part played by the small percentage of CO₂ in the air and the removal of CO₂ from the atmosphere by the weathering of the rocks. The amount now present was a mere relic of what was originally in the air, and there is certainly more than enough weatherable rock left to absorb every trace of this gas. It is interesting to speculate if all free C0₂ will eventually disappear and make life impossible. With regard to the estimation of CO₂ in beer, he had made a number by means of an apparatus published elsewhere (Journ. Soc.Chem. Ind., 1920,39,149T.) and found the method reliable.

Mr. Flamen remarked that some years ago (this Journ., 1903, 9, 70) he had designed an apparatus for estimating carbonic acid in beer. He found that for carbonated beer in bottle 1·6 to 1·7 volumes of CO₂ was a suitable amount, and the figures the authors had quoted corroborated this. The quantity of gas necessary in a beer depended very much on the gravity. The percentage of air in gas was of importance, and mention had been made in the literature that one part in a thousand of gas was a maximum quantity. In collecting fermentation gas it was difficult to keep below 0·2 or 0·3 per cent., and if accidentally 0·7 per cent, was present there was a falling off in the head-retaining properties and in the general appearance of the beer.

He had had experience in sending beer from Burton to London in insulated glass-lined and aluminium railway tanks holding about 60 barrels. The beer (original gravity 1033-34°) was carbonated ready for bottling with a top pressure of 6 Ib. It arrived at a temperature of 40° F. and gave very little trouble. It seemed to him in connection with loss of gas that the trouble of flatness experienced by London brewers who used Doulton jars would be overcome if the beer was delivered cold and with a small top pressure. After delivery it should be allowed to stand and reach a temperature suitable for sale; under such conditions there should be an ample amount of gas in the beer.

Mr. Barrington said he thought the amount of carbonic acid held in solution varied inversely with the attenuation. Carbonated beer was racked at a low temperature, but in Burton the beer had to be dispatched all over the country and by the time such a carbonated beer (if used) was in the houses a secondary fermentation would have started with a resultant cloudiness. On the other hand, naturally conditioned beer would be brilliant and would contain the same percentage of carbonic acid whether it had been carbonated or not. He asked the authors if a carbonated beer kept for six days would be perfectly brilliant or would the secondary fermentation cause a “kick”? It was known that if air got into the carbonic acid a serious waste of beer would arise from excessive fobbing.

Mr. Dryden said some 10 years ago he made a number of carbonic acid estimations in naturally matured bottled beers and he found the volume was roughly 1600 c.c. per litre. A competitor’s bottled ale contained 3000 c.c. per litre and when it was opened followed the cork which, of course, was undesirable. He asked the authors if much gas was driven off when priming was roused into the racking squares? He suggested the possibility of dividing the delivery tank into four compartments so that if a tank was delivering to four houses the beers would each have the same carbonic acid content.

Mr. Julian L. Baker said he would, with the Chairman’s permission, ask his friend and colleague, Mr. H. F. E. Hulton, to reply to the questions that had been asked. He would, however, like to point out that the system of delivery in bulk was greatly on the increase in London, and as experience was gained he had little doubt that houses at the beginning and end of delivery would receive beer with the same content of carbonic acid. In fact, steps were now being taken to effect this. He was aware that the system was not used in Burton, but he thought that a brief account of some of the difficulties might be of interest to the most important brewing centre in England.

Mr. H. F. E. Hulton said he noted that the Chairman was in favour of a fermentation gas as against chemical gas for carbonating purposes. It was rather difficult to see why the gas collected from squares should be so superior to chemically made C0₂, although most people seemed to prefer it. Possibly the presence of small traces of esters in the fermentation gas was responsible.

Dr. Stern said he wished the authors had said more about how to keep the gas in beer. There were, of course, many ways of avoiding losses with which they had experimented, although they had not included the figures in the present paper.

The subdivision of the tank-wagon was an idea that had occurred to many, but its drawback was that the same volume of beer was not delivered to each house. Mr. Flamen’s statement that the amount of gas required for a satisfactory condition in strong beer was less than in low gravity beer was one they could corroborate.

The question of the percentage of air in the C0₂ used to carbonate beer was of importance, and when this reached ½ per cent, it was noticeable in the flavour and the effect on the head was very prejudicial. The figure of 1600 c.c. per litre which Mr. Dryden mentioned was quite enough in their experience to ensure satisfactory condition.

A point in favour of the method described for the estimation of C0₂ in beer was its rapidity, as 30 to 35 minutes was sufficient for a single sample, while as regards fobbing while filling the gas tube, the fact that some fob comes over along with beer did not matter, since such fob was either beer or C0₂ and could therefore be measured eventually as one or the other.

As regards possibilities of minimising CO₂, losses, it was known that the employment of top pressure during the filling of the tank-wagon was advantageous, and another improvement was the retention of such top pressure on the beer in the tank during delivery. A third method was the maintenance of top pressure in the Doulton jar in the public-house cellar as tending to lessen the loss of CO₂ which occurred when beer was on ullage.

Votes of thanks were accorded to the Chairman for presiding over the meeting, and to the authors for their paper.