MEETING OF THE SCOTTISH SECTION, HELD AT THE NORTH BRITISH. By A. A. D. Comrie, B.Sc, F.R.I.C.

200 [J. Inst Brew. MEETING OF THE SCOTTISH SECTION, HELD AT THE NORTH BRITISH HOTEL, EDINBURGH, ON TUESDAY, 8th FEBRUARY, 1965 Mr. R. Allan in the Chair The following paper was read and discussed: BREWING LIQUOR TREATMENT By A. A. D. Comrie, B.Sc, F.R.I.C. {Messrs. Heron & Comrie, Trafford Laboratory, Manchester, 17) Received 10/A February, 1055 Reviewing the influence of liquor salts at successive stages of the brewing process, the special significance is emphasized of the ions of calcium and bicarbonate which, by interactions involving hydrogen ion interchange, affect the ph of wort and beer with consequent influences on wort composition and beer stability. Other ions may influence flavour. Methods for securing brewing liquors of compositions suitable for different types of beer are discussed in detail, special emphasis being placed on the removal of carbonate and on the importance of ph control throughout the brewing process. Introduction It is easy to understand that the quality of a beer depends very largely on the kind of malt and the type of hops that are used in its making, but the influence of the third major component, water, is less obvious. Perhaps an analogy with tea will be helpful. In tea making, almost every housewife knows of the important part played by the com position of the water in bringing out the flavour; so much so, that teas are blended specially to suit the water of the district in which they are to be sold. In something of the same way the flavour of a beer is affected by the composition of the brewing water, but here the analogy ends, for the salts in the water also influence the course of the com plex enzyme actions of the mash, and they help to regulate the physico-chemical changes that take place during boiling, cooling and fermentation of the wort. In the early days of brewing the com position of the water that was naturally available decided the type of the beer being brewed without the brewer being aware of it. Burton came naturally to pale ale, Dublin and London to stout, Pilsen to pale lager, and Dortmund to dark lager. Eventually the idea arose that the brewing water had something to do with it, and in time this idea was confirmed by analysis of the waters. Broadly speaking, it was discovered that soft, non-alkaline waters were suitable for pale lagers, while waters containing mainly car bonates were best used for stouts and dark lagers. With that knowledge, brewers started experimenting with the deliberate adjust ment of the composition of their waters, in order to brew the beers of their choice, and so began Brewing Liquor Treatment. To understand liquor treatment, one must know what contribution is made by each of the salts dissolved in the liquor, and the first thing to remember is that the salts are split up in solution into ions, and that it is to these ions that we owe the activities commonly ascribed to the salts. The ions of principal importance are those of potassium, sodium, calcium, magnesium, nitrate, chloride, sul phate and carbonate, and in the main they affect the operations of mashing, sparging, boiling, cooling and fermentation, and the flavour and stability of the beer. Influence of the Ions in Brewing Liquor Mashing, The ions of importance here are those of calcium, carbonate and, to a smaller degree, magnesium and sodium. For the present purposes sodium may be taken to include potassium, which will not be referred to separately. Calcium causes a fall in ph by its inter action in the mash with phosphates and

Vol. 61, 1955] comrie: brewing liquor treatment 201 protein from the malt. The phosphates form an equilibrium mixture of ions: -H+ -H+ -H+ H%VOAr^=^HtVOA'^^^.UPOA' -^=2^ PO4- - - +H+ -fh+ +H+ the majority being primary ions, HaP04~. In moving from left to right along this system, H+ ions are set free in three stages. By forming an insoluble tertiary salt, and a secondary salt which is only slightly soluble, calcium ions remove the PO4 and a proportion of the HPO4- -, so causing a move ment to the right to make good the loss. This liberates H+ ions and lowers the ph of the mash. Magnesium does not have the same effect because secondary and tertiary magnesium phosphates are comparatively soluble, and in fact, when present in quantity, magnesium tends to delay the precipitation of tertiary calcium phosphate until the wort is boiled, so diminishing the fall in ph in the mash tun and increasing it in the copper. Sodium has a somewhat similar action. The ions of calcium and magnesium, being bivalent, will also cause the precipitation of some of the protein of the mash, and this in its turn causes a slight additional fall in the ph. It was originally believed that the phos phorus concerned in the above reaction was entirely inorganic, but recent work has shown that it is mainly organically com bined phosphorus which is involved. This type of compound yields primary potassium phosphate on hydrolysis, and the ionic equilibrium which has been described is probably as good a way as any in which to represent the reactions with calcium, although it may not be literally correct. The carbonate ion operates in the reverse direction. The ionic equilibrium is in this case: -H+ +H+ -H+ +H+ and since H2CO3 is lost (as CO2) under the conditions of temperature and acidity of the mash, the equilibrium is displaced from right to left and H+ ions are absorbed, so raising the ph. Furthermore, roughly twice the number of H+ ions are removed in moving from COS- - to HaCOa as are set free by an equivalent amount of calcium in moving from a mixture of H2PO4- and HPO4~ - to PO4, which means that the carbonate ion is twice as effective in raising the ph as the calcium ion is in lowering it. In actual fact the carbonates in a water generally exist as bicarbonates, and there is only one H+ ion removed from solution in passing from HCOa~ (the bicarbonate ion) to HaCOs, but the number of bicarbonate ions is double the number when expressed as carbonate ions (i.e., Ca(HCO3)2 is con ventionally expressed as CaCO3) so the result is the same. like the precipitation of tertiary phos phate, the loss of H2CO3 as CO2 is partly delayed until the wort is boiled, when its loss has the effect of diminishing the fall in ph that occurs on boiling. Consider now what effect this rise or fall of ph in the mash is going to have, a- Amylase has a ph optimum at 5*7, js-amylase at 4-7, and the proteolytic enzymes of malt have optima between 4»6 and 5-0. The ph of a malt mashed with pure water is in the neighbourhood of 6*0 so that calcium ions, by lowering the ph, will favour saccharification and proteolysis, leading to increased extract from the malt and a rise in the amount of readily-soluble nitrogen in the wort. Calcium ions, as such, also have a protective action on a-amylase, and extend the range of optimal ph of all the enzymes. Carbonate ions, by raising the ph, will tend to increase the dextrin/maltose ratio and give less ferment able worts, and also leave a greater amount of undegraded nitrogenous matter in colloidal solution. Most of this will be precipitated before the wort reaches the fermenting vessel, but part of it will remain to lower the colloidal stability of the beer. From a purely quan titative point of view, less extract is obtained from the malt if the ph of the mash is raised. It may be noted that calcium and mag nesium ions will precipitate js-globulin rather than albumin in the mash (since the former bears a negative charge at the ph of the mash), and this is probably an advantage because j3-globulin is often considered to be the source of most of the haze-forming sub stances of beer. The phosphates are the main buffering substances in wort, and calcium ions, by precipitating phosphates, cause a reduction in buffering capacity which more than offsets the slight increase in buffering (due to amino acids) which results from the stimulation of proteolysis. As will

202 comrie: brewing liquor treatment [J. Inst. Brew. be seen later, this has a favourable effect on the ph of the beer. Calcium ions also initiate the precipitation of oxalate derived from the malt, which continues to come out of solution as beerstone throughout the brewing process. In the absence of sufficient calcium, oxalate precipitation may be unduly delayed and can in some circumstances cause haze in beer. Sparging. In the course of sparging, the amount of buffer substances remaining in the mash naturally diminishes and the influence of the ph of the liquor increases. The carbonate ion, by leading to a rise in ph, increases the extraction of colouring matter, astringent bitter-flavoured substances and silica from the grains, since these substances are all acidic in character, and peptizes a small amount of nitrogenous matter to a form of colloidal solution which is liable to give haze trouble later on. The chloride ion is considered to aid this process of peptization. As far as extraction of tannin and colour is concerned, there is here an analogy with the effects of different waters on tea infusion. Boiling and cooling. The reactions started in the copper are continued during cooling, so the two processes may fairly be considered together. Shortly after the start of the boil all the carbonate will have disappeared and precipitation of calcium phosphate will be complete, so that whatever effect either is having on the ph of the wort is stabilized. The lower ph produced by calcium ions lessens the rate of extraction of the soft resins of the hop, and slows up the speed of their conversion to the more soluble isoproducts. The wort will therefore be less bitter and contain less preservative; indeed, a highly gypseous liquor has been called a hop-waster since, to get the same degree of bitterness, more hops have to be used. This wastage of hop resin may be augmented by the increased flocculation of protein and protein-tannin which takes place at the lower ph, since the fiocculum adsorbs hop resin. Apart from this possibility, the im provement in copper break and cold break makes an important contribution to the health of the fermentation and the soundness of the beer. The improvement is due in part to the lower ph, and in part to the actual concentration of calcium and magnesium ions and their precipitating action on pro tein, as in the mash. The calcium ions also have the property of restraining colour formation during the boil, particularly of the reddish shades that enter into the colour of beer. In wort of a higher ph these effects are reversed. The bitterness is more pronounced and rather harsher, and the separation of unstable protein and protein-tannin com pounds is less sharp. In addition, more colour is extracted from the hops and produced by caramelization. Fermentation. If flocculation in the copper and during cooling has not been satisfactory, the fermentation will suffer because there will be progressive precipitation of substances which hinder the activity of the yeast. The presence of silica in more than very small amount, whether derived from the brewing liquor or extracted from the malt, has been known to cause particular trouble of this kind due to the formation of a protein com plex which is adsorbed on the surface of the yeast. Such clogging of the yeast will not only affect the fermentation but possibly the flavour of the beer also, and may interfere with the flocculation of the yeast. Calcium ions, on the other hand, seem to be essential to satisfactory yeast flocculation, while chloride ions have a tendency to check fermentation. A comparatively high initial ph due to the use of an untreated carbonate liquor has been said to improve the fermenta tion, but this is only likely to be true with a flocculent yeast, which tends to separate earlier if the ph is low. Closely related to the soundness of the fermentation is the health of the yeast, and the yeast will not be healthy unless it has certain mineral salts in its diet. Foremost among these are the salts of calcium, sodium, potassium, magnesium and phosphoric acid. Although it is likely that, with, sound brewing methods and materials, sufficient of all these salts is contributed by the malt, the nutrition of the yeast is safeguarded in this respect by the presence of modest amounts of mag nesium in the brewing liquor. It is doubtful whether increasing the calcium content of the liquor will have any significant bearing on yeast nutrition because much of the calcium is precipitated as phosphate before it reaches the fermentating vessel, and excessive amounts of calcium in the liquor are more likely to operate in the reverse direction and cause the yeast to suffer from lack of phosphate. This may be the reason for the few occasions on which it has been said

Vol. 61, 1955] comrie: brewing liquor treatment 203 that calcium sulphate in the liquor has an unfavourable effect on yeast and fermenta tion. Calcium ions have an important indirect effect on yeast nutrition by their stimulation of proteolysis in the mash tun, whereby the amount of yeast-feeding nitrogen in the wort is increased. No mention has yet been made of the nitrate ion. This serves no useful purpose in brewing liquor, and in amounts of more than about 3 grains per gal. it will cause progres sive deterioration of the yeast. The actual poisoning is probably due to nitrite produced by biochemical reduction of the nitrate, and its toxicity is enhanced by the presence of chloride ions in excess of about 20 grains per gal. Yeasts vary in their susceptibility, but a nitrate content of not more than 2 grains per gal. is generally harmless. Beer stability and flavour. The preserva tive value of the resins in the beer, responsible for checking lactic bacteria, is increased by fall in ph and reaches a maximum at ph 3-9. Calcium ions, by their precipitation of phosphate buffer substances, make it easier for the beer to approach this value in the course of fermentation. It should be noted.that the lowering of liquor ph by the removal of carbonate alone will not have the same effect, since the presence of calcium ions is necessary as well. The presence of carbonate in the liquor, however, by raising the ph of the wort, will leave more unstable nitrogen compounds in the beer and so lower its colloidal stability, just as calcium ions, by their action in the mash tun and the copper, will improve its colloidal stability. Exaggerated claims have sometimes been made regarding the influence of liquor com position on the flavour of the beer, but even when these are discounted there remains the fact that the flavour given by the malt and the hops is influenced to some extent by the presence of certain ions in the liquor which persist into the beer. These ions are those of magnesium, sodium, chloride and sulphate. The magnesium ion has a very marked sourto-bitter flavour, and can be detected in concentrations of little more than 1 grain per gal. Sodium has a slightly sour-saline effect on the palate, potassium is purely saline, sulphate is rather more dry and bitter, and the chloride ion is sweet and full. Three things follow from this: (a) the amount of magnesium must be kept low; (b) the sulphate/ chloride ratio should be higher in bitter beers than in mild ales; and (c) the balance of flavour from sodium linked with chloride is better than from sodium linked with sulphate. A fourth flavour-effect is due to the fact that the chloride ion can be tasted at a concentra tion lower than the sodium ion, from which it follows that an increasing concentration of sodium chloride will start by being sweet and end by being mawkish and unclean.* In directly, calcium has a beneficial effect on the flavour of the beer as a result of its influence on the ph of the wort, but excess of calcium, by its effect on proteolysis and hop extraction, produces a harsh, thin flavour, lacking in hop character. Effect of Trace Elements In addition to the ions which have been discussed, there are others which may be found in a water supply in very small amounts which are not without significance. Iron, and its frequent companion manganese, will upset yeast and give an off-flavour to beer if present in more than a few parts per million, but in solution it is hardly likely to survive the brewing process to any significant extent. As a liquor contaminant, therefore, it is more likely to cause trouble in pipelines than in the brewing process itself. Copper is known to be toxic to yeast if present in wort in excess of about 10 p.p.m., but the amount contributed by any natural water would be insignificant compared with the amount dissolved from the copper plant in the brewhouse. Fluorine in water has assumed dietary importance in recent years, and the time may come when it will be added (as fluoride) to water supplies in some parts of the country to ensure a concentration of about 1 p.p.m. This concentration has been found to be quite without effect on fermentation and yeast health, and the yeast itself is not harmed by concentrations up to 10 p.p.m. As far as fluorine is concerned, therefore, any water which is safe to drink is safe for brewing. Requirements of Brewing Liquors With these facts in mind, it is now possible to decide what is wanted and what is not wanted in a brewing liquor. For this reason some brewers prefer to add potassium chloride in place of sodium chloride, as the potassium ion is free from the slightly sour flavour of the sodium ion.

204 comrie: brewing liquor treatment (J. Inst. Brew. Pale ales and att beers of a delicateflavour. For these, carbonate must obviously be absent, or at least its effect on ph must be considerably outweighed by the amount of calcium present, and there should be more sulphate than chloride. For top fermenta tion ales some calcium is necessary even in the absence of carbonate, but lager of the pilsner type is best made with an entirely soft liquor. The harsher flavoured Dortmund lagers, however, have somewhat the same requirements as English pale ales. Mild ales. These will also benefit by the removal of carbonate or its counteraction by calcium, but less calcium is required than in pale ales and its place can partly be taken by sodium. In this type of beer there should be more chloride than sulphate. purity, and must be free from any pronounced taste and from more than the merest trace of oil. Liquor Treatment In order to work out any form of liquor treatment, it is first important to know what is in the liquor to begin with. This informa tion is contained in an analysis of the liquor, which will probably set out the mineral constituents in two forms: first, as the quantities of the ions present, and then as the quantities of the different salts needing to be added to pure water to get these ions in solution (remembering that the bicarbonates are conventionally expressed as carbonates: it makes no difference to the calculations). The two forms of expression TABLE Typical Water Analysis I As ions Conventional representation as salts Ion Grains per gal. p.p.m. Millivals Salt Grains per gal. p.p.m. Sodium, Na Magnesium, Mg Calcium, Ca Nitrate. NO, Chloride, Cl Sulphate, SO, Carbonate, CO,. 0-5 0-5 50 1-0 30 30 4-5 7 7 70 14 42 42 63 03 00 3-5 0-23 M0 0-88 21 NaNO, NaCl MgCl, Cab, CaSO«CaCO, 1-40 0-28 200 2-08 4-25 7-50 10-6 3-9 28-0 28-1 69-5 1050 Dark, strong beers and stouts. Here, the absence of carbonate is less imperative, and the Munich type of lager positively demands some carbonate in the liquor. Calcium is of minor importance and most of the salts may be present as chlorides. General. Excessive amounts of calcium in any brewing liquor could lead to fermentation and fining troubles and to thin-flavoured beer. Most brewers like to ensure the health of their yeast by having a small amount of magnesium in the liquor; but it must be small, and in no case should appreciably exceed 3 grains per gal. or the flavour of the beer will suffer. The liquor should contain not more than about 3 grains per gal. of nitrate, 10 of sodium, 3 of silicate, 25 of chloride, and no more of the trace elements than are permitted in a satisfactory drinking water. It should also reach the standards of drinking water in regard to bacteriological are necessary because, although we may think of the effects of ions, we have to add these ions as salts. The concentrations may be expressed as parts per million or grains per gallon, and since brewers are more familiar with the latter, that mode of ex pression will be used here. All that has to be done to convert grains per gal. into p.p.m. is multiply by 14. There is a third way of expressing the amounts of the different ions present, and that is as milli-equivalents or, more simply, millivals. The number of millivals of an ion is obtained by dividing its concentration in p.p.m. (or 14 times its concentration in grains per gal.) by the chemical equivalent weight of the ion. This form of expression is useful because it shows the balance of the ions in the liquor, since 1 millival of any ion is equivalent to 1 millival of any other ion. To give an example, suppose the liquor

Vol. 61, 1955] comrie: brewing liquor treatment 205 contains 8 grains per gal. of calcium and 12 grains per gal. of carbonate; it is not until these are expressed as millivals (Ca =s 8 X 14/20 = 5-6; CO3 m 12 X 14/30 = 56) that it is seen that these amounts are chemically equivalent. A typical analysis is set out in Table I. Removal of unwanted minor constituents. The only way to get rid of excessive amounts of nitrate, sodium and chloride is to pass the water through a bed of cation-exchange material and then through a bed of anionexchange material. If excess of silica is present, a special silica-exchanger will also be necessary. The result is a pure water which could be treated for any desired type of brewing. The treatment is rather more costly than others and, if it were used, care would need to be taken that the anionexchanger was not soluble in the water to any significant degree, as traces of this type of substance have been found to interfere with saccharification. Similar treatment would be required by a liquor containing an excessive amount of any of the heavy metals, but fluorine can be reduced to a safe limit by passing the liquor through a bed of hydrated calcium phosphate and apatite (a phosphatic mineral), and excess of iron and manganese removed by passing the aerated water through a special bed of granular material. Small amounts of iron are generally removed satisfactorily in the course of the removal of carbonates by lime. A high content of magnesium can be reduced by using a particular modification of the lime treatment known as the split process. In this, the whole of the lime required is added to a portion ($- ) of the liquor only, which is then filtered or decanted free of precipitate and blended with the remainder of the liquor. Traces of oil and objectionable flavours, such as those sometimes produced as a result of chlorination, may be removed by passing the liquor through a bed of granular activated carbon. Treatments of this nature are required exceptionally and are therefore of minor importance. By far the most neces sary form of treatment is that for the removal of carbonate. Removal of carbonate. This can be done in four ways. (1) By boiling: Ca(HCO8)a HaO + COa f (2) By treatment with acid, e.g.: Ca(HCO3)2 + HaSO4 CaSO4 + 2H2O + 2C02t (3) By treatment with lime: Ca(HCO3)2 + Ca(0H)2 *2CaCOs + 2H2O (4) By treatment with lime and acid salt: as (3), then, e.g.: Ca(OH)2 + 2CaCO3 + 4KHSO4 > 2CaSO4 + Ca(HCO3)s + 2K2SO4 + H2O Boiling can, in favourable circumstances, reduce the content of calcium carbonate to about 4 grains per gal., which is sufficiently small to be harmless, but it will not remove magnesium carbonate to the same extent and has no effect on sodium carbonate. This method is therefore only effective if most of the carbonate is combined with calcium, although the addition of calcium chloride equivalent to the sodium carbonate (0-86 pints of a saturated solution per 100 brl. for each grain per gal. of sodium carbonate) and the magnesium (3-8 pints per 100 brl. for each grain per gal. of magnesium) before boiling will increase the removal of carbonate. With the present cost of fuel the method is expensive, and it can only be recommended where the microbiological purity of the liquor is questionable. Various devices have been tried to make it more economical, and much improvement has been gained by heating the liquor under pressure with agita tion. In one case heat is conserved by re generation, and no more is used than would be needed to heat the liquor to mashing temperature, but special plant is required and there is a certain amount of scale formation. Acids used in the second method have included sulphurous, sulphuric, hydrochloric, phosphoric and lactic. Sulphurous and hydro chloric acids are difficult to handle because of their fumes, and hydrochloric may raise the chloride content of the liquor too much, whilst phosphoric and lactic acids produce salts which increase the buffer content of the wort and tend to keep the ph of the beer above its best range. Acidification with sulphuric acid is therefore the form most favoured, although lactic acid has been found to give good results with beers of delicate flavour, such as pilsner, and where increase in sulphate is undesirable owing to the large amount already present. Acidifi cation may be partly or entirely carried out

206 comrie: brewing liquor treatment [J. Inst. Brew. by means of an acid salt such as the bisulphate of sodium or potassium, so avoiding the difficulty of handling a strong acid and lessening the risk of over-acidification. The amount of acid or acid salt to add is best found by titrating the water with a standard acid solution (any strong acid will do), using as indicator methyl orange, which is unaffected by the carbon dioxide set free. It is not advisable to neutralize the alka linity of the water.completely to this indica tor or the beers will be thin and lack bitterness, and because too low a ph in the mash tun has been found to increase the extraction of unstable protein and may cause "gushing" in bottled beer due to over-active proteolysis. According to D. McCandlish & G. Hagues (this Journal, 1929, 61), if 200 ml. of the raw water are titrated with 0-1-N acid, the correct ph is given at a point 4 ml. short of the methyl orange end point. On this basis the weight of concentrated sulphuric acid required for 100 brl. of liquor will be 0-9 (T 4) lb., or approximately 0-05 (T I) gal., where T is the titration value in ml. An equivalent weight of good commercial potas sium bisulphate would be 2-9 (T 4) lb., and of sodium bisulphate, 2-45 (T 1) lb. After treatment, the water should be checked with bromothymol blue indicator, which should show a bluish-green colour. The acid must be drip-fed into the water, but the acid salts can be added at the same time and in the same way as any hardening salts. Sulphuric acid treatment is applicable to any types of water except those containing more than a trace of iron, and possibly those which already contain much sulphate, but it is less suited than other treatments to waters con taining alkali carbonates, since it converts these into the objectionable alkali sulphates. The carbon dioxide which is disengaged can have a corrosive action on the iron of the liquor tank, and in order to get rid of it effectively the treated liquor needs to be brought to the boil. Treatment with lime can take the form of simple addition of the lime, or it can be carried out in conjunction with some means of removing excess lime and adjusting the ph of the water to neutrality. The simple addition of enough lime to precipitate all the carbonate will leave the water too alkaline, and not more than about 0-5 lb. of calcium hydroxide per 100 brl. for each grain per gal. of carbonate in the liquor is advisable. The efficiency of the treatment \vill depend on the amounts of free carbon dioxide and mag nesium carbonate in the liquor, though it is improved by the prior addition of calcium chloride as in the boiling process. If much magnesium carbonate is present, the split method of lime treatment described earlier is advisable. The speed and efficiency of separation of the precipitated calcium car bonate can be greatly increased by certain mechanical means. In one of these, the raw liquor-lime mixture is continuously stirred in contact with previously precipitated calcium carbonate while it passes down a central vessel, then rises with diminishing velocity in a surrounding vessel so that by the time it enters the effluent pipe at the top it is perfectly clear. The time taken over the process is quite short, but the system is a continuous one requiring special plant and is only suitable where large amounts of treated liquor are needed. The same principle is used in a rather simpler plant, in which the raw liquor and the lime emulsion are pumped under pressure into the bottom of a tall conical tank con taining a bed of catalytic granules. The precipitated calcium carbonate is largely retained on the surface of the granules, which are kept in constant agitation by the swirling motion of the entering water, and the slight remaining turbidity is removed by nitration of the liquor after it has emerged from the top of the tank. This process is also a con tinuous one. The problem of obtaining removal of carbonate without leaving the liquor alkaline has been tackled in two ways. In one, the carbonate is precipitated with a small excess of lime and allowed to settle for 1-4 hr. The liquor is then passed through a bed of car bonaceous calcium zeolite, which removes the excess lime, and any remaining calcium and magnesium carbonates, down to about 1 grain per gal., leaving the water with a ph of 6-9-7-1. Other salts of calcium and mag nesium are unaffected. In the other process, the carbonates are precipitated with lime; any excess of lime, together with a fraction of the precipitated calcium carbonate, is neutralized with sodium or potassium bisul phate. The process is simple and it can, if necessary, be carried out in the one hot liquor tank equipped with means for mixing-in the reagents and holding back the precipitated calcium carbonate. It has the advantage of

Vol. 61, 1955] COMRIE: BREWING LIQUOR TREATMENT 207 coping easily with slight variations in raw liquor composition. The liquor is preferably heated to about 120 F. to accelerate the reactions, and the necessary calcium chloride is added as in the boiling process, followed by 0-7 lb. of calcium hydroxide per 100 brl. for each grain per gal. of carbonate that the water contains. Good mixing is an essential feature of the process, and if the tank is heated by naked steam the steam nozzles can be arranged to secure this. If steam coils are used, compressed air will do the rousing cheaply and efficiently. One jet of about J-in. orifice will serve a circle of about 8 ft. radius irrespective of the depth, and jetting for 1 min. is sufficient, so that the air is best taken from a compressed-air storage vessel in order to keep the compressor as small as possible. After a few hours, most of the carbonate will have formed a firm sludge, and the longer the time that can be allowed before adding the bisulphate the more complete will be the deposition, the less the amount of bisulphate needed, and the smaller the amount of carbonate left in the treated liquor. This is because the bisulphate reacts not only with the excess lime but also with the carbonate left in suspension, converting it into bicarbonate. The quantity of bisul phate required to bring the liquor to the correct ph can be found by titration of the lime-treated water with 0-1-n acid, and the calculation is the same as that given for the acid treatment. Alternatively, it can be calculated from the titration value of the untreated water and the amount of lime added. Deposition will be facilitated if a small amount of calcium carbonate is left in the tank from the previous treatment, but it is retarded by traces of zinc in solution, and galvanized metal is inadvisable inside the tank even in the absence of copper. If two tanks are available the liquor is transferred to the second one before adding the bisulphate, and whether one or two tanks are used care must be taken that the deposit remains undisturbed when the water is drawn off. The lime used should be a good quality hydrated lime powder, free from iron and silica. One final word about decarbonation. It should be applied to all the liquor mashing, sparging, make-up and dilution. The addition of salts. It will be remem bered that the reason for removing carbonate was in order to reduce the ph of the mash, and that the same effect could be produced by the addition of calcium ions. Where the carbonate content is moderate and there is no large concentration of other salts, the carbonate, instead of being removed, may be offset by adding calcium salts mainly sulphate if the liquor is to be used for pale ales and partly chloride if it is required for mild ales. The quantity of calcium needed for this purpose is such that the total calcium in the liquor after the addition is at least twice the equivalent of the carbonate present, and preferably more if some of the carbonate is combined with magnesium (2 grains of calcium are equivalent to 3 grains of car bonate). Where it occurs naturally, as in the well waters of Burton-on-Trent, such a "neutralization" of carbonate by the total calcium present gives good results, but on the whole it is preferable to remove the carbon ate, or at least reduce it to about 4 grains per gal. as calcium carbonate. There is one important difference between the two methods of dealing with carbonate. "Neutralization" of carbonate by adding calcium will cause more phosphate to be precipitated than will a decarbonated liquor, with its lower total calcium, and so will tend to give beer of lower ph. The actual amount of any type of salt required in a liquor is largely a matter of opinion, based on the character of the beer that is being brewed. In the days when gravities were commonly 1045-1055, as much as 20 grains per gal. of calcium would be demanded for pale ales, but the amount fell to about half this figure with the diminution in gravities. With a better realization of the importance of adequate flocculation during boiling and cooling, there has been a tendency recently for the amount to rise again, and it now varies between 7-14 grains per gal. for pale ales, 4-8 for mild ales, and 2-4 for stouts, these quantities being additional to what is required to offset unremoved carbonate. In pale ales the larger proportion of this will be as sulphate, but in mild ales it will be in more nearly equal amounts of chloride and sulphate, since the dry, bitter flavour of pale ales requires the sulphate/ chloride ratio to be about 2/1, and the full, sweeter flavour of the mild ales needs a ratio of about 2/3. This is a generalization, and where the taste is for a rather sweeter pale ale than in the past, the ratio should probably be nearer 3/2 than 2/1. The soft flavour of

208 comrie: brewing liquor treatment [J. Inst. Brew. stouts requires little, if any, sulphate, and the calcium will be present principally as chloride. Sodium chloride is responsible for part of the chloride in mild ales and stouts, and to a smaller extent in pale ales, where a large proportion would bring out a somewhat harsh hop flavour. If the liquor contains any appreciable amount of sodium as sulphate, it is advisable to replace part of what would normally be added as sodium chloride by cal cium chloride or potassium chloride. It is open to doubt whether any magnesium need be added, since the malt probably supplies all that the yeast requires, but many brewers prefer to make sure by bringing the mag nesium content of the liquor up to about 0-5 grains per gal. This may be added in the form of sulphate but is more often added as kainit, which usually contains 3-5% of magnesium. A difficulty attendant on the use of kainit is its liability to vary in com position, and the fact that it complicates matters by adding at the same time con siderable amounts of calcium sulphate and sodium and potassium chlorides. This is all right if the brewer wants these salts to be added, but it gives him little control over the proportions. These recommendations for the composition of brewing liquors are sum marized in Table II. As a rule, the amount of each salt used (within its range) is proportional to the gravity of the beer. When it is remembered that the mash tun reactions take place in a mash of fairly constant concentration whatever the original gravity of the beer is to be, it might be suggested that the actual mashing liquor should always receive the maximum treat ment, the final concentration for flavour in the beer being adjusted by adding less of the salts to the sparge and make-up. All the liquor should still be decarbonated. The calcium and magnesium salts are added to the hot liquor tank and well mixed in during the heating of the liquor. Where decarbonation is being carried out in the one tank, the salts are best added beforehand to avoid disturbing the carbonate sludge later, and in any case the calcium salts are best added first as the additional calcium will aid the decarbonation process. Calcium chloride, of course, is always added before decarbona tion. Sodium chloride is usually added to the copper, since its effect is one of flavour only, apart from the minor activities during mashing and sparging which have been referred to. Table III shows the forms in which the commercial salts are added to the liquor or wort, the amounts in lb. per 100 brl. which are equivalent to 1 gr. per gal. of the TABLE Compositions of Some Brewing Liquors* II Pale ales Mild ales Stouts Ions Salts Ions Salts Ions Salts Ca Mg Na.. 7-14.. 0-5.. 1-2 CaSO4 CaCl, MgSO, nsci.. 15-7-31-6.. 6-9-13-7.. 2-6.. 2-6-6 Ca Mg Na.. 4-8.. 0-6.. 2-4 CaSO4.. 6-10 CaCl,.. 7-14 MgSO4.. 2-S NaCl.. 5-10 Ca Mg Na.'. 0-5.. 2-8-4-8 CaSO4 CaCl,.. MgSO4 NaCl. 5-5-11 2-5 7-12 * All figures are as grains per gal., and the amounts of calcium are additional to what is required to offset any carbonate present. TABLE III Equivalence Values op Some Commercial Salts Anhydrous salt Commercial form Quantity equivalent to 1 grain per gal. of the anhydrous salt Proportions by wt. of ions in the salt CaSO, MgSO4 NaCl CaCl, Gypsum, CaSO..2H,O MgSO4.7H,O Kainit NaCl Solution, sp. gr. 1350-1360 0-66 lb. per 100 brl. 1-06 lb. per 100 brl. ca 2-5 lb. per 100 brl. 0-61 lb. per 100 brl. 0-82 pints per 100 brl. Ca 0-3 :SO4 0-7 Mg 0-2 :SO4 0-8 Na 0-4 :C1 0-6 Ca 0-36 : Q 0-64

Vol. 61, 1955] comrie: brewing liquor treatment 209 pure anhydrous salts, and the proportions by weight of the ions in each salt. The liquor analysis given in Table I may be used to provide an example of the form of treatment that might be given for a light pale ale brewing. Since the amount of carbonate is not large, it is decided not to remove it but to offset it by the use of additional calcium. Since 3 of calcium are equivalent to the 4-5 of carbonate present, and the liquor contains no magnesium car bonate, the amount of calcium needed to offset the carbonate may be taken as 6. In addition, for hardening purposes we want, say, 8 of calcium, making a total of 14 of calcium. There are 5 of calcium already in the liquor, so 9 require to be added. If 6 of these are added as sulphate and 3 as chloride we shall be adding 7/3x6 = 14 of sulphate (SO4) and 16/9 x 3 = 5-3 of chloride (Cl), bringing the totals of sulphate and chloride to 17 and 8-3 respectively. The addition of 3 of sodium chloride will increase the chloride to 10-3, which makes a suitable balance of sulphate and chloride. Now, to translate these amounts into terms of commercial salts. Table III gives, in round figures: 6 of calcium as sulphate = 13 lb. gypsum/ 100brl.(6 x 10/3 x 0-65); 3 of calcium as chloride = 7 pints solution/ 100 brl. (3 X 100/36 X 0-82). These are added to the liquor. 3 of sodium chloride = 1-5 lb. salt/100 brl. (3 X 0-51). This is added to the copper. Conclusion The scheme of brewing liquor composition which has been put forward is intended only as a general guide. There will be those who, while agreeing with it in principle, would disagree in certain details. Would it not be better, for instance, to decarbonate all liquors and add alkali to the copper wort in the case of dark beers? Any such scheme as this can only be used as a basis for experiment. A brewer should not expect all his troubles to vanish and his beers to be transformed as a result of adopting a particular kind of liquor treatment, though proper correction of an unsuitable liquor will start him well on his way. He will have to proceed by trial, remembering that any change may have unforeseen repercussions; thus, the removal of carbonate was found to cause trouble in one particular brewery because it made the strain of yeast that was being used flocculate too early. The surest indication of the brewer being on the right path is the ph of his worts and beers about 6-2 for first wort, rising to about 5-6 in the later spargings; about 5*0 for boiled wort, and 3-9 to 4-1 for beer and the flavour and character of the beers. It must be remembered that the beers are influenced more by the balance of the ions in the liquor than by the amount of any one salt. And although the brewer has to feel his way forward by trial towards the beer of his choice; any adjustment in liquor com position that it may be decided to try can be planned with understanding when it is known what contribution each ion is making to the whole.