Techniques learn how to: * Predict in advance how a mash will come out stinguish between high-quality and inferior brewing products + Evaluate pitching yeast * Compare brewing procedures and determine their particular effects. on flavor “A fascinating (and, sometimes, slightly controversial) insight into many informative aspects of brewing materials ... plus a host of brewing basics. A most enjoyable, readable, and stimulating addition tovthe art {and science) of brewing!” Dr. David S. Ryd Brewing, Research, and Quality Assurance, Miller Brewing Company “All aspects of the brewing process are covered—from raw mataiial acquisition to beer packaging and product evaluation. The appendices which consider numerical aspects of the process, are particularly This volume is recommended to both craft- and homebrewers.” Professor Graham International Centre for Brewing ai Heriot Watt University, Edinbuf ysis of Brewing Techniges is thought-provoking crore ervey of ball moi asa s learning, experimentation, an nail ‘an imaginative and readable: aw that will be’ a p advanced homebrewer.” A Browers Publications Book Brewing -Analysis Brewing Techniques |Analysis Brewing TechniquesBrewers Publications, Division of the Association of Brewers PO Box 1679; Boulder, CO. 80306-1679 (G03) 447-0816; Fax G03) 447-2825 © 1997 by George J. Fix and Laurie A. Fix [All sights reserved. Except for use in a review, no portion of this book may be reproduced in any form without written permission of the publisher. Neither the author, editors, nor the publisher assumes any responsibility for the use or misuse of information contained in this book. Printed in the United States of America W98765 4321 ISBN 0937381-47.0 Library of Congress Cataloging-in-Publication Data Fix, George J, 1939 Analysis of brewing techniques / Geonge Fox and Laurie Fix. Pp. cm. Includes bibliographical references and index. ISBN 0:937381.47.0 alk, paper) 1. Brewing—Handbooks, manuals, ete. 1. Fix, Laure. 1. Tide, ‘res70rs76 1997 63'3—de2 97-10923 ‘Technical Editor: Scott Bickham Book Project Editor: Theresa Duggan Copy Falitor: Terri Bates Eyden Assistant Editor: Kim Adams Interior Designer: Vicki Hopewell ‘Cover Designer: Stephanie Johnson Cover photography by Michael Lichter. Flasks courtesy of I's Pizza, Grill, and Brewery, Boulder, Colorado. Unless otherwise noted, all interior photographs are by George and Laure Fix. Contents Introduction. Wort Production Malt Cereal Grains Water Mashing Systems Hops. Wort Boiling, Yeast Strains. StOTABE ween Pitching Rates .. Propagation Oxygenation of Chilled Wort Slant Preparation and Inoculation, Evaluating Pitching Yeatt.... Fermentation and Maturation ... Fermentation By-products Fermentation Vessels 4. Cleaning and Sanitation .. Cleaning Agents Sanitizers . 5. Packaging Beer ining Agents Filtration Carboration Bottling Beer ..6. Evaluation of Beer Basic Data... : : 141 Microbiological Analysis, eee Evaluating Flavor «10.00 - 146 Improving Competitive Judging ° er Appendix A: Basic Units . Appendix B: Gravity Units .. Appendix C: pH Basics ‘Appendix D: Prediction and Measurement of RDF Appendix E: Measurement of CO, and 0; vw Contents Introduction ‘Welcome to An Analysis of Brewing Techniques. Our original intent ‘was to write a sequel to Principles of Brewing Science and title it Principles of Brewing Science MU: Practical Considerations. Principles of Brewing Science considered the chemical compounds relevant to beer and the reaction mechanisms associated with them. ‘The sequel would explore the practical implications of those com- pounds in terms of finished beer quality. However, as our research progressed, it became increasingly clear that such material would be best organized around brewing materials and the procedures used to process them. The end result was An Analysis of Brewing Techniques—a practical study of brewing materials and procedures. We quote Principles of Brewing Sefence at several points in this book; however, this was done (o give the reader a point of departure for a more theoretical exploration of brewing, The material refer- enced in Principles of Brewing Science is not essential to under- standing the practical issues addressed in An Analysis of Brewing Techniques. Thus, the two books arc independent of one another and can be read in either order. ‘The analysis of brewing materials is based on key quality indices, that enable the brewer to distinguish between high-quality products and inferior ones, This information is discussed in the Malt, Cereal Grains, Water, and Hops sections of chapter 1. An Analysis of Brewing Techniques also compares different brewing techniques (chapter 1, ‘Mashing Systems and Wort Boiling, and chapters 3, 4, and 5). Introduction wilChapter 2 is about yeast, and itis by far the most crucial chapter of the book. It touches on all aspects of yeast management, including available strains, yeast propagation and storage, and, most importantly, how to evaluate pitching yeast. The five appendices deal with the measurement of key brewing variables. ‘The procedures discussed ate of great practical impor- tance; however, they have been separated from the text in order to ‘maintain continuity. It is our view that brewing heer is an art—always has been and always will be. This belief defines our aspications for beer quality. An analytical approach has value in that it helps us reach our brewing goals, such as attaining a particular flavor profile, We feel that this is as true for unorthodox brewers of idiosyncratic beers as it is for con- ventional beer brewers. All brewers (novice or experienced) spect late and theorize about phenomena they observe in brewing. The theme of An Analysis of Brewing Techniques is that progress is most. uniform when such considerations are based on data, H. Loyd Hind, in his classic book Brewing—Science and Practice, notes that before there were systematic measurements, brewing was at best haphazard. He quotes Marshal Saxe (who was referring to military science in the nineteenth century): "It is a sci ence so obscure and imperfect that custom and prejudice, confirmed by ignorance, are its foundations and support, with sacrosanct dogmas no better than maxims blindly adopted without any examination of the principles on which they were founded (Hind 1950).” An Analysis of Brewing Techniques was written to help brewers avoid such circumstances by taking an analytical, data-based approach to brewing. One reason brewing is so fascinating is that no one can completely master it. As our knowledge of brewing grows, so do the mysteries associated with this ancient art. ‘While we hope this book will contribute to the reader's insight into brewing, further reading is definitely encouraged. With this in mind, we carefully selected and used the references cited within, focusing on their accuracy, relevance, and readability. George and Laurie Fix vil ‘An Analysis of Brewing Techniques Wort Production MALT Barley Malt Malted grains are the basis of most beers, and, indeed, beer as we know it today would not exist without the enzyme systems that form, when grains are malted. A variety of grains are malted, but for most beer styles, barley is the grain of choice. The two major issues asso- ciated with malted barley are the variety and the malting process. ‘When we compiled research for our book on Viennese beers, we came to reslize the important role that barley variety plays in beer Quality (Fix and Fix 1991. Itis as important to brewers as grape vari- ety is to winemakers, The problem for brewers is that barley varieties cultivated for malting and brewing are constantly changing, For example, in 1985 Triumph, grown in northern and western Europe, and Aura, grown in West Germany, occupied the largest growing areas (Brauwelt 1986). Yet, by 1993, Aura and Triumph had virtually disappeared (except in France), and Alexis had become the favored spring twoxow barley in England and Germany (Brauwelt 1993). And, undoubtedly, things will change in the future, Barley Protein Level Fortunately, a number of attributes are common to varieties having superior malting and brewing properties. The most important is bar- ley protein level. De Clerck stresses this point and cites 9-11% as the preferred range (le Clerck 1957). Although introduced over four: decades ago, this criterion is still widely accepted. Common beers Wort Production 1have been successfully produced from malt in the cange of 11-12%; however, malt whose protein levels exceed 12% are generally regarded as suitable only for cattle feed. Table 1.1 gives examples from the 1993 crop. Table 1.1 Barley Protein Levels in 1993 Crop Variety Type ‘Country Percent Protein Alexis Spring eworow [Alexis Spring woo zp Splng wor Paint Spring ico Winter worow Sprig woeow Une ates Spring cow Dnited Sates ———S=«RO-IBS “tn the Une Kingdom, pecent prota xometines expressed a percent nitrogen. Te commer sion factors 6:25, Fr extmp, 10% poi ele to 10/625, which equals 1.6% oye [As table 1.1 shows, domestic malt in the United States is typically high in protein Gcheer 1990). This is traditional and closely related to the widespread use of unmalted cereal grains, such as corn and rice. High protein malt usually has enough enzymes to convert its ‘own starch as well as the starch of the unmalted adjuncts. Since the latter does not contribute soluble nitrogen to wort, cereal grains malt, in effect dilutes what would otherwise be excessive wort nitrogen levels. Nevertheless, as all-malt, craftbrewed beers and analogous seasonals grow in popularity, there will undoubtedly be an increased demand for domestic malts that have protein levels more in line with de Clerck’s criterion, Thus, we are hopeful that future domestic malt ‘will match Europe’s best malts in terms of brewing quality. The Malting Process ‘The malting process is also important. Indeed, a top-quality barley malted at one plant may be significantly different from the same barley malted at another plant. The entire process starts with the steep, where the barley's moisture level of a moderately modified malt is brought to 38-42% of kernel weight. For problematic barley, 46-47% is sometimes used, but this invariably produces mat with inferior brewing characteristics. Steeping takes from eighteen to thirty-six hours 2 ‘An Analysis of Brewing Techniques ‘The next step in the malting process is germination, which is, done either on a germination floor (floor malting), as shown in figure 1.1, or in a germination compartment (pneumatic malting). It is dur- ing germination where serious modifications take place, the most important being; + Development of the grain’s enzyme systems + Breakdown of complicated proteins into simpler structures + Breakdown of complicated starch into simpler carbohydrates “The degree of modification depends on the type of malt being produced as well as on the brewer's requirements. However, what is important for any type of malt—well modified ot otherwise—is that germina- tion is uniform. Ideally, pulp kernels of a uniform size should result. A number of shortcuts have been proposed to cut down on the time and cost of getminating grain, such as using higher moisture levels in the steep or acding hormones like gibberellic acid to the germinating grains. Adding hormones is banned in many countries but used in others, The danger with such stimulants is that uniformity of modifi- cation, crucial to malt quality, suffers in forced, rapid germinations. ‘A variety of time and temperature programs have traditionally been used in the vegetation process (Le., steep plus germination), as table 1.2 shows. Two general trends regarding time and temperature are (I) the lower the temperature, the more uniform the modification, ee Figure 1.1. Floor malting, Photo courtesy of Crisp Malting ‘Wort Production sbut it takes longer to reach the desired modification level; and (2) the higher the temperature, the required time decreases, but so will the ‘modification uniformity. Table 1.2 Time and Temperature Programs ‘Malt type. Method Modification Temperature Time (days) Tleale Foor igh 29 FEO 0 Pikexer Moor ——sNomal—— 150 Extract—The total complex of carbohydrates in wort. Malt analy- sis sheets typically ist the malt yield, the fraction of a grain’s weight that is converted to extract. This is done in a standardized mash that is quite dilute (1 kilogram of malt per 6 liters of water) and is not sparged. ‘This figure, therefore, tends to represent the maximum extract, which is neither possible nor desirable to achieve in pra cal brewing situations. In addition, malt yields vary—malt from two- row barley typically has higher yields than sixrow barley—hence they rarely give insight into starch modification. A more usefull index of starch modification is the fine/coarse grind extract difference using a standardized mash. Low values of fine/course grind differences correlate with high levels of starch mod- ification. Typical values are given in table 1.6. 6 ‘An Analysis of Brewing Techniques Table 1.6 Fine/Coarse Grind Extract Differences, Starch Modification ‘ndeemodkeaton Unlike protein modification, there is almost universal agreement that high levels of starch modification are desirable, As a general rule, ‘most brewers are reluctant to use malt whose fine/course grind dif- ference is higher than 2.2% (de Clerck 1957). ‘The Hartong index, sometimes used to measure starch modifica- ion (particularly on the Continend, is the extract ia a L13 °F (45 °C) mash expressed as a percentage of fine-grind extract, Most brewers prefer a Hartong value of at least 38%, and values in the 38-42% range tend to correlate with good starch modification, Kiln Temperatures—Influence the three major malt properties the strength of its enzymes, its coloring potential, and its sulfur con- tent. The strength of a malt’s enzyme system is represented by an index called diastatic power (DP), which is usually expressed in degrees Linter (Fix, Principles of Brewing Sclence 1989). In general, diastatic power decreases as the kiln temperature increases. There is also a varietal effect: a high protein malt from six-row barley tends to have a higher diastatic power than a low protein malt from qwo-row barley. On the other hand, color, expressed in degrees Lovibond (L) (or Standard Reference Method (SRM, increases with increasing kiln temperatures. Some sample ranges are listed in table 1.7. ‘A malt’s sulfur content is strongly affected by kiln temperature and, in general, it decreases as temperature increases. Thus, this malt Table 1.7 Diastatic Power and Color Malt Type DP (Liner) Protein Level Domestic show 5185 ~Domesictworow ~*dSISSSCSCSC*CS ARS i Ec Te ae oe Munich as ~~—«dO AS [ate ate 0-2 90-110 Wort Production‘parameter becomes a primary issue only for pale malts. Malt has many ‘components that contribute to the sulfur flavor in beer, but dimethyl sulfide (DMS) is the component that is generally monitored, This ‘compound is created in wort production from malt precursors, and it is the DMS precursor level of malt that is most relevant (Fix, Principles of Brewing Science, 1989; Fix 1992, 15:3). Table 1.8 gives ‘ypical values Table 1.8 Malt Sulfur Levels Malt type Level (micrograms per gram of mat) Paleale | Moder Pisenae Itis important to emphasize that sulfur flavor in beer is very subtle, Selective gram-negative bacteria can create large amounts of DMS, leaving an unpleasant cornlike/cooked vegetable taste in beer. Domestic six-row malt also leaves a cornlike flavor, unless special steps are taken in wort production to reduce DMS levels. In striking contrast, select German pale lagers often have a highly pleasant ‘malty/sulfury flavor profile. A variety of sulfur components contribute to this, DMS being just one, Wheat Malt ‘Wheat malt is completely different from barley malt. First, protein lev- els in wheat malt are typically high. It is not unusual to find values in. excess of 13% (Oloff and Piendl 1978). Certain varieties grown in spe- cial microclimates result in wheat malt protein levels in de Clerck's range of 9-11%, but these are the exception and not the rule (de Clerek 1957). In general, soft white wheat varieties have lower pro- tein levels than hard red wheats. Values for fine-grind Gaboratory) extract are usually higher than those of barley malt. For the most part, this is a consequence of wheat malt not having husks. Asa result, car- bobydrates occupy a higher fraction of grain weight. This extra yield. is rarely seen in practical brewing situations, and many brewers find, they get lower yields from malted wheat than from malted barley. ‘One reason for this is that problems are encountered in malting ‘wheat. Typically, cytolytic modification is low, hence viscosity is high. 8 ‘An Analysis of Brewing Techniques Values in the range of 1.8-1.9 mPas are not uncommon (as compared. with 1.4-16 for malted barley) (Oloff and Piendl 1978). In contrast, proteolytic modification is high, and resulting Kolbach values are gen- erally over 40%, In order to avoid the vicissitude of excess protein modification, wheat malt is usually processed in such a way that car- bohydrates are undermodified. Differences in fine/coarse grind extract are often above 2.5%, which tends to relate to low yields in practical mashing systems. Roasted Malt Roasted malts play an important role in brewing, both for their cok coring potential and for their special favoring. In the roasting process, the enzyme systems of these malts are totally deactivated. As a con: sequence, they are often called adjunct malts. The market is saturated with trade names, thus the simplest way to classify roasted malts is by the two ways in which they are produced: + Group 1: Green malt (germinated but not dried) + Group 2: Regular malt (germinated and dried) ‘The flavoring of these malts differs from pale malts, primarily because of the Maillard products formed during roasting. Maillard products are N-heterocyclic compounds, such as pyrazine and pyra- zole (Fix, Principles of Brewing Science 1989). According to Narziss, these compounds can vary in magnitude depending on the length of time the malt is roasted and on the roasting temperature (Narziss, Zymurgy 1993). Lightly roasted malts have a sweet toffee/caramel taste; highly roasted malts have a burt taste often accompanied by sharp/acid tones. Tables 1.9 and 1.10 provide data from Blenkinsop's, research (Blenkinsop 1991). Table 1.9 Group 1: Green Malt Malt Type Color (D Taste 1S. tight Caras) ey “Terecearamel| a aa Tovfecesramel | 30-40 Ganmetautty 70-80 Naty/roasted ‘| 100-160 ~Tateerborar———_—| Wort Production °Table 1.10 Group 2: Regular Malt Malt Type Calor ¢L) Taste ‘nbe, 30-40 Bice cece aie 0-500 Sigh harpac dy a Brown malt, used in traditional porter formutations, falls into the group 2 amber category (Bergen 1993). It is roasted with direct heat from becch, oak, or other wood fires. Malt used for Bambergstyle rauchbier is produced similarly. Maillard products formed during this, roasting process are highly complex and not yet completely defined, and the flavors they impart are equally complex. Brewers often overlook the limited shelf life of roasted malt ‘Their unique aromatic compounds, which can be used to great advantage with fresh malt, are the first to suffer in extended malt stor- age. This is particularly true of crystal malt, or caramel malt Unpleasant, cloyingly sweet caramel flavor tones are signs of deterio- sation. Diacetyl, even at low levels, will tend to amplify these nega- tive flavor notes, as will oxidation, CEREAL GRAINS ‘There are two basic reasons to use unmalted grains. First, cereal ‘grains (com, rice, ete.) tend to have the same carbohydrate structure a8 malt but are often cheaper; therefore, selected use of these grains ‘can save the brewer money. The second and more interesting reason, to use them is that while cereal grains do have a definite protein con- tent, they tend to make only minimal contributions to the wort’s pro- tein pool. Thus, cereal grains used in conjunction with malt give ‘extra strength (0 the finished beer without the satiating effects that result when only malt is used to gain extra gravity. Some world-class lagers and ales are produced this way (Fix 1994; Protz, Real Ale 1991). Gelatinization and Saccharification ‘The classic concern about unmalted grains is the tear of incomplete .gelatinization and sacchatification, Narziss correctly points out that if these defects are present, the finished beer can have “broad bitter- ness” and a “raw character” (Narziss, Brauwelt 1993, 3). Flaked grains 10 ‘An Analysis of Brewing Techniques are gelatinized by passing the grains through hot rollers at 185 °F (85 °C), This is nearly ideal for grain gelatinization, and the concerns ‘expressed by Narziss generally do not apply to these grains. In par. ticular, they are readily converted in normal mashes at usual sacchar- ification temperatures (Ke., 140 °F [60 °C] or above). Yields in excess of 80% can be expected with these grains. Raw grits require a cooker mash in order to achieve gelatiniza tion, Boiling times vary from fifteen minutes for refined corn starch and refined grits up to forty-five minutes for common grits. A 10% ‘malt charge is typically used, and cooker mash thickness is usually ‘maintained with feed water during boiling to compensate for evapo- ration and water uptake by the grains. Narziss also stresses the importance of proper mashing proce- dures to avoid incomplete gelatinization and saccharification Warziss, Brauwelt 1993, 3). The best results are obtained when the mash pH ranges from 5.5-5.6. Lower values typically lead t0 high ‘wort viscosity Storage ‘The oil content of cereal grains isa concer, more so for corn products than for rice. Oil can become rancid during extended andor high:tem- perature storage. Today, products are usually produced in such a way that the oil content is below 1%. Brewers should request this informa tion from suppliers to avoid storage problems. It should also be noted that high of levels in cereal geains generally lead to inferior beer foam. Cereal Grain Types A number of unmalted cereals are used in brewing, the most com- ‘mon being flaked maize, refined corn grits and corn starch, corn grits, rice grits, unmalted barley, wheat adjuncts, and rye adjunets Flaked Maize—Although itis one of the most expensive grains, flaked maize is by far our favorite unmalted grain. This grain is prege- latinized and can be added directly to the mash. Since the flakes do not contribute to the wort’s protein pool, they are normally added once saccharification temperatures are reached. Flaked maize was once Widely use¢ by brewers in the United States, but its high cost has led to diminished use (Fix 1994; Nugy 1948). Flaked maize, however, is siill very pepular with brewers in Belgium and the United Kingdom where it is valued for the soft and subtle grainlike sweetness it imparts. We have obtained yields close to 80% trom flaked maize. Wort Production aRefined Corn Grits and Corn Starch—These products are pre- processed by wetmilling procedures and are essentially pure starch. As a consequence, yields from these products are often above 90%. They require cooking in order to achieve proper gelatinization, however boil times are brief, typically five 0 fifteen minutes. Dextrose is produced from refined comm grits and com starch by further processing, where the starch is degraded into glucose. When refined grits are mashed, they have a carbohydrate spectrum similar to that obtained from malt, Com Grits—Corn grits are the cheapest and most widely used adjunct in the United States. They are a by-product of the food indus- try and are produced by a dry-milling process. Typically, thirty to forty-five minutes of boil time are required in a cooker mash, and yields are similar to those of flaked maize. Rice Grits—This rather expensive grain has long been esteemed by US. brewers as the perfect adjunct. Rice grits tend to impart a clean, neutral flavor that does not interfere with malt flavor. These grains are produced the same way as table rice, that is, by separating. the starchy endosperm from the outer layers (husk, bran layers, and germ), The kemels are often broken in the milling process and are not used as table rice because of physical appearance. These “rejects” are normally sold as brewers’ grits. Shortgrain rice is preferred because medium. and long-grain varieties can lead to viscosity prob- Jems in cooker mashes. Even the shortgrain products are milled in the brewery to get a uniform, small granule size. For these, a fifteen ‘minute boil is suflicient to achieve gelatinization; longer boiling can Tead to viscosity problems. Yields are similar to those of flaked maize and com grits. Rice is also available in flakes, but rice flakes have not been as widely used as flaked maize. Unmalted Barley—This adjunct has long been of interest to brewers because of its similarity to malted barley. For example, unlike rice or com, unmalted barley definitely contributes to the ‘wort’s protein pool. Irhas been estimated that approximately 15% oF uunmalted barley protein is dissolved in the wort (Fix 1985). This is lower than the amount of protein obtained from malted barley around 35-45%), but much higher than the minimal amounts con- tributed by com or rice adjuncts. Experience shows that if unmalted barley is mashed with a suitably high concentration of malt, the pro- teins will be foam positive. Pregclatinized flakes are the easiest form to use. While they are ready for sacchariication, there is, nevertheless, the problem of barley 2 ‘An Analysis of Brewing Techniques gums, like betaglucan, that must be dealt with in mashing (Fix, Principles of Brewing Science 1989). The appropriate breakdown of betaglucans can be achieved with a low-temperature rest in the mash (preferably around 104 °F [40 °C}) where betaglucase activity is encouraged. The use of low protein, plump unmalted barley is also recommenced to avoid the excessive wort viscosities that result from barley gums Unmalted barley is sometimes roasted, so it has traditionally been an important component of many stout formulations. It is very simi- lar to highly roasted black malt, except that its flavoring tends to be more intensely sharp/burnt/acidic. Wheat Adjuncts—It is a wellknown fact that foam-building properties of wheat proteins are superior to those of barley. This effect has been exploited by brewers who use a small wheat charge to the grain bill to increase the foam quality of beer—a practice that is something of an art, AC too low a level, the increase in foam quality is generally too trivial to justify the effort. On the other hand, wheat proteins typically do not respond well to chill-proofing agents, so too high a level can result in a reduction of the becr’s resistance to chill haze. Decaces ago, malted wheat was widely used as a foam-building agent. However, given the high degree of proteolytic modification of ‘modern wheat malt, unmalted wheat is now preferred, In addition, unmalted wheat adds a discernible smoothness to finished beer. A striking example is Belgianstyle white beer, where ‘up to 50% of the grain bill consists of unmalted wheat (Lodahl 1994). ‘There are also several English-style amber ale formulations that use wheat, some with charges as high as 20%, With a wheat charge of 50%, it is important to use barley malts with diastatic powers above 120 °Linter; with a 20% wheat charge, diastatic powers of 90 *Linter are needed, Rye Adjuncts—Rye imparts a distinctively crisp and slightly spicy character to beer. Interest in the grain has been growing, which hhas led to a new beer style—American rye ale, This ale differs from roggenbier (German-style rye beer) in much the same way that American wheat ale differs from German weisse beer. American rye ale is typically made with a 10-20% charge of flaked unmalted rye, and it tends to have a dry, snappy finish. We find tra good deal more interesting than the average American wheat ale. Flaked rye has many of the properties of flaked barley. Tt deft nitely contributes foam-positive proteins to wort. Also, its betaglucan Wort Production 3content is high, hence a temperature rest near 104 °F (40 °C), or at east one well below 140 *F (60 °C), is recommended, Rye is also available as a malted grain, however, constituents derived from rye husks have a strong, spicy and bitter taste, so use caution’ Another advantage of flaked rye is that it is dehusked. WATER ‘There has always been something of a mystique surrounding beer and the water used to brew it, It has often been claimed that the histori- cally successful brewing centers gained their fame because of their ‘water. Even today, select brewers assert in their advertisements that, “it’s the water” that is responsible for the special qualities of their beers. But the fact is, most of the world’s best beers are brewed with ‘water that has been extensively treated. While today's watertreat- ‘ment programs lack the romance attributed to older brewing proce- dures, they do, without a doubt, offer a shorter path to consistent beer quality. We will review some of the major procedures used t0 produce brewing liquor (e., treated water that has been rendered suitable for brewing) ‘The first issue to address is water purity. Contaminates fall into four categories: (1) inorganic material, (2) organic material, () microbes, and (4) miscellaneous particulates. Heavy metals are the ‘greatest concem, Table 1.11 lists the maximum acceptable levels; ‘water not meeting these requirements is unsuited for brewing, Table 1.11 Maximum Acceptable Metal Levels in Brewing Water Element Concentration (mail) (0.050 Ted a (0.050 | “Manganese” (0.100 Mercury 0) (002 Zine 2m) “e300 ‘While levels higher than the concentrations listed in table 1.11 ‘will have a negative impact on yeast and adversely affect beer st bility, sublimated levels can have positive effects. Zine is the best understood example, Zine concentration in wort is derived primarily “ ‘An Analysis of Brewing Techniques from malt and possibly from the brewing liquor. At a level of 0.1-0.2 milligrams per liter, zine plays a vital role in fermentation. In parti ular, itis a erucial cofactor for carbon splitting, which is enzymati- cally induced by yeast during respiration (Fix, Principles of Brewing Science 1989). When barley crops are known to be poor, low zinc levels in wort are not uncommon, which can lead to a number of dis- orders (Pieskey 1977). Brewers who use a high fraction of uamalted cereals in standard gravity batches typically experience zinc defi- ciencies. Zine sulfate (ZnSO,) is often added to correct this, ‘The copper content of wort is a more subtle issue. It has been clearly shown that copper is toxic to yeast above 10 milligrams per liter, Also, at or above 1 milligram per liter, copper serves as a cata- lyst of oxidation, leading to permanent haze (Moll 1979). In contrast is the anecdotal evidence of brewers who experience elevated beer sulfur levels in breweries where there is no wort contact with cop- per (Miller 1993) Organic residuals are feared because of their foul flavoring and also because they provide food for microbes. Most water supplies are chlorinated to prevent or at least control these effects. Nevertheless, from a brewing perspective, many water supplies cannot be viewed as sanitaty. To make matters worse, residual chlorine will react with organic material to create products that are very harmful to beer flavor. A number of measures have been used to control water conta- ‘minants, but in the final analysis, some form of water filtration is the ‘most effective means. Here are the best options: (@) Activated carbon filtsation works by absorption, and it is usually used with silver impregnations to provide antimicrobial effects. ‘This is the best filtration system to remove chlorine, organics, and residual particulates, like sand and dust. (2) Reverse osmosis is a series of membrane filtrations that effec- tively removes organics, inorganics, and microbes. It also removes some water minerals, although generally it leaves water chlorine levels unchanged. (3) Deionization is a two-stage process whereby water minerals are exchanged with hydroxide and hydrogen ions. This process also removes organic and inorganic residuals ay well xs microbes. As with reverse osmosis, water chlorine levels are generally unchanged, It should also be noted that many home water soft- ceners use only the first part of the deionization process, where Wort Production %calcium ions are replaced with sodium ions, Such systems are not suitable for processing brewing liquor. ‘The most widely used water filtration system in brewing consists of @) or G), followed by option (1) to remove chlorine. Such water is ready for general use, for cleaning, and for rinsing equipment Since it has no residual protection against microbial infection, great ccare should be taken on a regular basis to ensure brewery water lines are sanitary. ‘Water that has been filtered still may not be suitable for brewing, ‘The final step is to adjust the mineral content of the water in ways that are favorable to the type of beer being brewed. The most rele- vant cations (positively charged ions) are calcium, magnesium, and sodium, Salts are combinations of these with anions (negatively charged ions). The most relevant are bicarbonate, sulfate, and chloride. Water hardness has traditionally referred to the ability of soap to form suds; the harder the water, the less suds. Hardness is correlated to the calcium and magnesium content in water. Today, hardness is reported as the sum of these ions measured as an equivalent amount of calcium carbonate, It can be computed with the following hardness CaCO, = 5/2 x < Ca" > + 25/6 x < Mgt > where < Ca" > and < Mg’* > are the calcium and magnesium con- centrations in ppm (milligrams per liter). Calcium carbonate has a molecular weight of 100, while that of calcium is 40. Their ratio, which is 100/40 = 5/2, is the first coefficient in the formula. The coefficient 25/6 is the ratio of the molecular weight of calcium car- bonate to that of magnesium. These coefficients convert the ion lev- ls into an equivalent amount of calcium carbonate. Alkalinity refers to the number of alkaline ions present (most notably bicarbonates) in the water. Itis typically determined by titration, Pure water supplies can differ dramatically with respect to min- ral content, hardness, and alkalinity, as illustrated in table 1.12. Sample 1 is spring water from Ann Arbor, Michigan, and is charac terized by a high carbonate hardness (i.e., high levels of calcium, magnesium, and bicarbonate). Sample 2 is artesian water from Pittsburgh, Pennsylvania, It is also hard, but unlike sample 1, it has much lower bicarbonate levels. Water like sample 1 is sometimes called “temporarily hard water,” because if it is boiled for a sufficient 6 ‘An Analysis of Brewing Techniques length of time, the calcium ions will complex with the alkaline mate- rial and precipitate out as calcium carbonate. This, of course, will reduce both the hardness and alkalinity of the water. Since the bicar- bonate content of sample 2 is much lower than sample 1, boiling will lead to a lower reduction of hardness. As a consequence, water like sample 2 is often called “permanently hard water.” Table 1.12 Mineral Content of Pure Water Samples Element Sample1* Sample 2 Sample 5 ‘Glen coy) ry 0 36 [Meneame ae 24 Sulite cme) «OG 198.0 290 | Ghone ma 900 140 bonate mg) «AOD 460 250 [__Alkainey Gs Caco 335.0 740 160 | [Marines sco) ~~ SSS«SITO 340 pil a | "Sample 1 ispring wate fom Ana Arbor Michigans Sample 2isatesian water fous Pats, Pennsylvania; and sample 3 public water rom Atington, Texas Water samples 1 and 2 were remarkably free of pollution, and the problematic contaminants cited previously were absent. Neither of these samples was filtered, and the data given in table 1.12 repre- sent the samples as collected at the source. Unfortunately, this is not always the case for artesian and spring water, and brewers using ‘water from such sources should have it carefully analyzed for purity. Sample 3 is from civic water supplies in Arlington, Texas, Samples from the tap typically have residual microbes (more during ‘warmer months), organic matter, and chlorine. While sample 3 meets Environmental Protection Agency drinking water standards, ie is “unsuitable for brewing, The data shown in table 1.12 represent water after filtration (1) and @) discussed previously). This treatment almost completely removed minerals, and water like sample 3 is regarded as soft. ‘The water samples in table 1.12 roughly parallel the differences seen in water (untreated) from classic brewing centers. Selected examples are given in table 1.13. Munich water, like sample 1, is tem poratily hatd. Water like this can also be found in Dublin and London. Wort Production 7‘The samples from Dortmund and Pittsburgh (sample 2) are consid- ered permanently hard, while Pilsen water and the treated water Gample 3) are soft. The big question is how mineral content affects the brewing, process. While stylistic considerations come into play, the most important are the effects water has on mash and wort pH. For the many reasons we will discuss in the next three sections, for most beer styles it is highly desirable to achieve a mash pH of 5.2-5.4 and a wort pH of 5.0-5.3. Following are the relevant mechanisms: A) Interactions between calcium and malt phosphates increase the mash hydrogen ion content and, hence, lower the pH. This mechanism is also present in wort boiling 2) Bicarbonates act as a buffer to any change in mash pH. @) Acids present in malt tend to lower pH, which increases as malt color increases. One can also directly acidify the mash (or mash water) to achieve the same effect. Given a particular water supply and a grain bill, these mecha- nisms compete to determine the mash pH and, later, the wort pH. ‘Therefore, the question is what steps can be taken to ensure pH falls into the desired range. We have found that Kolbach’s criteria have proven to be the most useful in this area (Kolbach 1953). In particu- lr, he introduced a notion of residual alkalinity (RA), defined as: RA = ~ < Ca > / 135 - < Mg > /7, He then showed that the mash pH will increase or decrease as the residual alkalinity is increased or decreased. The approximate rule is that a change in residual alkalinity of 180 milligrams per liter as cal- ‘cium carbonate (hardness) equals a change in pH of 0.3. To see how these concepts can be used, consider sample 3 in table 1.12. For pale beers, mash and wort pH fall into the desirable range if < Ca‘ > equals 50 to 150 milligrams per liter and residual alkalinity is less than 50 milligrams per liter as calcium carbonate, In sample 3, RA = 25 - 9.6) /35 - 2.4) /7 = 24.4 mg/L as CaCO, ‘This comfortably satisfies the criterion that residual alkalinity be less than 50 milligrams per liter as calcium carbonate, The calcium content 8 ‘An Analysis of Brewing Techniques Table 1.15 Classic Brewing Center Water Supplies Munich Dortmund tem ra 2s ages © Sale Goat 0 ‘Chioride (w/t) 160 730 ‘typical of soft water does not meet the specified criterion, The most common procedure is to add calcium chloride hydrate salts. Calculations show that the addition of 20 grams per hectoliter (0.8 ounces per barre!) will increase the calcium level by 50 milligrams per liter (Fix, Principles of Brewing Science 1989). Gypsum is some- times uscd, and here, 22 grams per hectoliter (0.88 ounces per barrel) have the same effect. ‘The above consideration requires modification when the fraction, of dark malts in the grain bill increases, since these malts bring extra acidity to the mash. In this context, calcium carbonate hydrate salts are often used, and the addition of these at 25 grams per hectoliter ounce per barrel) will increase the calcium level by 50 milligrams per liter. Clearly, stylistic considerations (discussed later) come into play, but as a gencral rule, with worts color in the range of 10-15 *L, 4 mixture of calcium chloride (or gypsum) and calcium carbonate should be used. For darker worts, the exclusive use of calcium car- Donate is recommended. ‘Water sample 1 in table 1.12 brings up entirely different issues. Here, the residual alkalinity is quite high; in fact, RA = 409- 141.8 / 35 - 38.1 /7 = 399.5 mg/L as CaCo,, Kolbach’s rule would suggest that for pale beers the pH will be too high, by around 0.6. In test brews, using this water with pale malts led to mash pH levels in the range of 5.75-5.9, which is unaccept- able. Since the calcium context of this water is fine, itis usually best to consider reducing the bicarbonate level for pale beers. One way to do this is through deionization, which will create the same situation as in water sample 3 in table 1.12. Acidification water (or mash) is an altemative. In Bavaria, where ‘water like sample 1 is common, the old practice of biological acidification Wort Production 9of the mash has reemerged as a serious and widely used procedure (liver and Daumen 1988, 3). However, itis a tricky procedure, and only special strains of bacteria can be used (Oliver and Daumen 1988, 4), Tis crucial that these microbes create only lactic acid and no other products. While the microbes used will typically be killed in the boil, their products, on the other hand, tend to spill over to the fin- ished beer. ‘The addition of food-grade acids to the brewing liquor is more ‘common. This procedure has to be approached empiticaly, since ‘water pH and mash pH do not always correlate, In water like sample 41, we have found that 0.1 liter per hectoliter (1 gallon per 1,000 gal ons) of 30% lactic acid suffices. This is reduced to 0.06 liter per hiec- toliter (1 gallon per 1,600 gallons) if 85% phosphoric acid is used. ‘The exact amount needed depends on the residual alkalinity as well as on other factors. [As the color of the wort increases, water like sample 1 becomes more attractive in its native untreated state. It was reported that a turn-ofthe-century brewery in Ann Arbor made sensational dark ales and lagers with this spring water (Ann Arbor News 1975), and this has been our experience with test brews as well. ‘Water sample 2 in table 1.12 also requires different considera- tions. Like sample 1, this water is hard, but with a much lower resid ual alkalinity: RA=75 ~ 88/35 - 48/7 = 65.6 mg/L, ‘Water like this can be used untreated for a variety of pale ales, light amber ales, and lagers. However, test brews indicate less than inspir- ing results when the wort color goes much above 6-8 °L. In this case, dcionization is the only viable option, followed by the measures described previously for dark worts and soft water ‘The most common stylistic issue concerns the use of calcium chloride versus calcium sulfate (gypsum). The chloride ions generally promote beer smoothness, while sulfates tend to promote a harsh hop bitter and drying aftertaste in beer. Thus, calcium chloride tends to be used widely in modern brewing, particularly with more delicate beer styles. There is, however, one major and important exception— Burton ales. The composition of Burton’s water is given in table 1.14, and, as can be seen, the sulfate level is much higher than in the previ ‘ois examples, The effect on finished beer flavors is striking. In England, 20 ‘An Analysis of Brewing Techniques this is often called the “Burton sulfur taste" (Protz 1991), but itis sig- nificantly different from sulfur flavors derived from organic sulfur constituents like DMS. It has a decidedly mineral tone, but it interacts with the high level of Goldings hops that are used in a very unique way, Water like this would be a disaster for most ale and lager for- mulations, but it is largely responsible for Buston’s fame as a brew: ing center. Table 1.14 Burton Water Element ‘Magnesium ea epbes eee cecn ‘Bicarbonate aks 8 [A stylistic issue with Pilsener formulations is whether or not to treat soft brewing water with salt. The archtypal Pilsener, Pilsner Urquell from Pilsen, uses the soft water shown in table 1.13 as is. In the United States, selected small regionals (e.g, the Straub Brewing, Co,, St. Marys, Pennsylvania [Fix 1982]) have done the same, ‘There is a loss of efficiency in wort production (most notably in yields), but the flavoring of the finished beer is indeed quite special. In particu. lar, we found in test brews with various Pilsener formulations that ‘water like sample 3 produces a softer finish if untreated, In Germany both hard water and soft water Pilseners are brewed. Kénig Pils from Duisherg is a widely praised example of a hard water Pils. The water at Duisberg is very similar to that of nearby Dortmund. On the other hand, Veltins, another Rhinelander Pilsener, is a mar: velous example of a soft water version, The direct comparison of these worldclass beers is an excellent Way to get a measure of the effects of brewing water. ‘The effect of bicarbonate is somewhat contradictory, This min- eral tends to promote a rounded malt flavor, but at the same time, a sharper and harsher hop bittemess. Traditionally, the moderately hopped sweet stouts of London and the Munichstyle darks exploited these effects to great advantage. Sill unresolved is the role sodium ions play. Many experts assert ‘that sodium ions have no effect, while others assert that they promote Wort Production afsmoothness, particularly if the water comes from a sodium chloride environment, Our experience has been completely to the contrary, particularly if sodium levels are 75 milligrams per liter or higher. ‘We found a broad harshness not found in equivalent brews made with water having lower sodium levels. Moll reports similar results (Moll 1979), MASHING SYSTEMS From a conceptual point of view, mashing is best scen as an exten- sion of malting in that it continues the carbohydrate and protein mod- ification (degradation) started in malting, The main differences are the times and temperatures involved. Because of this, choosing the best mashing system depends a great deal on the way the grains have been malted. Choosing the Best Mashing System There are a number of parameters that can be used to measure the effectiveness of mashing systems. The most important are carbohy- diate yield, percent fermentabilty, maltose/maltotriose ratios, and protein levels. Carbobydrate Yield—In addition to modifying grain carbohy- rates, we also need to dissolve them in the mash. This is « three~ stage process whereby starch molecules (1) take up water (liquefy), 2) rupture (gelatinize), and @) disperse into the medium. The efi- Ciency of the mashing system determines the yield obtained. Yield the percent of the grain’s weight that goes into extract dissolved in the mash (this and related units are discussed in appendix B) Generally, yields in the 60-70% range can be expected (problems that lead to deviations are discussed later). Percent Fermentability—Not all of the extract dissolved in the mash can be metabolized by yeast. Moreover, the type of mashing. system used strongly influences how much fermentable extract is actually obtained. The relevant parameter is the real degree of fermen- tation (RDP), (RDF is discussed in appendix D, a8 isa forced fermenta- tion procedure for measuring it.) Most mashing systems produce RDFs in the range of 63-67%. However, many Bavarian lagers as well as some strong ales have values below 60%. Low-calorie beers, on the other hand, usually have RDFs in excess of 70%. Thus, the optimal value of RDF depends on stylistic issues. 2 ‘An Analysis of Brewing Techniques Maltose/Mattotriose Ratios—Theoretically, maltose and. mat- totriose are both fermentable sugars. Most lager strains take in both sugars with equal ease, and the ratio of these sugars is not very important. Selected ale strains, on the other hand, behave quite dif ferently. The uptake of maltotriose is much slower than that of mal tose, and the metabolism of the former may be incomplete. Test brews show that for such strains, the ratio should be above 3:0. In cases where the ratio dropped below 3:0, the RDF actually achieved wes lower than predicted by the procedure in appendix C, which means there were fermentable sugars left in the finished beer. This is not desirable, because the beer's balance will be skewed, and because residual sugars could support undesirable microbial activity Proteins Levels—As with carbohydrates, we want to dissolve ‘malt proteins as well as appropriately modify them. In this context, a definite balancing act is required. There are three groups of proteins. ‘The most complicated are the high molecular weight proteins found in unmalted barley and wheat. Their major function in beer isto inter- act with malt phenols and to create haze; therefore, itis advantageous to degrade these proteins or to remove them by mechanical precipi tation. The next level down in complexity are the medium molecular ‘weight proteins, These are highly favorable in that they play a con- structive role in beer foam (see chapter 5's discussion of carbona- tion), and they also play an important (but poorly understood) role in the quality of the beer's malt flavor. The last group consists of amino acids, the building blocks of al proteins, Fix, Principles of Brewing Science 1988). Amino acids have both positive and negative effects. ‘They are a crucial source of nitrogen for yeast, and deficiencies in the finished wort can lead to disordered fermentations (bid.). On the other hand, unmetabolized amino acids that spill over into beer are foam negative. In addition, some yeasts (particularly nonculture strains) tend to use the carbon skeletons of excess amino acids to ‘metabolize harshly flavored fusel alcohols (see chapter 3). As already mentioned, FAN is the unit used to measure amino acid content Experience has shown that FAN levels in wort need to be above 150 ‘milligrams per liter and preferably over 250 milligrams per liter, for a sutficient amount of amino nitrogen to be available for yeast metabo- lism. However, values in excess of 350 milligrams per liter can cause problems. Fortunately, FAN levels can be strongly influenced by ‘mashing programs, so this isan area where brewers have some control Wort Production 25Mash pH ‘One of the most important operational variables in the brewer's con- trol is the mash pH. As noted in the previous discussion about water, the mash pH, for all practical purposes, is determined by the water ‘mineral content and the grain bill. Our experience, with a large num- ber of test brews, is that pH is established very carly in the mash and is buffered from that point forward, Moreover, we found it important to have a pH below 5.5 for all malt worts. Theoretical considerations using mathematical models indicate that this condition is necessary for proper enzymatic activity. This was confirmed in test brews hav: ing a high pH in the mash, with values in the range of 5.6-5.9 lead. ing to erratic results. The yields varied considerably (usually on the low side), and control over RDFs was weak. The fact that a high mash pH may lead to a high wort pH is also a problem (this will be dis- cussed shortly). Temperature Programs “Assuming that a proper pH has been established, a brewer's primary control comes with the time and temperature program used, Following are some general guidelines for various temperature ranges and their effects. (A rest in the range of 95-104 *F (5-40 °C) has traditionally been called an acid rest. We find this term to be somewhat mis- leading, because the pH in test brews was not greatly affected, at Icast for rests of thirty minutes or less. Nevertheless, a num- ber of desirable activities do occur in this temperature range. For example, it is optimal for beta-glucanase, which leads to a better modification of gums, such as beta-ghicans, which in turn improve lautering as well as beer filterability. The major amylase enzymes are not active at these temperatures, however, there is ‘strong evidence that liquefaction of these enzymes is enhanced by a rest at these temperatures, thus preparing the enzyme sys- tems for highertemperature rests. There is virtually no prote- olytic activity in this temperature range, so it is well suited as a low-temperature regime for use with malt that has a high degree of protein modification. @) Atemperature range of 113-131 °F 45-55 °C) has traditionally been regarded as the regime for protein rests. Here the term is justified. Ifa gentle, low oxygen pickup mashing system is used, 24 ‘An Analysis of Brewing Techniques then a nontrivial amount of protein breakdown will occur, ‘Theoretically, a rest in the 113-122 °F (45-50 °C) range favors the breakdown of medium molecular weight proteins into smaller ‘nits, while a rest at 122-131 °F (50-55 °C) favors action on the high molecular weight proteins (de Clerck 1957). In actual ‘mashes and in mathematical simulations, this distinction typically does net exist; and both activities take place, although with a diminished effect, as the temperature is increased. For example, tests show that a rest at 131 °F (55 °C) will not increase the chill proofing stability over brews that had a protein rest at 122 °F (50 °© for the same amount of time. However, only trivial amounts of protein degration was seen at 138 °F (59 °C). Glucanase enzymes remain active over the entire range of 113-131 °F (45-55 °O, but this activity sharply decreases with higher tem- peratures and stops completely at 140 °F G0 °C) ) The 140-148 °F (60-64 °©) range has traditionally been regarded as optimal for betaamylase activity. However, test brews have shown that there is nontrivial betaamalase activity at tempers: tures as low as 131 °F (55 °C). In any case, a rest in this range ‘can be used to increase the maltose/maltotriose ratio. More gen- cally, it can be used to increase the RDF. @ The range for optimal alpha-amylase activity is 154-162 °F (68-72 °O). Spending the bulk of the time at higher tempera tures in this range, as compared to 140-147 °F (60-64 °C), results ia lower RDFs than splitting the time equally between the ‘two regimes. ‘We also can control the amount of oxygen uptake and the amount ‘of hotside abuse that takes place during mashing, The evidence docu- menting the negative effects of hot-side aeration is extensive. For ‘examples see Fix 1992, 15:5; Narziss, Brauwelt 1993, 3; Huige 1992.) ‘Of particular relevance to mashing is the need to avoid shearing forces created by excess stirring, as they have a negative impact on ‘enzymatic activity. It was once believed that the traditional protein rest was of limited value, and data were presented to support this conclusion. This was in direct conilict with other data that showed there was significant proteolytic activity during rests at 113-140 °F (45-60 °C), Hotside aeration was the key in resolving this conflict. In a gentle low oxygen mashing system, protein degradation does take place, but in a less favorable environment, this may not be the case (Uhlig, 1992). Wort Production a(Our test brew results illustrate these ideas. We used a 504iter pilot system, however, the data have been normalized for I-bartel and Lhectoliter batches for convenience. A target extract of 12 °P (SG equals 1.048) was selected for the finished wort. Anticipating an approximate 65% yield, we used an equivalent of 50 pounds of malt for the L-barrel batch and 20 kilograms for the 1-hectoliter batch. ‘The total water used exceeded the batch size by a factor of 1.33, half of which was used in the mash and the other half was used for sparg- ing, On average, there was a 2% loss of extract during sparging, and a 15% loss of moisture. ‘Thus, of the 1.3 hectoliters (1.3 barrels) of ‘water, only 1.1 hectoliters of sweet wort were sent to the Kettle. ‘This volume was reduced 10% by evaporation in the boil, giving a final volume near 1 hectoliter (1 barrel). In some mashing systems, precise temperature control is prob- ematic, which creates a situation where (wo or more temperature ‘regimes may coexist in the mash ata given rest. In such circumstances, ‘yields, RDPs, and other data can vary in erratic and unpredictable ways. In the test brews cited here, our mash tun was fitted with thermally insulated jackets, and gentle mixing kept the thermal gradients below 2.°F (which is the same as a difference of 1 °C). The data are averages over a number of brews; nevertheless, as temperatures were carefully controlled, variances from batch to batch were under 2%. 10440) Bea es oe eT sume etme) Figure 1.2. The 104122-140-158 (40.50.60-70) program, 26 An Analysis of Brewing Techniques Test Batch Series 1~This series used a European Pilsener malt produced from low protein (10%) two-row barley. It had a moder- ate degree of protein modification (Kolbach index of 38%) and a high degree of carbohydrate modification (fine/course grind extract difference of 1.7). The best results with this malt came from a mash using all four temperatures cited previously—104, 122, 140, and 158 °F (40, 50, 60, 70 °C). The time and temperature history is shown in figure 1.2. ‘The malt in test series 1 was doughechin with 0.5 hectoliter (0.5 barre of water at 104 °F (40 °O). After a fiftecn-minute hold, another 8 liters (2.6 gallons) of boiling water was added to increase the tem- perature (0 122 *F (50 °C). This was then held for fifteen minutes, after which another 8 liters (2.6 gallons) of boiling water was added. ‘This, plus a small amount of external heat, raised the temperature to 140 °F (60 ‘C), which was held for thirty minutes. External heat then ‘was used to raise the temperature to 158 °F (70 C*). The final rest was thirty minutes. The results, given in table 1.15, are typical of the data ‘we obtained from this type of system. Our favorite program with this type of malt omits the rest at 104 °F 0 °C and spends thirty minutes each at 122 °F (50°C), 140 °F (60 °C, and 158 °F (70 °C). The 16 liters (5.2 gallons) of extra water can be used to affect the two transitions, so very little if any external heat is required, There is a slight loss of yield (66-67% as opposed to 67-68%), but all other factors remain the same. It is important in a three-temperature program to keep the first rest temperature below 131 °F (55 °C) to ensure good beta-glucanase activity and above 113 °F (45 °C) to ensure good proteolytic activity. A two-temperature pro- gram with rests at 140 °F (60 °C) and 158 °F (70 °C) does not lead to satisfactory results. To obtain acceptable yields, the total lautering time needs to be increased significantly with extended recirculation. ‘This ultimately extracts an appropriate amount of grain sugars but also sharply acids to finished beer astringency. Test Batch Series 2—This series used a low protein (10%), highly modified two-row ale malt. The Kolbach index was 44% and the fine/course grind extract difference was 1.6. The best result with this ‘alt was obtained with @ 104-140-158 °F (40-60-70 °C) program, with thirty minutes at each temperature. Since most of the malts available today have Kolbach indices in excess of 40%, the 40-60-70 °C pro- ‘gram has become our most frequently used system. Sixteen liters (3.2 gallons) of boiling water were used to increase the temperature from. Wort Production ar ae.Table 1.15 Test Batch Series 1 Results Table 1.16 Test Batch Series 2 Results fe Variable Th brew 1 bUL brew Variable thLbrew 1 bh waa aa 067 ales Volume wee brew cen ex HD 186? oar cart = _ : 1077 Fee RD ag sen Tig ha 08 eee ag _ “pm | So ae ee | = a os Yes ae ee sweet Wort Votune oS Percent act WD 1a mectwon vole ttt ee eta iv we Percent exit w/w) P ae | = = % er | FeReaTenet WA) TRS AG. SOBRE - ae “Ler | fee elect DSi 3 Femecaeerie 27 SGT nt = Toulon asks 359 Tol ext aa Yet 5% 9% bs fee aoe a Mel 5% oar | Fed Wort votume vn oe scent extract (W/W ___Selame LOnL LO bbL iss = tract (W/W) ‘Percent extrict (w/w) 18 a 6 Sosa 1082 es Malone | Gu RDF ry (@ ROF can be vad yang the ne pent at 140° 6°) and 158 °F CO “O. For aoe Megane nus the forme and een aes tthe ater wl ney SRR tor wc 67-0s ange the revere wlowe it round 2 SeuTinmen a 190° Gp ove may be ove foal wah gh degree of sa union “Tee ate, however, mone Pisce mats at Dave mh Paeee aratydeae mnt ith Beco gan xtc ere it ee hen mate may ee move ten say mine fe the sine es ae Ge) Dean alcan quy ane hase ay the ned bess wre ted yO Fen ee ane 1122140 5 ADS 60709 progam av let DU rk rcs pon deeds E wos 104-140 °F (40-60 °C), and external heat was used for the last tran- sition. Typical data are given in table 1. 16, cama ‘The comments regarding RDF cited in table 1.15 also apply t0 i this program. One major difference between test series 2 and test series 1 concems protein modification, Test brews indicate that with, sido highly modified malts, time spent in the 113-131 °F 45-55 °C) tem perature range should be kept at a minimum—even a fifteea-minute rest at 122 °F (50 °C) had negative effects. The foam quality was poor, Sooo and, even more striking, the underlying malt flavors took on a dull ee and neutral character. The rest at 104 *F (40 °C), however, proved to 28 ‘An Analysis of Brewing Techniques Figure 1.3. The 140-122-158 (4060-70) program. Wort Production Ebe beneficial. "The results were much better than with the classic sin- ale temperature widely used with ale malts like this one. Figure 14 illustrates our test results with the 150 °F (66 °C) program, and data are given in table 1.17. ‘Omitting the low-temperature rest resulted in a yield toss of carly 15%, Some brewers compensate for this by using extended recirculation, by extensive lautering, of by crushing the malt into a fine powder. A more preferable option is simply to use more matt ‘When we tested this, we found litte difference between the finished beer quality of a single-temperature mash and a three-temperature mash, However, we want to emphasize that this applies only to low protein, low betaglucan, and highly modified two-row malts. Test Batch Series 3—Our final test series used a domestic two row malt with a relatively high protein content (11.7%), a high degtce of protein modification (Kolbach index of 40%), and a mod- erate degree of carbohydrate modification (fine/course grind extract Table 1.17 Test Batch 150 (66) Results Variable thi brew 1 bbL brew Mash Volume 67 me 0.67 bs Percent extract Gw/w) v2 wer a se ToT 108 [Pacem WA) AS kyl. DL_| Toes ae 309 5 re sO difference of 2.2). With this malt, we achieved the best results with a 95-131-140-158 °F (35-55-60-70 °C) system. To avoid excess protein degradation, the rest at 122 °F (50 *C) was replaced with a rest at 131 SB G5 °C) and was held for fifteen minutes. The low-temperature rest was taken at 95 °F (35 °C) and was held for thirty minutes. Test results are given in table 1.18, Some brewers find that their equipment is not suitable for multi- temperature infusion mashing, due primarily to overshoots and large temperature gradients, In such cases, decoction mashing is highly recommentled. Darryl Richman’s book, Bock (1994), describes a welkconsicered. program, easily accommodated with elementary equipment. While it is true that decoction mashing leads to deeper color and to flavoractive constituents not found in infusion mashes, the compounds responsible for these, namely Ncheterocyclics, also are generated during wort boiling. We have found that making suit able adjustments in thermal loading during the boil takes care of this (discussed in section Wort Boiling) Table 1.18 Test Batch Series 5 Results Variable th brew Mas Vote 067m a i = a ec or ex 1320 foes nanos reer neeene enn Sweet Wort Yotume ti Lab sweet Wort Volume Lah erent exact (wh) Tor 102 [en cect ere) PCP ee 1.040 | ss aaneaarermosl Percent exci WA) ‘LD Ryt_—_—275 Ths /OBL_| [Pane enet wh) TERME 204TH = Toaletact 12g act SSSC*R 323 Yield 3% - 45% Go| Finished Wort___Volume Lon |___Finishee Wort Lone 10 a a i Rerent enact) 7 ioe | 5 os oh 36 1019 oe fotos 89 a Mahone 3 3a or [sats 1F oo ow so ‘An Analysis of Brewing Techniques Wort Production 3co Temperate FCO 04440) Figure 1.4, The 150 (66) program. 1580) 90 os ‘ime Gis) Figure 1.5, The 95-131-140.158 @5.55-60-70) program. Fa ‘An Analysis of Brewing Techniques (Our test batches illustrate that the optimal mashing system depends very much on the degrec of modification of the malt being used. However, it should be noted that in addition to the test batches ‘we have described, there are alternatives that produce nearly optimal sults. The crucial points that emerged from our test brews that have ‘wide applicability are: (@) Mash pl should be kept below 5.5 in order to produce consis- tent and reproducible results. (2) If the malt’s Kolbach index exceeds 40%, serious consideration should be given to minimizing the time spent in the temperature range of 113-131 °F (45-55 °O). @) If the Kolbach index is well below 40%, a rest in the 113-131 °F (45-55 °C) range is advantageous. @) Inall cases, the inclusion of some type of low-temperature rest, (below 113 °F [45 °C) will increase yields. HOPS ‘The constitvents of hops, which are dissolved in wort during brewing, are resins, essential oils, sugars, proteins, and tannins. The contribu tions from the last three are minor compared with what is extracted from malt; consequently, the resins and oils are of greater relevance to brewing, It was traditionally believed that resins are responsible for hop taste, while oils define aroma, Since oils are quite volatile, they are rapidly removed during wort boiling. This led to a distinction between bittering hops, which are added early in the boil for their resin con- tent, and aroma hops, which are added late in the boil for their essen- tial oils. Aroma hops are also sometimes added during beer storage. However, research shows that these are oversimplifications of what actually occurs. For example, when hops are added to beer in storage, both hop aroma and flavor increase, and brewers commonty under- estimate the effect late hop additions have on a beer’s hop bitterness. In addition, we have found that bittering hops contribute to hop aroma, even when they are added early in the boil. In fact, the more pungent the oil fraction, the greater the effect. Consequently, we believe that all hops should be evaluated by a common standard. In par- ticular, ifa bittering hop falls below a brewer's standards with respect to aroma, it should be replaced with one that meets the brewer's expectations—regardless of when itis added in the brewing process. Wort Production cyResins Hop resins can be extracted from ground hop cones with methanol ‘and petroleum ether (ASBC Methods of Analysis 1992). The following, diagram shows how resins are broken down, hop | resins soft resins hard resins In fresh hops, soft resins dominate, and hard resins typically make up less than 1% of the total resin, As hops age, the hard resins increase and the soft resins decrease, while the total remains cssen- tially constant, Soft resins, which have the greatest relevance (o brew- ing, are made up of alpha acids, beta acids, and gamma acids, soft resins alpha acids beta acids gamma acids ‘Gamma acids are typically a small percent of the total. They have not been characterized chemically and are believed to be of litle importance ‘Alpha acids are of greatest relevance because they contribute fla- vors associated with hop bitterness. Theit concentration can be determined by titration with lead acetate (ASBC Methods of Analysis 1992) or by photometric methods (Verzeli and de Keukeleire 1990). ‘Alpha acids are only weakly soluble in wort, and during wort boiling they are isomerized into the more soluble isoalpha-acids. Generally, the tofal concentration of iso-alphaacids in wort (or, for that matter, ‘peed is a good indicator of the intensity of a hop’s flavoring, which is measured by a bittering unit (BU) index. Bittering units are mea sured photometrically (which tends to include @ small amount of resins other than iso-alpha-acids) or chromatographically (which is a 54 “An Analysis of Brewing Techniques more specific assay), and they are typically reported in milligrams per liter. Following are some general guidelines: Hop Bitterness Intensity BU Imai) | Med ‘Song White a beer’s BU correlates with the strength of its hop flavor, it does not indicate hop flavor quality. The primary reason for this is that the isoalphaacid fractions in beer are treated as equivalents, when in fact they differ chemically and in other ways as well. The major fractions of alpha acids found in hops are: alpha acids —_—_» wort boiling humulone ad-humulone co-humulone iso-alpha-acids Isochumulone iso-ad-humulone iso-co-humulone Co-thumutone has one less carbon atom than humulone and ad- humulone, which implies that co-humulone has a greater polarity and is more soluble in wort in its isomerized form. Rigby (1972) studied co-humulone and found that the flavoring potential of iso-humulone and iso-ad-humulone tends to be rounded and mellow, while iso-co- humulone tends to impart a crude bitterness. Rigby's results were consistent with the hop varieties in existence at the time of his research; and, indeed, all of the world’s most highly valued classic hops have low co-humulone fractions. Nevertheless, his work was ‘criticized by several experts because specific mechanisms linking, high co-humulone with harsh bitterness were lacking (Verzeli and de Keulkeleire 1991; Wackerbauer and Balzar 1992). A number of new hop varieties have been introduced that shed additional light on the subject. Wort Production #In particular, i is now clear that a low co-humulone fraction does not ccessarily ensure a refined bitter flavor, there are simply 100 many variables involved. On the other hand, all varieties with high co- Ihumulone have the bitter hop taste Rigby predicted Beta acids can be separated into fractions as well: beta acid Jupulone adupulone co-lupulone In addition, colupulone levels, like co-humulone levels, tend to be a varietal characteristic with the classic aroma hops having lower frac- tions than bittering hops. "The role played by beta acids is not entirely clear. These resins in fresh hops are not readily converted into wort-soluble substances. AS ‘hops age, the alpha acid and beta acid levels decrease and oxidize, thus becoming more soluble in wort. Beta acids are included in a larger class of compounds called nonhumulone bitter substances ‘Wackerbauer identified the following schools of thought (Wackerbauer 1993): @) Nonhumulone bitter substances are important and leave a pleas- ing hoppy taste if the hops have an alpha/beta ratio (a/b) that is, sufficiently low. ‘The classic aroma hops satisfy this condition, a/b = Qéalpha acids) / (% beta acids) @ Nonhumulone bitter substances are not important. G) Nonhumulone bitter substances are important and are the main cause of unpleasant hop flavors, We did a number of test brews in order to sort out these con- ficting viewpoints. Here are our conclusions: (@) We found the higher the co-humulone percent—a BU in the mid 20s or higher—the cruder the bitter flavor. No conclusion could be drawn when a BU was below 20, 36 ‘An Analysis of Brewing Techniques (2) Wort pi is as important (maybe more so) as hop composition, Even fresh, high-quality aroma hops contribute harsh tones in beers waerein the finished wort pH is high. (@) An important key to controlling wort pH is to establish a proper mash pl. Ifthe latter can be kept below 5.5, the wort pH should finish below 5.3. We found this condition is sufficient to avoid hharsh tones. Nonhumulone bitter substances depend a lot on hop rates. If the BU is below 20, and if aged aroma hops are used, then a rather pleasant hoppiness can be detected that will not be pre- sent if hops are fresh. This is consistent with the practice of some large commercial brewers that do not use aroma hops ‘until they are at least one year past harvest. On the other hand, nonhuntulone bitter substances do add to beer harshness if aged hops are used in high concentrations. Flavors such as barnlike, cheesy, dank, earthy, grassy, and musty can usually be detected, ‘Thus, at least two schools of thought cited by Wackerbauer may not be conflicting, The validity of either viewpoint depends on the context. As previously stated, both alpha acids and beta acids decrease as hhops age. The extent to which this occurs is difficult to predict accu- rately since it depends on hop variety, hop type (e.8., cone or pellet), and storage conditions. An added complication is that attenuation of these compcunds does not follow first-order kinetics. There are time periods where losses occur at very low rites followed by time peri ‘ods where the rates are much higher. We found the only reliable way to deal with this issue is to store hops in a cold (preferably below 32 °F [0 °C), low-oxygen, low-humidity environment. [tis also useful to ‘measure the alpha levels of whole hops every four to six months and pellet alpha levels every twelve months. Essential Oils The essential oils of hops that are dissolved in beer definitely affect beer aroma and, either directly or indirectly, the overall perception of hop flavor as well. Oil fractions vary significantly with hop variety as ‘well as with the region where hops are grown. Often, the total vil content of a hop is quoted, but this is not nearly as important as the fractions present in a hop. Therefore, our analysis concentrates on the latter. Wort Productionsscntial oil composition can be broadly classified as: essential oils oxygen hydrocarbons constivents” other monoterpenes sesquiterpenes Hydrocarbons and oxygen-bearing constituents are the most rel ‘evant to beer, although sometimes sulfurbearing constituents result ing from heavy pesticide use have been isolated in both hops and the beer brewed from them. Hydrocarbons are made up of monoterpene and sesquiterpene. Monoterpene is essentially myrcene, a compound that has an extremely crude and unpleasant aroma. The major sesquiterpene fractions are humulene, farnasene, and caryophylene Sesquiterpene has a refined and elegant smell. For this reason, the sesqquiterpene/monoterpene (s/m) ratio is important: '8/m_= (total sesquiterpenes ~ ml/kg) / (otal monoterpenes ~ mL/kg). However, the sesquiterpene/monoterpene ratio should be used carefully because farnasene is difficult to measure accurately, even with chromography. Linalool is the major hop alcohol found in vietually all beers. In isolation it has a distinct, floral/hoppy smell that is valued by those wh like becr. Geraniol is another alcohol with floral tones, but in isolation, it has other tones that recall cheap, rank perfume. Many ester fractions have been identified, but their effects are difficult to characterize. ‘There is some evidence that hop esters are 38 ‘An Analysis of Brewing Techniques responsible for some of the unique aromas of selected varieties, ‘Therefore, itis interesting that all the classic aroma hops have very low total ester levels, which is why the ester/oxygen-bearing com- ‘ponent (¢/obc) ratio is important: ¢/obe= (total esters ~ ml/kg) / (otal oxygenbearing components - ml/kg). Oxidation products of humulene, such as humulene epoxides, humulenol, and humuladienone, play an important role in aged- ‘hop aroma. In isolation, these compounds have a grassy and haylike ‘smell that is casy to distinguish from the cheesy and valeric tones of oxidized resins. Hop Varieties In applying these ideas to the major hop varieties available today, it is convenient to group hops with similar quality ratios Gee tables 1.19-1.23). Our grouping method was strongly influenced by the work of Peacock (1992), Foster and Schmidt (1994), and Haunold and Nickerson (1998). Al of the quality factors of classic aroma hops (table 1.19) fall into the excellent category, which is probably why these hops are valued by brewers the world over for their refined taste and aroma. Subgroups in table 1.19 are differentiated by their characteristics; for ‘example, Geiman hops tend to fallin the lower range while U.S. and Table 1.19 Group I: Classic Aroma Hops Hop Variety (@) German Falla, German Hersoracker (©) Gemnan Taunanger, Crechoslvakian Sse, (© UK Eat kent Goudy {UK and Sovenian Seyrhin Galings, UK. Fags US, Characteristics (Low to average alpha evel: 3-65 [i tow arin 0-12 [0 igh in tos 25-4 © tow che ion Wort Productionnr neenennnannnrunnmtrnvTMIPNeNOPURErrrrreeerrreererrererrsscerrecrerevvssi1/s2;s:bH -screivisnetvissesreie UK. hops tend to be in the upper range. Saaz-type hops tend to have high famasene levels, and this is widely felt vo be the reason for their special aroma, (it should be noted that farnascne is virtually destroyed in the pelletizing process, so the aroma is best seen in whole hops.) Farnasene is also present in Hallertautype hops, (, pbutat very low levels; (¢) and (@) have mid-range farnasene levels. The Iinalool level of all five subgroups is high, but (d) has very high levels Table 1.20 Group I: Special Hops i Hop Variety those of *noble-type” hops, particularly in the crucial areas of percent co-humulone and sesquiterpene/monoterpene ratio, It is our opinion that despite their high-brewing quality, these hops are best taken on their own terms and not seen as replacements. For example, U.S, ‘Tettanger hops are quite different from their German counterparts and, for example, have a much fower farnasene content. Mt. Hood, a Hallertau clone, has a much higher farnasene content than the latter, while Liberty, another Hallertau clone, does not. Ultra, one of the newest varieties, comes the closest to the classic Hallertau profile, and many people think it will emerge as the best of the group III vari- ties. Crystal, on the other hand, is unique among the group III hops. [thas a faste and smell that is definitely European, but it tends to be more intensely hoppy than the others in this group. Like Ultra, Crystal is also finding wide acceptance, particularly with craft brewers. Table 1.21 Group Ill: U.S. Versions of "Noble-Type” Hops Hop Variety [ Ge) avenge sn aio: 1-18 Als ) 2alen Lot Sal © . om sente Ss Freshy baked ead Renzothiante Quinolinete, rubbery odor NMethyhiazticine Sugav/methlonine products Cooked cabbage , Brauwelt 1992, 4). Inferior roasted malts are also sus} Te Iso suspected ‘The key to successful wort boiling is to avoid excess and to find a . Extracting hop constituents and removing DMS require at least thermal loading. We found that percent of volume evaporationne during the boil is a very useful control parameter. Here are some use ful guidelines: (@) ‘The best general recommendation is an evaporation rate of 911%, This can usually be achieved with a ninety-sinuse Poll @ If decoction mashing is used for a beer style normally produces by infusion mashing (e.g., British-style ales), keep the evapors: tion rate in the range of 6-7%. {) Tfinfusion mashing is used for a beer style normally produced by a decoction mash (e.g, bock), an evaporation rate as high as 12-14% is advantageous. . (& Inall cases, avoid evaporation rates in excess of 15% seer ocisions we have seen expineed ts incor cect ici mast wed when the el oxi ea med according to @) and @) Ao sing oa ae nage let fom) ae incre et fo ay pobems in fementaon, 54 ‘An Analysis of Brewing Techniques Yeast STRAINS Each yeast strain has its own personality and characteristic flavor sig- nature, which is why yeast strain selection isthe most important step in developing a successful beer formulation. Fortunately, there are a large number of strains available as well as a number of suppliers ‘This section provides an overview of the more important options. A systematic yeast strain classification system relevant to brew- ing is difficuleto provide because changes can occurin yeast simples, «iter during maintenance by the supplier or during brewing. A good ‘ase in point is provided by Poirier and Lang (1978). Their brewery, 4 mediumsized regional in New York, used the same lager strain for years, purchased from a single supplier. Occasional off-flavor nuances—particularly in their delicately flavored beers—suggested that the pitching yeast may not have been a pure culture. Additional analysis done by consultants revealed that at least six strains were present. Table 2.1 shows the individual flavor profiles they found in ‘the Christian Schmidt strain Poirier and Lang were using, It has been our experience that (1) is the closest to the original ‘culture and the most pure version. The elevated diacetyl levels teported with (4), (5), and (6) indicate they are very likely respiratory- deficient mutants—the fault of the brewery and not the supplier. Strains (2) and @) are likely the result of the supplier's maintenance procedures. skews in the overall beer flavor protile produced by (3) are not uncommon, Because of these and other related examples, we decided to take 4 strictly functional approach to yeast classification, characterizing Yeast 3Table 2.1 Substrains of Christian Schmidt Strain Flavor Profile Origin sed ele ‘lean, y/etery,appretaiy sry ~ Gan, very Desy a estersulur blanc ‘lean, este, sgl sultry 9m favor, very unpestant ince or, favor ee Okay various strains on the basis of the pure versions’ behaviors. Any spe- cific version from a specific supplier may or may not share the char- acteristics described, depending on the purity of the sample. Functional Attributes ‘Flavor Profiles—Under normal circumstances, each strain displays its own propensity for creating and/or reducing fermentation prod- ucts, such as esters, fusel alcohols, diacetyl, and sulfur compounds. By and large, this defines the flavor signature of each strain in its pure form, Preferred Temperature and Environmental Conditions—Each strain has a preferred fermentation temperature for optimal behavior. Other important issues are alcohol tolerance, oxygen demand, and sensitivity to wort composition, “Apparent Attenuation—Most brewing strains fully metabolize monosaccharides and disaccharides, however, the ability to metabo- lize trisacchatides and trace sugars varies, This effect is captured by a method called apparent attenuation, Apparent attenuation is the per- cent drop in apparent extract (Lc., as measured by a hydrometer without alcohol corrections) that is typically achieved by the strain under normal circumstances, Floceulation—There ase significant differences in the floccula- tion characteristics of brewing strains, ranging from highly flocculant, to powdery, to nonflocculant. The exact mechanism that induces flocculation is poorly understood, but it is known to be a genetic characteristic that can change with mutation (Russel 1995). ‘Stability Mutations in pure yeast cultures are usually induced by unfavorable environmental conditions. Examples include elevated temperatures, osmotic stress due to high ethanol levels, and starvation 56 ‘An Analysis of Brewing Techniques due (0 lack of nutrients. Nevertheless, each strain tends to display its own propensity toward mutation, with some strains being more for giving of adverse environments. ‘The following are sources of the yeast we used in our evaluations: () Siebel Institute of Technology 4055 W. Peterson Ave. Chicago, Illinois 60646-6001 @) Wyeast Laboratories PO Box 425 Mount Hood, Oregon 97041 @) Hefebank Weihenstephan £8050 Feising 12 Germany @ Food Research Institute Colony Lane Norwich, United Kingdom NR4 7UA (S) Brewers Resource (BrewTech yeast strains) PO Box 507 ‘Woodland Hills, California 91365 ©) Yeast Labs 1308 W. Madison ‘Ann Arbor, Michigan 48103, ® White Labs 7955 Silvertone Ave San Diego, California 92126 ‘The above list is by no means a complete set of yeast suppliers, Our goal was not to test all strains available from all suppliers, but sather limit our study to the major strains in their pure form, We have found the yeast strains listed to be reliable, but this is not to suggest that alternates not on the list are not also reliable. Lager Yeast Genetic research of yeast strains has centered ground chromosome fingerprinting as a biological discriminator Casey 1996). Studies show there is considerable heterogeneity in the chromosomes of ale YeastTE and wheat beer yeast. However, the fingerprints of lager yeast were ‘closely related, indicating a strong similarity to the pure culture intro- duced in 1883 by Hansen (Raines, Yeast Culturing 1992). The group { classification is consistent with the homogeneity of lager yeast In practical terms, we found that today there are only two function: ally different versions—malty and dry/crisp. Group 1 yeast is the lat ter, group 2 yeast the former, Group 1—These strains typically produce beers with a crisp, ‘clean, and refreshingly dry finish. Under normal conditions, diacetyl, ester, and sulfur levels are well below threshold. As a consequence, this group is preferred for American and Scandinavian-style lagers as well as German-style Pilseners. Versions: American/St. Louis and North German + Optimal temperature: 50-54 *F (10-12 °C) + Apparent attenuation: 75-78% + Flocculation: Slightly powdery + Stability: Good during first five to six generations “American/St. Louis lager yeast is a product of one of the world’s largest brewing companies. Beer produced with this yeast tends to have a refreshingly snappy finish that recalls apples. Many feel this fla- vor is @ consequence of brewing procedures (kracusening with a rather brief twenty-one-day cycle time), and not an intrinsic part of the strain’s flavor signature, Sources and order numbers: BrewTech C1620, Wyeast 2007, White Labs WLP020, and Yeast Lab 134 ‘North German lager yeast comes from a brewery in northern Germany that is well known for its clean, crisp, and hoppy Pilsencrs. Source and order number: BrewTech CL660. Group 2—These strains leave a full and rounded flavor that com- plements the malt, Typically, there is a subtle sulfur residual, but this normally has a clean malt character, Group 2 strains tend to be very sensitive to the wort's amino acid pool and have a tendency to mis: behave if it falls much below 200 milligrams per liter. Thus, they ‘work best with all-malt beers or high-gravity adjunct beers. As a rule, all versions display good alcohol tolerance. Group 2 strains are almost ‘universally used in Bavaria for both pale and dark beers. We believe they are quite appropriate for amber lagers as well. On a historical note, these strains traditionally found acceptance in regional brew: ties in the Midwest; and, along with midwestern sixrow barley, they gave definition to the unique character of those beers. 58 ‘An Analysis of Brewing Techniques Versions: W-34/70, W-206, W-308, Christian Schmidt, and Califor- nia Steam + Optimal temperature: 45-50 °F (7-10 °C) + Apparent attentuation: 78-82% + Flocculation: Good + Stability: Good ‘The W-34/70 yeast strain is now the most widely used lager yeas in Gemany in hs found accepance in the United Sites well (Donhauser et al. 1988). Sources and order names: Weihenstephan 34/70, Wyeast 2278, and Yeast Lab 131. ‘The strain W-206 has found wide acceptance in Europe. It is sim- ilar in character to W-34/70, except it tends to have slightly higher ester levels. (We suspect it is very close to strain (2) in table 2.1.) Sources and order numbers: Weihenstephan W-206, Wyeast 2206, Siebel BRY-206, and Yeast Labs 132. : ‘The strain W-308 was once widely used in Germany but as all Dut disppeared there. The major complaint is its tendency to ecome unstable when reused in practical brewing situations. We Fel every close o Chitin Schmit and to wn (1) i able 2.1 ources and order numbers: Wye c sc don feast 2308, White Labs WLPO22, and ‘The Christian Schmidt strain has been the workhorse of small- and medium-sized breweries in the East and Midwest for decades. Ironically, only recently have complaints emerged about its instability In the example involving the New York brewery, Christian Schmidt ‘was rejected as their production yeast in favor of strain (2) in table 2.1 because of its instability. Sources and order numbers: Siebel BRY- 118, Wyeast 2272, and BrewTech (unnumbered). In our opinion, California Steam yeast is a stable version of ‘Christian Schmidt, and both are closely related to the lager yeast ‘culture originally brought to the United States (One Hundred Years of Brewing 1974). This strain has been conditioned to work well at felatively high temperatures for lagers Gc., 60-62 °F [16-17 °C), “which is defintely not the ease with W-34/70 or W-206, However, ‘many brewers are switching to strains like this because the higher temperatures tend to reduce diacetyl faster and cut down on pro- duetion tes Gehl 1995). Sources and order numbers: BrewTech C1690, W} c ot ‘yeast 2112, White Labs WLPO22, and Yeast Yeast 30