CHAPTER 6

Beer quality and flavor

6.1 INTRODUCTION

There are two distinct aspects to beer quality. The first is its relative
nature, kind, or character and the second, its degree of excellence. In the
former case, we are, for example, talking of beer types in terms of the
method of production, place of origin, or style. Examples would be top fer-
mented, double decoction, Pilsner, and Burton ale. This aspect would also
include many quality parameters that are perceived even before the beer is
consumed. For example, color, clarity, degree of carbonation, and presence
(or not) of foam. The brewer might also refer to beer types as lager, ale, or
dark quality. In recent years, there has been an increased market share for
low and no alcohol beer types. Clearly the quality referred to in these ways
is largely determined by the brewing practice. The same applies to other
beer parameters such as pH and alcohol content. Traditionally, beers arose
as characteristic styles depending on the water supply, available raw mate-
rials, method of temperature control in brewhouse, fermentation, and stor-
age. Today, it is technologically possible for a single brewer to produce a
whole variety of beer qualities from the same equipment. Quality assur-
ance procedures are used to set overall specifications and often techniques
are available for fine tuning them. Table 6-1 gives some examples.
These readily measured and controlled parameters also contribute to
the degree of excellence of the product. This aspect of quality is, of course,
largely subjective and additionally takes into account the flavor of beer.
With the exception of control of bitterness, control of overall flavor is the
most difficult aspect of quality assurance. It still largely relies on the selec-
tion of raw materials and yeast, consistent processing, and hygienic oper-
ations. Nevertheless, the majority of brewers would hold, and sensory sci-
ence shows, that their beers have unique and characteristic flavors and
every effort is taken to maintain this aspect of quality. Flavor evaluation
is used to monitor uniformity, changes in raw materials, or process and
development of new products.

95
M. J. Lewis et al., Brewing
© Aspen Publishers, Inc. 2001
96 BREWING

Table 6-1 Methods of measurement and adjustment of typical beer quality

assurance parameters
Parameter Method of measurement Method of adjustment
Color Spectrophotometer or color Grist composition (colored malts
comparitor to darken; non-malt starch to
lighten); kettle adjuncts
(caramel to darken, non-malt
syrups and reduced boiling to
lighten)
Clarity Spectrophotometer or haze Filtration following prolonged
meter storage at 0 to -1°C, fining
in tank or cask
Removal of protein precursors
with adsorbents (silica hydro-
gels, nylon filters) or pro-
teases (papain), adding tan-
nic acid
Removal of polyphenol precur-
sors with adsorbents
(polyvinyl pyrrolidone, AT)
adding antioxidants (ascorbic
acid)
Foam Generate foam with gas bub- Higher malt in grist; use wheat
bles and time its collapse flour adjunct, alginate
extracts
Adjustment to dispense gases
and taps
Hop bitterness Spectrophotometric, Use pre-isomerized hop
chromotographic extracts after fermentation
Hop aroma Sensory evaluation Late hop addition to kettle; dry
hop; add aroma extracts to
beer
Carbon dioxide Measure volume of gas Lower by venting fermenter;
content absorbed by alkali increase by sealing fermenter
In line carbonation

6.2 FLAVOR

Perception of flavor involves the individual senses of touch, taste and

smell. The sense of touch is used to perceive the so-called mouth-feel
characteristics, which include smoothness, astringency (drying), temper-
 Beer quality and flavor 97

ature, and the tingling sensation given by effervescent CO 2, Taste is per-

ceived by the taste buds of the tongue and four primary tastes are recog-
nized (Figure 6-1). These are sweet, salt, sour, and bitter. Standard solu-
tions of sucrose (cane sugar), sodium chloride (table salt), dilute
hydrochloric acid, and quinine define these tastes. More recently, a fifth
taste standard has been recognized and named umami, and the standard
for this is glutamic acid. It is commonly depicted (as in Figure 6-1) that
sweetness is perceived predominantly at the front of the tongue and bit-
terness at the back, with salty and sour tastes at the front and rear edges
of the tongue, respectively. However, it is likely that this is not the case
and the taste receptor cells may not have such a localized distribution.
Taste buds (small onion-shaped structures) contain 50 to 100 taste recep-
tor cells and there are only a few thousand buds on a tongue. It is also
clear that taste receptors operate in different ways. Salt and sour receptors
respond to ionic stimuli (influx of sodium and hydrogen ions respec-
tively), whereas sweet and bitter receptors respond to complex organic
molecules. From genetic experiments in mice, it is clear that there is a
large number of receptors for identifying different bitter substances. The

c
o 0
o 00
o o
o
BITTER o

~
C
:D

SWEET

Figure 6-1 Simplified diagram of the taste sensory areas of the human tongue. C,
circumvallate papilli at the back of the tongue.
98 BREWING

same situation almost certainly will hold in humans. So, individually, we

may have up to 90 different receptors for bitterness, each perhaps recog-
nizing a specific compound. However, we cannot distinguish the individ-
ual compounds and all of them are perceived simply as bitter. Perhaps the
taste receptor cells produce all of the receptors, thus providing a broad
range and high sensitivity. The taste response is most sensitive to bitter
compounds. This is claimed by some to have a selective advantage since
many plant poisons (alkaloids) are bitter and their presence may be
detected before a lethal dose is reached! Sensitivity to bitterness is clearly
important as far as beer flavor is concerned since one of the most charac-
teristic flavor notes is hop bitterness. The receptors responsible for detect-
ing sweetness remain to be elucidated.
It is quite common for individuals to confuse the senses of taste and
smell. The flavor of any beverage taken into the mouth is virtually simul-
taneously discerned by both senses. This happens as volatile vapors pass
up into the olfactory organ at the back of the nasal cavity (Figure 6-2). It
is possible to test the validity of this by taking a mouthful of beer (or other
suitable beverage) and pinching the nose to stop the circulation of the
vapors (taking due care not to choke in the process!). Similarly, the aroma
of a beverage can be separated from its taste by sniffing a sample. The
nerve endings in the olfactory organ are connected directly to the brain.
Different nerve endings respond to different chemical substances. The
interaction is commonly believed to involve the three-dimensional shapes
of the molecules; however, the chemical receptors in the nerve endings
may be able to sense the vibrational frequencies of the molecules and this
defines aroma. Simultaneous stimulation of different types of nerve end-
ing is thought to produce a different sensation (aroma) to single interac-
tions. The sense of smell is therefore potentially more complex than that
of taste. Both senses are influenced by physiological, psychological, and
genetic factors. Thus, a head cold restricts the passage of vapors to the
olfactory nerves and a spicy meal not only affects smell but desensitizes
the taste buds.
Psychological influences are more difficult to define but most people
would accept that mood, environment, and company influence the per-
ception of a meal or drink.
Genetic factors are clearly important. Certain well-documented exam-
ples illustrate that different individuals differ in their perceptive abilities.
Thus, the compound phenyl thiocarbamide (Figure 6-3) is excruciatingly
bitter to about 40% of a European population, whereas the remaining 60%
cannot taste it at all. Some individuals are born without a sense of smell
(anosmics) and many individuals experience specific anosmias (unable to
 Beer quality and flavor 99

To
trachea

Figure 6-2 Diagrammatic representation of the olfactory organ. Arrows (broken

lines) show the air flows. N, nostril; P, upper palate (roof of mouth); M, mouth; 0,
olfactory lobe (sensory nerve endings); OC, olfactory center in the brain.

perceive certain aromas). Given instances like these, it is quite plausible to

propose that some individuals will have different sensitivities to the same
substances (indeed this can be readily demonstrated by experiment) or
even, for example, lack the ability to perceive some particular aromas. The

CrI ~
N-C-NH
II
S
2

Figure 6-3 Phenyl thiocarbamide. Some 40% of Europeans find this compound
excruciatingly bitter; the remainder cannot taste it
100 BREWING

consequence of this is that the human population is not uniform in terms

of its discernment of flavor. Add to this the fact that our description of
sensory phenomena is to a large extent determined by our individual
experiences and associations, then it is clear that using sensory evaluation
to describe, let alone quantify, quality in terms of "degree of excellence"
is no simple matter. Indeed, so complex is the subject of sensory science,
that it may be tempting to ignore it and rely on our opinions of what an
excellent flavor is. However, in this chapter we will resist this temptation.
In any event flavor must be measured to control it, and opinion is no sub-
stitute for scientific method!

6.3 WHAT GIVES A BEER ITS FLAVOR?

The predominant influences on overall beer flavor are derived from

hop bitterness and aroma, malt components (before and after metabolism
by yeast), and the use of specialized malts and adjuncts. Thus hopping
rates, the use of dry hopping, and/or addition of specialized malts for
color and flavor generate specialized types of beer. Adjuncts are often
used to enhance sweetness or as diluents of malt flavors. Two extremes
would be using starch in mashing to dilute overall flavor and mixing fruit
pulp in the kettle boil to extract a specific flavor.
Malt is a source of many flavor compounds; one that has received much
attention is dimethyl sulfide. Its cabbagy-vegetable, sweetcorn aroma is a
predominant flavor note in many European lager beers. The characteris-
tic flavors of all beers also derive both directly and indirectly from the
products of yeast metabolism. Alcohol, by virtue of its high concentration,
makes a major contribution, but it is the minor products of yeast metabo-
lism that give a beer its characteristic flavor. Table 6-2 gives an indication
of the range and concentrations of materials encountered; some specific
examples are referred to in Chapter 18. There are certainly in excess of 700
such constituents of beer. Most of these are present at levels just below
those at which they are easily perceived. However, acting both synergis-
tically and antagonistically, and together with hop and malt constituents,
they give the overall beery character and the specific flavors associated
with the beverage. When these individual components are present at lev-
els of double or more those where they can be easily discerned, they
exhibit specific flavor notes. These may be characteristic (therefore desir-
able, flavors) or uncharacteristic of the beer (and therefore undesirable,
taints). It follows therefore that one brewer's desirable flavor may well be
another's taint!
 Beer quality and flavor 101

Table 6-2 Some constituents of a typical beer

Componenfl Typical amount (gil)
Ethanol (alcohol) 35
Carbon dioxide 5 (30 produced)
Organic acids (e.g. pyruvate) <0.1
Aldehydes (e.g. acetaldehyde) <0.1
Esters (e.g. isoamyl acetate) <0.01
Higher alcohols (e.g. isobutanol) <0.01
Diketones (e.g. diacetyl) <0.0002
Sulfur compounds (e.g. dimethyl sulfide) <0.00005
aThe other main product of fermentation is of course yeast, about 1.5 g wet weightjl is
produced

6.4 ANALYSIS OF FLAVOR

An experienced individual may smell and taste a beer and declare that
it is within specification. It is as well to remember, however, that she/he
is subject to physiological and psychological influences like anyone else!
Two basic types of objective analytical test may be used: these are the
difference test and the profile analysis. The value of difference tests is
greatly enhanced if tasters are trained to identify different characteristic
beer flavors. Profile analysis must be conducted by a trained panel.
Training involves firstly identifying individuals who are capable of dis-
criminating different beer characters. For such a process, reference mate-
rials are used (some examples are given in Table 6-3). These may be pre-
sented as aroma samples in small vials or added to a beer with no strong
flavor characters of its own. In this case, the aroma and taste of the com-
pound may be evaluated. As well as using pure chemical standards, the
enterprising sensory analyst may choose to use raw materials (e.g.,
ground malt) or other foodstuffs (e.g., liquid from canned sweetcom) as
standards. Such materials are often more readily obtained and have the
distinct advantage that they themselves are foodstuffs.
The taster is also trained to identify flavors using a common set of
descriptive terms. In brewing, the terminology of the Flavor Wheel
(Figure 6-4) has gained general acceptance. However, it is still not
unusual to find different terms being used in specific instances. Proposals
have been made to extend the flavor wheel to include the importance of
mouth-feel characteristics (Figure 6-5). Training in identification of par-
ticular taints may also be given. The diverse origins of many beer taints
are indicated in Table 6-4.
102 BREWING

Table 6-3 Some reference standards for evaluation of flavor and aroma
Standarcfl
Taste
Sweet Cane sugar (sucrose) 0.75% in water
Salt Table salt (sodium chloride) 0.18% in water
Sour Sour milk (or citric acid 0.05% in water)
Bitter Isohumulone (or quinine sulfate 0.001% in
water)
Aroma
Diacetyl (butterscotch, Diacetyl (2,3 butane dione)
toffee, buttery)
DMS Dimethyl sulfide
Cooked cabbage Over-cooked green vegetables
Phenolic, hospital-like Trichlorophenol
(medicinal, antiseptic)
Smoky Guaiacol
Floral (rose) 2-Phenylethanol
Estery (apple) Ethyl hexenoate
Estery (peardrop, banana) Isoamyl acetate
Fatty acid (cheesy, old hop) Isovaleric acid (old hop oil)
Sulfury (yeasty) Commercial or home-made yeast autolysate
Sulfury (sufitic, sulfur dioxide) Potassium metabisulfite
aFor aroma standards the concentrations used depend on whether sniff samples or beers
spiked with the standards are evaluated. Sniff samples are better for initial training. Samples
are dispersed in odor-free solvent in sealed vials

Objective tasting requires specialized facilities and a clear code of prac-

tice. Randomized procedures are used to code samples and the prepara-
tion and presentation are done by different individuals. Each taster con-
ducts the analysis in isolation, in dim light in a room of constant humidity
with filtered and odorless air. Sessions are held at times away from those
used for meals and coffee breaks. Two types of differential tests are often
used: the duo-trio and the triangular. Both use three samples, the former
uses a control beer as a reference and a control and test beer; the taster is
asked to match the reference. The latter does not use a reference and two
glasses contain the control and one the test (or vice versa). In the absence
of a difference, 50% of answers will be correct in the duo-trio test and 33%
in the triangular test. Statistically significant deviation from these scores
indicates a difference in the beers under test. These tests are often used to
compare the effects of process changes on beer flavor. Typically 15 to 30
individuals, after adequate training, may be used in such tests.
A typical procedure for sensory evaluation of beer as recommended by
the American Society of Brewing Chemists is shown in Table 6-5. Clearly
 Beer quality and flavor 103

~
.... .!:1
~0 "0

-
~ .c
() 0 ~ 0
~
~ ~
Q].
0 «
OJ 0
~ ~. 0
Q.
~

~.
~
~ -< 0
~ '" o

0310 Grainy

~
F' '6
;,::
CJ 0
.....
0
~
0
0 o.
"3
l 0
V)
0
Ul
~
(/l
c;.
c:
~

CJ ~ g ~

r:
0

0
0

Figure 6-4 Beer flavor wheel. Copyright American Society of Brewing Chemists;
with permission.

from the foregoing, the conditions used to evaluate beer flavor are far
removed from those in which the product is normally consumed!
Flavor profile analysis uses small panels of three to five trained tasters.
They will score a sample for particular attributes. In this way, a profile of
attribute scores will be obtained. This analysis is more comprehensive and
is used to monitor flavor consistency, look for specific effects of process
104 BREWING

o
~

Carbonallon
Sling

Figure 6--5 Proposed expansion of the flavor wheel to include mouth-feel charac-
teristics as of equal status to odor (aroma) and taste characteristics.

changes, and analyze beer types. In this way, an objective evaluation of a

beer's flavor can be obtained, which is far more comprehensive than can
be obtained by for example chemical analysis. Additionally, simple visual
comparison of "profiles" may be made using suitable graphical methods
such as spider plots (Figure 6-6).

Table 6-4 The origins of beer taints

Basic ingredients Water; malt (OMS); roasted malt (phenolic); yeast
(contamination; autolysis); poor quality hops and extracts
Process aids Clarifying agents; finings (sulfur); primings (iron and
infection)
Process Filter aid; trub; fermentation; conditioning; mashing; high
gravity
Contamination Coolant; detergent; air; sterilant; poor hygiene and
sanitation
Packaging Air; leaking containers; poor washing of containers;
microbiological contamination
Dispense Tubing; detergent; sterilant; air; glass rinsing and drying
 Beer quality and flavor 105

Table 6-5 Sensory evaluation of beer, as recommended by the American Society

of Brewing Chemists (ASBC)
Some basic conditions for flavor evaluation
1. Silence
2. Dim light
3. Participants separated in booths
4. Room ventilated with odorless air of constant humidity
5. No smoking
6. Participants not to use perfumed cosmetics
7. Participants not to have consumed food or drink immediately prior to the
session
Rules for triangular tasting (ASBC)
1. Compare odors before and after swirling
2. If an odor difference is detected, taste the one (or two) samples having
the least odor
3. Decide on the difference
4. Fill in tasting form

Body

Malty Bitterness

Figure 6--6 Spider plot (radar plot) profiles of two beers. Intensity scores (0 none
to 4 very intense) for different attributes (body, hoppy, bitterness, malty and
fruity) radiate from the center. Note that 'hoppy' refers to the aromatic contribu-
tion of hops. The inner solid line is a U.s. light beer brewed by a major company;
the outer dashed line is a U.s. pale ale brewed by a microbrewery. Although the
trained panel scored them both equal in fruity character (esters), they are
markedly different in the other four characters!
106 BREWING

6.5 BEER TYPES

The popular brewing literature abounds with recipes for traditional

beer types. These products are produced largely by varying the composi-
tion of the malt (and adjunct) grist. This is achieved by including special-
ized malts, caramels, and sugars. Other variations involve changing the
nature of the hop varieties used, varying the times of hop addition to the
kettle, and dry hopping. The nature of the conditioning process and use
of priming sugars also markedly affect beer flavor.
The strength of the beer is also a great variable. In most markets, there
has always been the requirement for beers of lighter flavor and less intox-
icating properties (e.g., small beers and mild beers produced from weaker
worts). These, typically, were consumed by men involved in heavy man-
ual work. In response to the concerns of the modern consumer about diet
(calorific intake) and regulations on blood alcohol and driVing, the brewer
produces a range of different beers. These include dry beers, low calorie
(low carbohydrate, lite beers), low alcohol, and no alcohol beers.
Conventional modern brewing plant is used and either simple modifica-
tions to the process or the introduction of additional processes are used to
make these products.

6.5.1 Dry beers

In many countries, the brewing practice optimizes total extract produc-
tion sometimes at the expense of fermentable extract. The consequence of
this is that more extract in the form of dextrins passes into beer. These
components add mouth-feel and (because of the action of amylases in the
mouth of the consumer) sweetness to the beer. Lowering the mashing
temperature increases the amylolytic degradation of malt starch by (3-
amylase, so the beer contains less dextrinous material and is said to be
drier. Other flavor consequences also undoubtedly result from the change
in mashing conditions and lower hop rate.

6.5.2 Cold filtered beers

These products are produced using efficient filtration processes to
ensure removal of yeasts and contaminating microorganisms from fin-
ished beers. The practice ensures that pasteurisation is not required so
any characteristic flavors that may result from the latter are no longer
present.
 Beer quality and flavor 107

6.5.3 Lite, low calorie, low carbohydrate beers

These terms are applied to a whole range of products. At their simplest,
they are beers brewed from wort of lower gravity (containing less fer-
mentable extract) than the "normal beer." Alternatively, extracting malt,
containing high soluble protein, at low temperature (35-40°C), leaves starch
behind but solubilizes many malt components contributing to beer flavor.
These include simple sugars, but small amounts of additional highly fer-
mentable sugars can be added to obtain a fermentable wort. The majority
of lite beers are however produced by greatly lowering (in some cases to
zero) residual dextrin content. The dextrins in beer arise because the malt
amylases are limited in their action on malt starch (see Chapter 13).
However, if enzymes capable of degrading these dextrins are added to the
process, the polymers are removed and the sugar produced is fermented by
yeast. Enzymes capable of hydrolyzing the ex-l-6 and ex-l-4 glycosidic
bonds of the dextrins are found in unkilned or lightly kilned malt. Extracts
of this material may be added during mashing or at the fermentation stage.
Alternatively, enzymes produced commercially from fungi or bacteria may
be employed (often referred to as amyloglucosidases).
As a rough approximation, the calorific content of beer can be
expressed as:
Calories in 10 cl =4 X % (w Iv) solids X 7 X % (w Iv) alcohol
In typical beers, residual dextrin accounts for the bulk (75%) of the
solids. The calorific value of beer is largely due to alcohol, but dieticians
do not consider alcohol to be a carbohydrate. Using enzymes therefore
removes the carbohydrate; hence, the beers are low carbohydrate. The
calorific value of the alcohol produced by the fermentation of the addi-
tional sugars is less than that of the sugars themselves; therefore, the beer
would be higher in alcohol but lower in calories (but not by much). In
practice, less malt would be used in this process to achieve a typical alco-
hol concentration in beer, thus further lowering the calorific value. How
much reduced in calorific value such beers are clearly depends on the
calorie content of the standard brew. In general, most low calorie beers are
20-30% lower in calories than the standard product. The term "lite" is also
applied to beers with lower alcohol content.

6.5.4 Low alcohol beers

Unfortunately, world-wide, there is no agreed standard terminology for
these products. "No alcohol" beer ("alcohol-free" beer) in most instances
108 BREWING

will contain less than 0.05% alcohol by volume. A category of beers of

alcohol content 0.05-0.5% is referred to in the u.K. as "dealcoholized" and
beers with up to 1.2% alcohol as "low alcohol." Additionally, there is a cat-
egory of beers from 1.2-3.0%, often considered to be reduced in alcohol.
Beers with alcohol content of 2-3% are usually produced by normal pro-
cessing of appropriate grists or by fermenting weak worts and adding back
the volatile constituents collected from the vent of fermentations of high-
gravity worts. Beers with alcohol content of 1.2% and below are generally
produced using additional technologies to remove alcohol. However, this
requires considerable investment in specialized equipment. Smaller brew-
ing operations can achieve these objectives by other techniques. Thus
restricting fermentation of normal worts by using yeast unable to ferment
maltose is a practical proposition (down to 0.5% ethanol); compared to
diluting high-gravity wort before or after fermentation (down to 2% alco-
hol); or compared to mashing at high temperature (80°C) to produce a wort
of low fermentability (1 % alcohol); and compared to the cold contact
process of mixing wort with yeast from a normal fermentation at low tem-
perature (-1°C) for several days. In this latter process, provided that alco-
hol-free yeast is used, beers of less than 0.05% alcohol can be produced.

6.5.5 Ice beers

In a patented process introduced in North America, green beer is con-
tinuously chilled to form tiny ice crystals in a scraped surface heat
exchanger (without any significant beer concentration) and then exposed
to beer-ice in a reactor. It is claimed that this rapid process stabilizes the
beer physically and matures its flavour.

6.5.6 Removal of alcohol

Three technologies are used to remove alcohol from beers brewed to a
normal specification. They are: (1) vacuum evaporation (also called vac-
uum distillation); (2) dialysis; and (3) reverse osmosis.

Vacuum evaporation
Vacuum evaporation (Figure 6-7) uses plate evaporators, which are sim-
ilar in appearance to plate heat exchangers. Beer pre-heated to 35°C is
passed under vacuum through a plate evaporator heated with steam at
50°C. The evaporated stream passes to a separator where the dealcoholized
beer (DAB) is withdrawn and the alcohol-rich vapor passes to a condenser
 Steam ----1._~_ s
50°C '

Condensate

35°C
C Coolant
p
OSC O:l
~
'""t
~
BD:~:I ~I ;;::
O.5°C e..
~.
l':l
Alcoholic ;:::
:;::..
condensate
"-;:::!;,
Hot l':l
Glycol water g
'""t

Figure 6--7 Vacuum evaporator for removing alcohol from beer. E, plate evaporator; 5, separator; P, heat exchanger; C, con- ......
denser; DAB, dealcoholized beer.
~
110 BREWING

to yield an alcohol-rich by-product. The DAB can be re-circulated to obtain

further reduction in alcohol content. It is more usual, however, to use a
multistage process. This typically will have in series three evaporators and
three separators. Such a process gives more control and better energy uti-
lization than a single stage system. The by-product is typically denatured
and sold for vinegar production. The vacuum distillation process readily
produces product with less than 0.05% by volume alcohol.

Dialysis
The dialysis process (Figure 6--8) operates at low temperature and uses
the selectivity of a semi-permeable membrane (or hollow fibers). Small
molecules pass through the membrane into the dialysis medium. This flows
to a vacuum distillation column where the alcohol is continuously removed
and the stream of alcohol-free dialysis liquor returned to the dialysis unit.
Non-volatile beer components at the same concentration as in the inflow-
ing beer eventually circulate in the dialyzing stream. Since the driving force
for dialysis is a concentration gradient when the concentrations of non-
volatile materials across the membrane are the same, only those volatile
materials (mainly alcohol) removed in the distillation column continue to
pass across the membrane. The final product may contain as little as 0.5%
alcohol. Typically, beers of about 2% alcohol are produced in this way. The
nature of the membrane, rate of flow of beer, and dialysate can be manipu-
lated to control the process. This process avoids heating bulk beer.

Reverse osmosis
Reverse osmosis also uses a semi-permeable membrane, but one which
is more selective than that used for dialysis. Water and alcohol permeate to
a far greater extent than other beer components. Beer is passed to the sur-
face of the membrane at high pressure (40 bar) and this forces the perme-
able substances across. The permeate is removed and the beer is concen-
trated about twofold. Initially, the flow of water occurs at a higher rate
than alcohol. However, the transfer of alcohol increases with its increasing
concentration on the inside of the membrane. Pure, demineralized, and
deoxygenated water is then passed outside the membrane (diafiltration)
and alcohol continues to pass from the beer concentrate. The process is
stopped when the beer has the required alcohol content such that dilution
to the original volume gives the desired low alcohol specification (usually
around 2%). Membrane processes require scrupulous sanitation and regu-
lar replacement of membranes which deteriorate and become blocked.
Reverse osmosis has the principal advantage that no heat is applied to the
system, but it suffers from the disadvantage of operation at high pressure.
 Beer quality and flavor 111

Alcohol-
rich
_--4.....-- Alcohol-
rich
dialysate vapour
VC

0 1 I
I I
.!- I

n
I I
I I
I
I
.L

ii
.J-I
J.
I
I
I
I
I
I -4
I I
I I ---+-....- Sleam
I I
I

Alcohol-
reduced
dialysate

Coolant

Figure 6-8 Dialyser for removing alcohol from beer. D, dialysis unit; M, mem-
brane; Vc, vacuum distillation column; H, heat exchanger; DAB, dealcoholized
beer. Arrows indicate directions of flow.

Low alcohol products lack the preservative properties of ethanol and as

a result are less stable especially in microbiological terms. Consequently,
rigorous CIP and microbiological quality control are needed. The produc-
tion of low alcohol beers (especially less than 0.05% alcohol) obviously
incurs extra cost. However, there is usually a significant saving since tax
is not paid at the same rate as for standard beers (less alcohol); moreover,
no tax is levied on no-alcohol beers.