You are on page 1of 8

Den Skandinaviske Bryggerhjskole

The Scandinavian School of Brewing


Ballings Formula - scrutiny of a brewing dogma
Henning Nielsen, Axel G. Kristiansen, Karen Mette Krieger Lassen and Claudio Erikstrm

ABSTRACT
The design of Ballings formula is explained and the traditional critics are listed.
A survey of full scale fermentations at Carlsberg Brewery in Copenhagen is made to estimate the
difference between the wort strength after yeast pitching and the Original Extract calculated by
Ballings formula for the fermented beer. For a major brand of Lager Beer the formula calculated an
average of the Original Extract after fermentation to be 14.4 %Plato, when the wort strength was
14.3 %Plato.
This fair result was in line with similar surveys made 34 and 64 years ago at the same brewery.
For HGB dilution of the beer the formula fails to calculate an Original Extract, which is diluted to
the same extent as the content of Alcohol and of Extract. This could be a problem to the control of
beer losses.
DEDUCTION OF THE FORMULA
In 1843 1865 Professor C.I.N. Balling at the university in Prague elaborated on the following
mass balance for the fermentation of beer:
Loss of Extract = Alcohol + CO2 + dry matter in new yeast
Fermentation trials in his laboratory and in breweries showed him that 2.0665 g of Extract made
1.000 g of Alcohol and 0.9695 g of CO2. Furthermore, 0.11 g of dry matter in new yeast was
produced (Ref. 1).
Consequently, he assumed the mass balance to be:
2.0665 g (lost Extract) = 1.000 g (Alcohol) + 0.9565 g (CO2) + 0.11 g (dry matter in new yeast)
For 100 g of beer he calculated the mass (gram) of the original wort to be:
= g beer + g CO2 + g dry matter in yeast
= 100 + 1.0665 x A%mas
For the same 100 g of beer the calculated Original Extract in the wort to be:
= 100 x (A%mas x 2.0665 + ER%)
The calculated wort extract as % of the wort was named the Original Extract (in g/100 g or
%Plato) and found to be:
OE%P = (2.0665 x A%mas + ER%) / (1 + 1.0665 x A%mas/100)
This is Ballings grosse Formula, which is recognized and used worldwide. The formula is
authorized by EBC and ASBC as a part of the prescribed beer analysis (Ref. 2, 3).

The calculated Original Extract of the beer has become an important parameter to describe the beer,
and it has been adopted as an integrated part of the beer taxation system in many countries, just as it
has found application in the loss control system of many breweries.
The abbreviations applied in this paper, e.g. OE%P for the calculated Original Extract, will be listed
and explained in the final part of the paper.
FORMER CRITICS
During the 140 years of service Ballings formula has been criticised for several theoretical defects
as well as for deviations found, when comparing the calculated Original Extract to the actual wort
strength. (ref. 4, 5, 6, 7, 8, 9, 10, 11, 12).
In the following we will list the most important theoretical objections made to Ballings formula,
and we will estimate their consequence for the calculation of OE%P for a beer, which has
A%mas = 5.10 % and ER% = 4.64. For this beer Ballings formula without any corrections will
calculate OE%P to be 14.40 %.
Objection A
Ballings mass balance is valid for fermentation of monosaccharide (Dextrose).
However, in the brewery we mainly ferment disaccharide (Maltose), which requires
the up-take of water.
If the up-take of one molecule water per molecule of Alcohol is included in the mass
balance, OE%P will be calculated to be 13.93 instead of 14.40 %P.
Correction A to the calculated OE%P: -0.47
Objection B
Balling incorporated a yeast growth of 0.11 g dry matter per g Alcohol. Normal lager
beer production only allow for a yeast growth of 0.07 g dry matter per g Alcohol. If
this lower yeast growth is included in the mass balance, OE%P will be calculated to
be 14.23 instead of 14.40
Correction B to the calculated OE%P: -0.17
Objection C
Balling did not allow for any evaporation of water during the fermentation. From
small, open fermenters evaporation of more than 2% has been reported for top
fermented beer. From large, closed beer tanks the evaporation of water will be related
to the amount of CO2 released. From 100 g beer 5 g of CO2 is escaping. An amount of
evaporated water equal to about 5% of this (0.25 g) could follow. A%mas, ER% and
OE%P should consequently all be increased by factor 0.9975. Instead of OE%P =
14.40 it should be 14.43
Correction C to the calculated OE%P: + 0.03
Objection D
Balling did not allow for any evaporation of Alcohol during the fermentation. The
evaporation of Alcohol will depend upon the evaporation of water. The Alcohol has a

higher vapour pressure than water, and the Alcohol content of the fermenting beer is
smaller than in the fully fermented beer.
The escaping amount of Alcohol could be equal to about 5% of the evaporated water,
or about 0.25% of the total Alcohol. Instead of OE%P = 14.40 it should be 14.41
Correction D to the calculated OE%P: + 0.01
Objection E
Balling incorporated a loss of CO2, which is equal to the theoretical loss of one
molecule CO2 per molecule of Alcohol. This loss is 0.9565 g CO2 per g Alcohol.
However, some of the CO2 produced will not escape due to chemical reactions and to
insufficient removal prior to the analysis.
We assume that 2% of the CO2 will not escape. If this reduced loss of CO2 (0.9565 x
0.98= 0.937 g CO2 per g Alcohol produced) is included in the mass balance, OE%P
will be calculated to be 14.32 instead of 14.40
Correction E to the calculated OE%P: -0.08

Objection F
Ballings analysis for Alcohol was based upon Meissners Alcohol Table from 1816,
whereas we today use the OIML table from 1973, which we consider more correct.
The Meissner Table gave too high figures for the content of Alcohol. (Ref. 7)
If Balling had used the OIML table, he would have found less Alcohol and thus 4-5 %
more dry matter in yeast per g Alcohol. The amount of CO2 per g Alcohol was based
upon theoretical consideration (1 molecule CO2 per molecule Alcohol). So he would
hardly have changed that.
Consequently, he should only have modified his formula slightly due to
Meissnerstable.
Correction F to the calculated OE%P: -0.02
Objection G
Balling calculated the sugar content in %Balling, where we today calculate in %Plato.
The difference is due to the use of two different sugar tables. 14.4 %Plato = 14.35
%Balling and 4.64 %Plato = 4.57 %Balling.
We consider the Plato Table to be more correct than the Balling Table.
Even if Balling had used the same sugar table, as we do today, he would not have
changed his mass balance, as this is based upon measurements of Alcohol, CO2 and
yeast only.
However, he may have been less satisfied with the calculated results, when he
compared them to his analysis of the true wort.
No correction to be made for the use of the Plato Table

Objection H
Ballings formula contains an Original Extract calculated as percentage of the mass of
the wort. However, the formula as well includes the content of Alcohol and of Extract,
which both are calculated as percentage of the mass of the beer. This discordance
makes problems, when beer is diluted. We will revert to this special problem later in
this paper.
No correction to be made for undiluted beer.
The above objections could add up to a total correction of 0.74%Plato, if they are all fully valid. It
is a general trend that the objections to the formula almost only indicate negative corrections to the
calculated value for Original Extract. Consequently, we should expect the calculated Original
Extract generally to come out too high compared to the wort strength.
In 1942 and again in 1971 evaluations of Ballings formula were made at the Carlsberg Brewery in
Copenhagen. Both studies found that the application of Ballings formula led to too high values.
However, they both found the deviations from the true Original Extract to be smaller than 0.2
%Plato. (Ref. 7, 13)

THE NEW SURVEY


At Carlsberg Brewery in Copenhagen a new survey has been made in 2006 to evaluate the
correctness of Ballings formula for the present process. This survey was limited to involve only
one major brand, which is an international Lager Beer. After fermentation the calculated Original
Extract of this beer is specified to be 14.40 %P, just as the Real Degree of Fermentation is specified
to be 70%. The brand is later diluted to a lower calculated Original Extract. The survey comprised
about 300 brews and a total of 50 fermenter fillings in the spring of 2006.
The fermentation was performed in large cylindro-conical fermenters, each holding 6-8 brews. All
pitching yeast was added to the first brews, and the subsequent brews were added within 12 to 18
hours. The main fermentation temperature was about 13C.

Fermentation tanks at Carlsberg Brewery in Copenhagen

During a number of fermentation the Real Extract (ER%) and the Alcohol content (A%mas) was
measured by means of a Beer Alcolyzer, (Anton Paar) instrument and a SCABA Beer Analyzer
(FOSS) instrument. Both were well calibrated and more accurate than the sampling from a multibrew beer tank will ever be.
The true measured Original Extract was considered to be equal to the OE%P calculated from the
content of Alcohol and Extract found immediately after pitching the yeast to the wort.
The estimation of this true value was based upon as well the monitoring of the calculated Original
Extract during the various fermentations as on the measured wort gravity in the brewhouse. The
brewhouse value was not directly applicable, as the amount of chase water entering the brews in the
whirlpool and in the wort line is varying.
The filling of 6-8 brews into each tank over a period of 12 to 18 hours was a complication for the
estimation of the true wort strength.
Additional a separate study was made to evaluate the initial take-up of extract by the yeast prior to
fermentation.

When surveying and summarizing the calculated OE%P from the many fermentations, the values
were related to the content of Alcohol rather than to the time. This was preferred to minimize the
disturbance caused by the normal variations in fermentation time.

Figure A

OE as function of A%mas
14,45
OE calculated
after fermentation

14,4
OE%P

14,35
14,3

True OE

14,25
14,2
14,15
0

A% mas

Figure A illustrates the development of the average calculated Original Extract, when the content of
Alcohol increases from nothing to A%mas = 5.1 in the fully fermented beer. The initial value
(OE%P = 14.30 at A%mas = 0) is considered to be the true Original Extract, and the final value
(OE%P = 14.40 at A%mas = 5.1) is the calculated value for the fully fermented beer.
The comparison of these two values (14.40 14.30 = 0.10) leads to our positive evaluation of
Ballings formula.
As expected the calculated OE%P is too high compared to the true value. However, the deviation
found was smaller than what should be expected from the list of objections above to the Ballings
formula. The deviation found was agreeing well with studies made in the same brewery 34 and 64
years ago.
Ballings formula is designed for a complete fermentation, and during the fermentation additional
deviations between the true and the calculated OE%P must be expected. The highest deviation was
found around A%mas = 1.5. In this phase of fermentation the ratio of extract removed by the yeast
is considerably higher than the ratio of Alcohol released. The calculated Original Extract in this
phase of fermentation will be smaller than as well the true Wort Extract as the final calculated
Original Extract.

THE PROBLEM OF DILUTION


When a HGB-beer is diluted, the content of Alcohol and of Extract will be reduced. Consequently,
the calculated Original Extract must be reduced as well. However, the reduction of OE%P will not
be proportional to the Alcohol and to the Extract.
For the beer in question a HGB-dilution makes the following changes:
Table 1

Table of Dilution
Alcohol
Real Extract
Calculated Original Extract
%Plato
%mas
OE%P
5.10
100%
4.64
100%
14.40
100%
3.62
70.98%
3.29
70.98% 10.37
72.01%

HGB beer
Final beer

As will be seen from Table 1 the Original Extract calculated after the dilution is 72% of what it was
before the dilution. This in spite of a dilution ratio only of 70.98 %. As mentioned above this
discordance is caused by the design of Ballings formula.
If we calculate the mass of Original Extract for 1,000 hl beer before the dilution, it will be 14,533
kg Original Extract, whereas the beer after dilution will have a volume of 1,411 hl and contain
14,753 kg Original Extract. The dilution has increased the mass of Original Extract by 220 kg or
1.5 %. When we use the mass of Original Extract for the control of the beer loss in a brewery using
the HGB-technique we are consequently misled. The calculated loss will be 1.5 % too low!
If the loss should be 6.0 %, we will calculate only 4.5 %
CONCLUSION
A recent survey of the fermentation of a major international brand of lager beer at Carlsberg
Brewery in Copenhagen showed that the average Original Extract calculated by Ballings formula
was 14.40 %Plato, whereas the average true Original Extract measured in the wort after pitching
was 14.30 %Plato.
The result was in line with similar studies made in the same brewery 34 and 64 years ago.
Ballings formula has often been severely criticized due to the design of the formula. However, the
formula provides fairly satisfactory results at the brewery in Copenhagen.
At the subsequent HGB dilution of the same brand the calculated Original Extract is not diluted to
the same extent as the content of Alcohol and Extract. For this reason the loss of beer will be too
low, when the loss is calculated as the difference between the content of calculated Original Extract
for the outgoing beer minus the content of Extract in the wort. For the beer in question the
calculation of the production loss would fail 1.5%, e.g. showing 4.5 % instead of 6.0 %.

ABBREVIATIONS
ER%
OE%P
A%mas
%Plato
ASBC
EBC
HGB

=
=
=
=
=
=
=

% Real Extract in beer


Calculated Original Extract in wort
% Alcohol by mass in beer
% Extract according to Plato
American Society of Brewing Chemists
European Brewery Convention
High Gravity Brewing process

In %P, g/100 g
In %P, g/100 g
In %, g/100 g
In %, g/100 g

REFFERENCES
(1) Balling, Carl I.N. Die Grungschemie I-IV, Prag 1845, Die Grungschemie I-III. 2. Aufl.,
Prag 1854, Lehrbuch der Bierbrauerei I-II, 3. Aufl. Prag 1865
(2) European Brewery Convention. Analytica-EBC, Fifth Edition, Oct. 2004. Method 9.4 Original, Real and
Apparent Extract of Beer.
(3) Method of Analysis of the American Society of Brewing Chemists (ASBC), Method BEER-6B , 1992, Am.
Soc. Brewing Chemists, St. Paul, MN, USA
(4) Schnfeld, F., Die Nachprfung der Ballingschen Formel zur Stammwrzeberechnung, Wochenschrift fr
Brauerei, No. 28 p 209-221, 1911
(5) Doemens, A., ber die Bieruntersuchung nach Balling, II. Mitteilung. Zeitschrift fr das gesammte
Brauwesen, 34, 369, 385 (1911)
(6) Berglund, V.,Nogle danske Attenuationsundersgelser. Svenska Bryggarefren. Mnedsblad, 56, 331
(1941) and Schweizer Brauerei-Rundschau, 53, 109, 135, (1942)
(7) Trolle, B., Doctoral Thesis at Danish Tech. University, Studier over Attenuationsforholdene ved lgringen
med srligt Henblik paa en Revision af Ballings Formel, (1943)
(8) Trolle, B. Eine genauere und allgemeine Balling-Formel. Brauwissenschaft, 2, 1948, p.34
(9) Brofeldt, M., 128 ber die Berechnung der Stamwrze in Bieren
Brauwissenschafft Jg. 12 (1959) Nr. 5 121
(10) Schild, E., Schneider, G., ber die Berechnung der Stammwrze von unter- und obergrigen Bieren
unter Bercksichtigung der Alkohol- und Wasserverdnstung wrend der Hauptgrung
Brauwiss., 12, 298 311, 1957
(11) Silberhumer, H., Schwarz, H., Wrzeextract und Stammwrze im Bier, Mitteilungen sterreichisches
Getrnke Institut, No. 11/12, 1993, p146-147
(12) Nielsen, H., Panting Laurents, K., Ballings Formula, when used at Faxe Brewery
Scandinavian Brewers Review (2004), 61 (6), 39-40
(13) Rosendal, I. private communication, 2006

AUTHORS
Henning Nielsen, M.Sc. is a lecturer at The Scandinavian School of Brewing in Copenhagen and an independent
brewing consultant. Previously, he has been Director of Production at Bravo International / Heineken Brewery in St.
Petersburg, Russia, Director of Consultation at Alfred Jrgensen Laboratory in Copenhagen and Technical Director at
Faxe Brewery in Denmark
Karen Mette Krieger Lassen, M. SC. is a lecturer at The Scandinavian School of Brewing in Copenhagen and
Production Manager at the Carlsberg Brewery in Copenhagen. 1993 2006 Master Brewer and Laboratory Manager
with Carlsberg Danmark A/S.
Axel G. Kristiansen, M.Sc., Director of The Scandinavian School of Brewing in Copenhagen since 2004. 1984
2004 Master Brewer in various technical management positions with Carlsberg Breweries A/S in production in
Denmark, in UK and in Italy.
Claudio Erikstrm, M.Sc., since 2006 Laboratory manager with Carlsberg Denmark A/S. 2002-2006 Laboratory
manager with MILANA A/S

You might also like