June 2013


The accurate measurement of alcohol in wine is relatively easy for experienced laboratory operators with the correct equipment and techniques. The use of this measurement on labels, however, is a regulatory minefield. In this article we discuss the best technique to measure alcohol and the best way to use the data correctly.

Why measure alcohol content?

Many parameters are measured when making wine such as sulfur dioxide, pH, malic acid and alcohol. Some of these, such as sulfur dioxide and pH, are measured to ensure the wine is stable and will not deteriorate over the span of its life. The alcohol in wine, whilst obviously providing some stability against microbial growth, is not a critical stability parameter in the normal range of table wine. So why do winemakers need to know this value accurately?

In Australia and most other countries, the only analytical measurement that is required by government regulation to be printed on a wine label is the ethanol (commonly referred to as alcohol) content. All alcoholic beverages are subject to taxes which vary between different types of drink; for example in Australia beer is taxed differently to wine or spirits; this taxation varies between countries however that discussion is beyond the scope of this article.

Measurement techniques

The techniques for the measurement of alcohol in wine are many and varied. This topic was covered by the author in a previous article in this journal in 2004 and little has changed with these techniques in the intervening 9 years.

In fact, one of the most common techniques is ebulliometry (though it can also be one of the least accurate) which has been performed by virtually the same piece of equipment from the same French manufacturer for over 150 years (picture 1)! The author was fortunate to visit the Dujardin and Salleron factory in France this year and observed the original instruments from the 19th century – not much has changed since then, although an electronic version of the classic ebulliometer is now available.

Accuracy of measurement

The most commonly used techniques discussed in the 2004 article were:

  • HPLC
  • GC
  • NIR
  • Ebulliometry
  • DichromateTitration
  • Distillation/pycnometry
  • Distillation/hydrometry
  • Distillation/densitometry

These techniques vary in accuracy, cost of setup, ease of use and range of alcohol content covered. Setup costs depending on the technique chosen can vary from less than $1000 to over $100,000 and accuracy is one critical parameter that must be considered. The abovementioned techniques vary widely in their accuracy from more than +/- 0.5% to the most accurate at +/-0.1%. A good accredited wine lab in our experience would be providing this testing with a tolerance of typically +/- 0.1%, although this will depend upon the equipment used and experience of the lab.

How accurate does the final result need to be?

The “tolerance” for the alcohol label statement from different wine producing and importing countries is shown in Table 1. This data is a summary obtained from the “Label Guides” from the Wine Australia website:

CountryTable WineFortified WineSparkling Wine
European Union0.8%0.8%0.8%

* content above 16%. **content above 14%. NB:European Union requirements have recently changed. Previously the alcohol content was required to be stated to the nearest 0.5% and now it is to be stated to the nearest 0.1%.

The logic behind the variance in the different tolerances for different countries and wine styles escapes the author; in fact the logic (or lack of it) is seriously flawed. For instance, why is the tolerance for fortified wine 0.5% and yet table wine a tolerance of 1.5%? From our experience in Vintessential Labs, fortified wine is as difficult, if not more difficult, to measure to the same level of accuracy as table wine.

Standard Drinks Legislation

In addition to a statement of alcohol content in Australia and New Zealand, wine labels are required to contain a Standard Drinks statement. The Australian and New Zealand Food Safety Authority (ANZFSA) Standard 2.7.1 specifies the labeling requirements for alcoholic beverages for Australia and New Zealand. The standard states that the label must include a statement that “standard drink means the amount of a beverage which contains 10 grams of ethanol, measured at 20°C.” It also states that “the label on a package …must include a statement of the approximate number of standard drinks in the package”.

In regard to Standard Drinks, below is an example of a label statement from FSANZ:
“For a 750 mL bottle of 12.5% wine:

What to put on your label

If you have a wine that is tested by an independent ISO17025 accredited laboratory (such as Vintessential) and you receive a laboratory report that states your wine has 14.0% alcohol what are you obliged to report?

From our understanding this means you could use on your label a figure anywhere from 12.5% to 15.5% (i.e. a tolerance of +/- 1.5% not taking our precision into account). We are not implying that you should use a number different to what we provide, however it does appear that this is a possibility.

So if you use 12.5% as the alcohol statement on your label, what number would you use for calculating your Standard Drinks statement? Should you use the 12.5% on your label or the accurately reported 14.0%? If you use 14%, the standard drinks would be 8.3; if you use12.5%, the standard drinks would be 7.4 – almost one whole standard drink different!

Our interpretation of the Standard Drinks Statement clause in Standard 2.7.1 is that you should use the “actual” result that is provided. If you then use a different number for the Alcohol Statement, it would certainly create an anomaly between the 2 statements on your label.


As an upshot of this anomaly we highly recommend that producers selling wine in the domestic market use the accurate alcohol content provided by a reliable laboratory on their labels for both alcohol content and Standard Drinks statements. This will avoid potential repercussions from both the marketplace and regulators. With increased scrutiny of the wine industry by anti-alcohol action groups, the potential difference in reporting of Standard Drinks could become problematic.

Endnote: Caution re “Low Alcohol Wines”

With the current interest in producing and marketing “low alcohol wines”, an interesting clause in ANZSFA standard 2.7.1 may be cause for reflection for some wine marketers: “An alcoholic beverage which contains more than 1.15 % alcohol by volume must not be represented as a low alcohol beverage”.

That is, any wine produced and sold in Australia or New Zealand should not be referred to as a low alcohol wine. From some discussion in recently articles of certain wine magazines, this does not appear to be the current situation.


Howell 2004

Wine Australia


Article from the  Australian and New Zealand Grapegrower and Winemaker, Issue 594, 2013
Page Number(s): 70-72

Author: Greg Howell

Copyright© 2013 Vintessential Laboratories. No part of this publication may be reproduced in any form or by any means without the permission in writing from the copyright owner.