Dave's Preferred Priming Procedure

Please note that a fuller writeup of this information, including complete conversions to (shudder) those screwed-up British engineering units, appeared in the July/August 1996 issue of Brewing Techniques.

Introduction

Most brewing textbooks instruct brewers to prime their beers with an amount of priming sugar on a volume per volume basis. The level most commonly cited is 3/4 of a cup per 5 (US) gallons. This can give rise to problems, however: not all sugars are manufactured in the same way, so that 3/4 cup of one type might in fact be a very different amount than 3/4 cup of another type. It is not clear whether it should be a packed cup, a loose cup, or what. It seems extremely obvious to me (and to some others I've talked with) that a better way is to weigh the amount of sugar being used, and prime on the basis of weight sugar per volume beer--for example, grams per litre (g/L). This document is intended to show how this can be done in practice, and why it is a superior method. I acknowledge the helpful input from the following net.brewers: Mark Hibberd, who really should be thought of as a "coauthor" of this page (and who in fact coauthored the BT article mentioned above)--his document on priming available at The Brewery (latest update soon to be installed) saved me a lot of legwork; John DeCarlo, whose question motivated the water-adsorption tests described below; and Dick Dunn, who provided additional evidence of how different sugars might compact (he reports seeing dextrose/glucose compact up to 30% with just a gentle tap!).

Now, it is of course true that if we proceed in the same way every time, we will eventually arrive at a set of amounts that work for us. However, in order to reduce the trial-and-error aspect as much as possible, it would be desirable to start more or less from first principles so that we can have a better idea what we are doing, particularly when entering as-yet-uncharted territory.

Part 1. Arriving at the desired priming rates.

First a brief rundown of the main idea: carbonation is most usefully described in terms of volumes of CO2. A beer carbonated to 2 volumes would have, say, 2 litres of CO2 in every litre of beer. See Dave Miller's books for a more complete discussion. Mark Hibberd's priming guide gives the following handy guide to carbonation levels in a range of styles:

        -------------------------------------------
        Beer style                      Volumes CO2
        -------------------------------------------
        British-style ales              1.5 - 2.0
        Porter, stout                   1.7 - 2.3
        Belgian ales                    1.9 - 2.4
        European lagers                 2.2 - 2.7
        American ales & lagers          2.2 - 2.7
        Lambic                          2.4 - 2.8
        Fruit lambic                    3.0 - 4.5
        German wheat beer               3.3 - 4.5
        -------------------------------------------

Mark's guide also shows that beer that is ready to bottle, having had CO2 bubbling through it more or less continuously, will be CO2 saturated, and that the amount of CO2 dissolved is a function of the temperature of the beer. At lower temperature, the beer can dissolve more CO2. Accordingly, we must take this into account, and prime enough only to add the appropriate number of volumes to that already present, thereby arriving at our desired value.

Here is a list of the saturation values from Mark's paper. These numbers represent how many volumes of CO2 are in the beer at the listed temperatures before we add any priming sugar:

Temp                           Temp
(degC)    Vol. CO2             (degC)    Vol. CO2
------    --------             ------    --------
   0        1.7                  12        1.12
   2        1.6                  14        1.05
   4        1.5                  16        0.99
   6        1.4                  18        0.93
   8        1.3                  20        0.88
  10        1.2                  22        0.83

The reaction that produces CO2 during carbonation is one in which one mole of glucose, C6H12O6, goes to 2 moles of ethanol, CH3CH2OH, and 2 moles of CO2. A little stoichiometric algebra shows that we will add 1 volume of CO2 for every 3.7 g/L glucose added to the beer. So now that we are armed with the temperature dependence data and the amounts from this reaction, we can produce a general predictive relationship to use in our brewery.

The plot below shows how many volumes of CO2 will be produced in the finished beer by priming at the level on the x axis. Each line is labeled for the temperature of the beer being primed, and incorporates the amount of CO2 present prior to priming. We choose the carbonation level we desire, then find the line that corresponds to the beer's temperature, and finally read off the g/L priming rate that will give the desired carbonation.

Example 1. We have a lager at 4°C and want it carbonated at 2.75 volumes. We find 2.75 on the y axis, then move over until we hit the 4°C line, then read down to get about 4.5 g/L. Note that if this beer were at 20°C, we would have to prime at about 7 g/L to get the same level of carbonation, because of the lesser amount of CO2 in the beer before priming at that higher temperature.

Example 2. We have a pub bitter at 16°C that we want carbonated at 2 volumes. This time it requires about 3.7 g/L.

The lines on the plot above can be expressed as equations as well. To calculate the priming rate in g/L, first find (from Mark's table above) the saturation level at the temperature of the beer--let's call it v0. Then choose the volumes of CO2 that correspond to the desired carbonation level--let's call that v. Then

                           v - v0
           Rate in g/L = -----------
                           0.27027

You can confirm this using the two examples given above. For Example 1, the expression gives (with v0 = 1.5 at 4°C and v = 2.75) 4.6 g/L, and for Example 2 (v0 = 0.99 at 16°C, v = 2) we have 3.7 g/L.

If one is priming with sucrose, i.e. table sugar or brown sugar, it turns out that more CO2 is produced per g/L priming. I will not reproduce the graph for this case--it's just like the one above except the lines are displaced. The corresponding equation to use if priming with sucrose is:

 
                           v - v0
           Rate in g/L = -----------
                           0.286

Conversions of all this to screwed-up-British-engineering units is left as an exercise for the reader; but the conversion from g/L to dry ounces per US gallon is:


                    1 g/L = 0.133 oz/US gallon

See why the metric system is better? :-}

Part 2. Testing the reliability of volume vs. weight measurements

John DeCarlo raised the important question that perhaps, because sugar can adsorb water vapor from the air and thus increase its weight, volume might be more reliable a measure than weight. To test this, I did a simple experiment.

First I placed several grams of three types of sugar in open containers for a couple of weeks, so that it could adsorb as much water as possible. I did this with brewing dextrose, plain white table sugar, and brown sugar. The white and brown sugars are both sucrose, of course. Then, I placed the vials of sugar on a hot plate set at 80°C for 24 hours to drive off the adsorbed water. I at first tried to use a drying oven set at 110°C, but this is above the melting point of dextrose, so I was forced to use the hot plate. I took no special steps to ensure that the sugar was totally dry before being exposed to the air, because this most closely mimics the situation for most brewers; and it provides a worst-case result, which is what I am after here.

The results, tabulated below, suggest that the amount of water uptake was negligible, assuming that 24 hr at 80°C is sufficient to drive it off. The amounts ranged from 0.05% by weight for white sugar to 1.2% by weight for dextrose. I conclude from this that the uncertainty on the weight of sugar from adsorbed water is well within the noise of the types of scales used by most homebrewers.

Here are the data from this experiment:

Sugar type         Brown          White          Dextrose
----------         -----          -----          --------

Wt of sugar
at start, g        1.930          3.797            2.706

Wt after 
24 hours, g        1.922          3.795            2.673

Percent wt
loss               0.40           0.05             1.22

Conclusions

Because the effect of carbonation is so important to the overall impression of a beer, it does not make sense to take a chance on having the carbonation come out other than desired.
The governing relations for determining how carbonated a beer will be, as a function of the weight of priming sugar used per unit volume, are known and easy to use.
Accordingly, we can control very accurately the carbonation level of our beers, once we have a feeling for what a given number of volumes of CO2 ''feels like''.
There do not appear to be any problems with using weights as a result of adsorption of water by sugar. The amount of adsorption found in a simple experiment was trivial.

Email me your questions and comments, praise and criticism--or just let me know you were here.

Last updated Sunday 1/17/99.