FermCalc - Hydrometer Temperature Corrections and Alcohol Content Calculations
The Alcohol Content panel has four functions:
To calculate the alcohol content using the SG drop method, you need to know the initial and final hydrometer SG readings. These calculations are only valid if you have not added any additional water or sugar between the initial and final readings. To calculate the alcohol content using the hydrometer & refractometer method, you need simultaneous hydrometer and refractometer readings. To calculate the alcohol content using the boiling off method, you need two hydrometer readings obtained by the procedure described below. FermCalc can make temperature corrections to the hydrometer readings for all three methods.Back to top
Making a Calculation
To make a calculation, follow these steps:
The temperature-corrected specific gravities and the approximate alcohol content of your wine will appear in the output fields as you type.
All gravity/density values are converted to specific gravity when they are entered, and are subjected to a lower limit of 0.80 and an upper limit of 1.5545. Temperatures must range between 0ºC and 55ºC. If any of the entered values are outside of these ranges, output fields are highlighted in red and an error message is displayed.
Calculation details are provided below.Back to top
Hydrometer Temperature Corrections
FermCalc allows hydrometer readings to be corrected for temperature in all three of the alcohol content calculation methods. FermCalc uses tables published in the AOAC Official Methods of Analysis (Williams, 1984) to correct the readings. These tables have a reference temperature of 20°C, so first the readings need to be corrected for the hydrometer calibration temperature.
The specific gravity measured with a hydrometer is the ratio of the density of the liquid divided by the density of water at the hydrometer calibration temperature, or:
sgTc = specific gravity measured with a hydrometer with calibration temperature Tc
To correct the reading to a different calibration temperature, we simply need to multiply the hydrometer reading by the ratio of the water densities at the new and old calibration temperatures, or:
sg20 = specific gravity corrected to a reference temperature of 20°C
Water densities are calculated using the OIML formula (OIML, 1973) assuming 0% alcohol.
After correcting for the hydrometer calibration temperature, we can correct for the sample temperature. According to the AOAC tables, temperature corrections to hydrometer measurements are functions of both the temperature and the Brix level of the must. Based on these tables, I developed the following equations to correct hydrometer readings.
B = corrected Brix
The plot below compares the AOAC data with the calculated Brix corrections from equation (3) above.
Back to top
Alcohol Content by the Specific Gravity (SG) Drop Method
FermCalc includes four different methods for calculating alcohol content from the drop in specific gravity that occurs during fermentation. All four of these methods require hydrometer measurements of the initial and final specific gravities. (These methods do not work with refractometer measurements.) Agreement between the methods is generally excellent.Back to top
Standard SG Drop Method
This is the most commonly used method, and is described on pages 79-80 of First Steps in Winemaking by C. J. J. Berry (1987). It estimates the alcohol content by dividing the drop in specific gravity by the constant 0.00736, or:
av = alcohol content, % by volume
A quick internet search will turn up a number of variations of equation (4) that use constants that are both higher and lower than the value of 0.00736 used above. However, I can only find experimental verification for the 0.00736 value. Ritchie Products Co. (2004) claims to have compared the results of equation (4) to the results of gas chromatography for a wide range of wines, and the results were within 0.3% vol/vol.Back to top
Duncan & Acton Method
This method is described on pages 64-66 of Progressive Winemaking by Peter Duncan and Bryan Acton (1967). The Duncan & Acton method calculates the alcohol content from the initial and final specific gravities divided by a factor F that is a function of the corrected initial specific gravity. The equations are as follows.
av = alcohol content, % by volume
Combining equations (5) through (7) above yields the following equation:
Back to top
The Balling method is normally used for beer but gives results that agree very well with the other methods. The equations used in FermCalc were taken from Michael Hall's article "Brew by the Numbers" in the Summer 1995 issue of Zymurgy magazine. In the original equations, specific gravities are expressed as degrees Plato, which FermCalc treats as being the equivalent as degrees Brix. The method requires the calculation of a parameter called "Real Extract", which is an estimate of the residual solids content after fermentation has finished, as follows:
q = attenuation coefficient
The alcohol content (% by weight) is then calculated as:
where aw is the alcohol content in % by weight.
The result of equation (11) is then converted to % alcohol by volume as described here.Back to top
Cutaia, Reid & Spears Method
Cutaia, Reid & Spears (2009) analyzed data from 532 beers to develop equation (12) below relating alcohol content to the initial and final specific gravities.
The alcohol content of the beers ranged from 3% to 7% by weight (approx. 3.8% to 8.7% by vol.). As with the Balling equation above, the specific gravities for the analyzed beers were expressed in degrees Plato, which is assumed by FermCalc to be the same as degrees Brix.
The result of equation (12) is then converted to % alcohol by volume as described here.Back to top
Estimation of Solids Content (True Brix)
After we know the alcohol content, we can estimate true Brix, which represents solids content in % by weight, by using the model developed by James Hackbarth (2011), which is described below. This is done by treating the specific gravity SG and alcohol content A as known values and iteratively solving equations (34) through (40) below for the true Brix E. FermCalc uses the average alcohol content calculated by the above four methods for this calculation.Back to top
Alcohol Content by the Hydrometer & Refractometer Method
FermCalc includes four methods of estimating alcohol content and residual solids (true Brix) from simultaneous refractometer and hydrometer readings. All of the methods are designed to be used after fermentation, but they should be able to yield reasonable estimates of alcohol content during fermentation as long as there is enough alcohol to affect the measurements and the sample is degassed enough that the measurements are not affected by dissolved CO2.Back to top
Rogerson & Symington Method
Rogerson & Symington (2006) developed a method to estimate alcohol content and residual solids (true Brix) from based on refractometer and hydrometer readings on 35 port wines. In the words of the authors, "It is not applicable for the analysis of dry wines, whether fortified or not, which contain insufficient soluble solids for Baumé determination by hydrometer, and is yet to be evaluated for sweet table wines, such as sauternes." However it is included in FermCalc because many home winemakers seem to find it useful for monitoring fermentation progress and calculating alcohol content.
FermCalc first converts the hydrometer reading sg to degrees Baumé using the following equation.
where Bé is degrees Baumé.
Alcohol content is then calculated as:
av = alcohol content, % by volume
True Brix, Bt, which represents the estimated residual solids content in % by weight, is then calculated as:
Back to top
Son et al. Method
H. S. Son et al. (2009) developed tthe following six empirical equations based on refractometer, hydrometer, and alcohol content measurements on 30 wines before and during fermentation.
av = alcohol content, % by volume
Equations (20) and (21) allow calculation of alcohol content and true Brix directly from hydrometer and refractometer readings. However, I found these equations to be inaccurate, yielding estimates of alcohol content that appear too high in the lower-alcohol range and too low in the upper range. Instead of using equations (20) and (21), I developed alternative equations from equations (15) through (19) that appear much more accurate. Combining equations (18) and (19) to eliminate Bi gives:
Combining equations (16) and (17) to eliminate Bi gives:
Equations (22) and (23) are used by FermCalc to calculate alcohol content and true Brix.Back to top
This method was published online by Werner Roesener (2001) and is very popular among home winemakers, but there is no documentation regarding the derivation of the equations. My testing indicates that it yields results that are very similar to the other methods.
Simplifying the original equations we get:
where s is the dissolved solids content in g/L. FermCalc converts the solids content in g/L to true Brix in percent by weight using equation (26) below.
Back to top
Barth & Race Method
This method was originally developed by Georg Barth (1905) in Germany for analyzing beers. The original equations are:
Where ri is the measured refractive index. The equations were later modified by J. Race (1908) to yield more accurate results for beers with alcohol contents greater than 4.5% by weight.
FermCalc uses equations (29) and (30) to calculate alcohol content and true Brix because they were intended for higher alcohol concentrations and might be more applicable for winemaking calculations. The result of equation (29) is converted to % alcohol by volume as described here.Back to top
Alcohol Content by the Boiling (Spirit-Indication) Method
This method was first proposed by M. E. Tabarié in 1830 as a simplified alternative to the distillation procedure. It is based on the principle that alcohol causes the same depression in specific gravity in wine as it does in pure water.
The method involves evaporating (boiling off) a portion of the wine sample until all of the alcohol is evaporated, and then replacing the evaporated volume with distilled water. The difference between the specific gravities of the wine and the volume-corrected residue are then used to estimate the specific gravity of the distillate, which represents the specific gravity of a pure water/ethanol mixture, from which the alcohol content can be estimated. The experimental procedure is summarized below.
It is recommended that a narrow-range hydrometer be used for the specific gravity measurements since small errors in these measurements can result in large errors in the results.
In addition to estimating the alcohol content, we can also calculate the solids content (true Brix) of the wine from the specific gravity of the volume-corrected residue. First we just need to convert the residue specific gravity measurement to a Brix value by using the Brix conversion equation. This conversion yields the solids content in % by weight of the residue. We then need to convert this value to the solids content in % by weight of the wine by multiplying by the ratio of specific gravities, or:
Bt = true Brix (solids content) of wine, % by weight
The four methods that FermCalc uses to calculate the alcohol content from the specific gravity measurements are described below.Back to top
Tabarié (Division) Method
Tabarié originally proposed estimating the specific gravity of the distillate from the ratio of the wine and residue specific gravities, or:
sgd = specific gravity of distillate
FermCalc determines the alcohol content in % by volume from sgd by using the OIML formula (OIML, 1973).Back to top
Blunt (Subtraction) Method
T. P. Blunt (1891) suggested that the Tabarié division formula always underestimates alcohol content, and suggested calculating sgd from the difference between sgr and sgw instead of the ratio, or.
FermCalc uses the OIML formula to estimate alcohol content in % by volume from the results of equation (33).
S. Harvey (1892) presented experimental results suggesting that Blunt's formula is more accurate than Tabarié's. However, a few pages later in the same issue of The Analyst, A. H. Allen presented data for sugar and alcohol solutions ranging from 24% to 53% alcohol by weight suggesting that the Tabarié formula was more accurate. While Blunt is generally credited with showing that the subtraction formula is more accurate than the division formula, the subtraction formula was clearly in use well before Blunt wrote his paper (Mulder, 1857).Back to top
This method is described on pages 124-126 of The Art of Making Wine (Anderson & Hull, 1970), and is attributed to the researches of Dr. William Honneyman (1966). This method is very similar to the Blunt subtraction method described above. The main difference is that this method uses Dr. Honneyman's table below to estimate the alcohol content, and the Blunt method uses published alcohol tables. The two methods yield very similar results, but the Honneyman method yields slightly higher alcohol content than the Blunt method. Whether this is by design or not isn't clear. Anderson & Hull state that "The method as described here lacks some of Dr. Honneyman's refinements, but it yields results close enough for amateur winemakers." What those refinements are remains unclear. (I've been trying to get a copy of Dr. Honneyman's original report but so far I've been unsuccessful.)
If the difference (sgr - sgw) is greater than the maximum value in the table, FermCalc extrapolates the table using the OIML formula as a guide.
James Hackbarth (2009) showed that the Tabarié approach works reasonably well for dry wines and low-extract beers, but is inaccurate for beverages with higher alcohol and sugar concentrations due to solute-solute interactions that take place at the higher concentrations, an idea that was first proposed by Leonard (1897). Based on extensive laboratory experimentation and detailed analysis of the results, Hackbarth (2011) developed a new model for estimating the specific gravity of a mixture from its sucrose (extract) and alcohol concentrations. Since the spirit-indication procedure gives us measurements of the wine specific gravity and the extract concentration (true Brix), we can treat these as known quantities and use the Hackbarth model to solve for the alcohol content.
The Hackbarth model utilizes the following equations.
Z = correction for solute interactions
FermCalc solves equations (34) through (40) iteratively using the alcohol content calculated from the Blunt model as the initial estimate. The iteration loop is repeated until the difference between the value of SG calculated by equation (40) and the wine specific gravity sgw is less than 10-8.Back to top
The OIML Calculator uses the general formula for calculating the densities of mixtures of ethanol and water found in International Recommendation 22: International Alcoholometric Tables by the International Organisation of Legal Metrology (OIML, 1973). The general formula is:
ρ = density of ethanol/water mixture, kg/m3
Given an alcohol content and a temperature, equation (41) can be solved directly for the density of the mixture. Given a density and a temperature, FermCalc calculates the associated alcohol content using an iterative technique.
© 2007-2013 Steve Gross