The above graph is a culmination of data from three sources. Two independently run tests isolating mash temperature holding all other variables as contestant as practical. These are blue diamond and green triangle marks. The red squares are data from beers that I have brewed. You can see that the three all have a very similar shape and that all three are roughly the same, however it's where they deviate that I find interesting.
Kai's data shows a maximum attenuation at about 148 degrees Fahrenheit while the Wyeast data shows maximum attenuation at 153 degrees Fahrenheit. There aren't really enough data points in my data to draw a conclusion here, but it seems to peak about where Kai's data is.
The diastatic power of the malt, mash time, and yeast strain, all play into the equation as well. Which likely account for these deviations.
It can be seen from this data that in the typical mash range of 145-160 the attenuation can be varied by 10%. This is the largest contributing factor that I have seen. Yeast strain can vary attenuation about 4%. The use of crystal malt in a typical 10% addition changes atteunation about 3%. And mash thickness and fermentation temperature do not seem effect final attenuation a noticeable amount.
If you use an excel sheet for calculating attenuation then it's easy enough to add the full equation, but a basic rule of thumb will work for most cases. Especially considering that there are a number of other factors involved with attenuation.
Rule of thumb:
1% less attenuation for every degree above 151. (starting with the yeast's average attenuation)
(Just think of Bacardi 151 to remember the base mash temp)
The equation :
The results of my last six brews are shown below. Notice that the mash time effects total apparent attenuation, but not to the extent that of temperature.
Saison Terri II
Below is the data that I have collected with 3 gallon batches of beer.
Below is data from experiments run by Greg Doss of Wyeast. (1)
For completeness Kai's data is summarized here with the same scale as the other data.(2)