A policy of collecting and evaluating all analytical data and taste test results on a company wide basis was adopted by South African Breweries in the early 1980s and remained in force up to the time of writing, enabling long term statistical studies to be carried out. The study described in this paper covers the changes in the composition and flavour of two of the company's leading brands (which are produced in all its breweries) between 1982 and 1994, based on the average analytical and sensory results for each brand and brewery for each six month period during that time. These changes are correlated with changes in the brewing materials and methods used, in some cases revealing that changes made in response to quality problems did not solve these problems because they were based on a misunderstanding of their causes (e.g. when the free amino nitrogen (FAN) content of the finished beer increased, this was initially attributed to excessively high wort FAN, and the FAN level in the malt specification was reduced, but in fact it was later found that the real problem was the poor physiological condition of the yeast, which was consuming less FAN than normal, and this was finally cured by adding a yeast stimulant). Many of the changes described arose from the adoption of high gravity brewing in 1987/87, not because of high gravity brewing as such but because of other changes in the fermentation process which were intended to minimize the increase in beer loss caused by the additional yeast crop volume generated by the more heavily pitched high gravity fermentations, but resulted in sluggish fermentation and a deterioration in beer flavour.
Keywords : beer brewery flavour high gravity brewing history quality statistical analysis