Predicting cheese yield

The manufacture of many cheeses involves an approximate 10-fold concentration of the solids in milk. Thus, a very rough rule-of-thumb suggests that one should obtain about 1 kg cheese from 10 L milk. Various formulae have been developed down through the years to estimate more accurately the yield of cheese of specific varieties. A widely used formula is the Van Slyke yield formula:

Where Ya is actual cheese yield (kg per 100 kg milk), F and C are the fat and casein contents of the cheesemilk (with added starter culture), respectively, %FR is fat recovery (%FR/100 is often 0.93 for Cheddar cheese), a is a coefficient to account for casein loss (often 0.1 for Cheddar) and b is a coefficient to account for cheese solids non-fat, non-protein (often 1.09 for Cheddar). This formula often underestimates yield of high moisture cheeses. Also, 93% recovery of fat may not be achieved. Factors must be varied for other cheese varieties and can also vary between factories and application of a generic yield formula may not accurately predict yield in all plants. If pronounced seasonal variation, theoretical formulae give less accuracy, mainly due to errors in estimating casein concentration.

Plant specific formulae are developed by analysing historical data on cheese yield in relation to milk composition, milk quality, fat and protein recovery and cheese salt and moisture contents. These must be updated regularly but give very accurate predictions for a particular plant. However, they are not applicable to other plants and are dependent on accuracy of historical data. Also, it is important to remember that accurate prediction does not mean that a plant is operating at maximum efficiency!
Further reading:

Banks, J.M. (2007). How can cheese yield be predicted? In Cheese Problems Solved, P.L.H. McSweeney (ed.), Woodhead, Cambridge, pp. 105-6.