Composition and Yield Control
19 Yield Prediction
Purposes of Calculating Predicted Yields
Predicting yields is one of the most important tool to predict cheese processing costs and therefore being able to set pricing and estimate budgets. But calculating yields has many purposes:
- Provide a target against which to judge actual yields and determine mass balance within the plant.
- Flag errors in measurement (e.g., the weight of milk or improper standardization).
- Provide an early signal of high or low moisture content allowing adjustment on subsequent vats. This can be checked with rapid moisture tests (microwave) which is sufficiently accurate for this purpose.
The Van Slyke and Price Formula
The formula widely used for Cheddar cheese and often Colby type cheese is the Van Slyke formula that was published in 1908 and has been used successfully ever since. The Van Slyke formula (Equation 19.1) is based on the premise that yield is proportional to the recovery of total solids (fat, protein, and other solids) and the moisture content of the cheese.
F = Fat content of milk (3.6Kg / 100 Kg)
C = Casein content of milk (2.5 Kg / 100 Kg).
0.1 = Casein lost in whey due to hydrolysis of κ-casein and loss of fines
1.09 = A factor which accounts for the other solids included in the cheese
M =Moisture fraction (0.37)
Exercises
Calculate the Van Slyke predicted yield for Colby cheese made from milk containing 3.3 % by weight fat and 3.2% protein by weight assuming a target moisture content of 40%.
This formula has several important limitations:
- First, the formula is based on casein, but commonly used milk analyzers only measure total protein. There are several solutions for this:
- Insert a coefficient to estimate casein (C) from protein (P). A commonly used coefficient is 0.78 which is commonly used value for casein number, that is the amount of casein as a percentage of total protein. To apply this in the Van Slyke formula replace C with 0.78P. Note that in Ontario, according to a survey of 1097 bulk tank samples that we completed in 1998, the average casein number is 76.1 (Hill, Fulton, Melichercik, and Szijarto, 1998, Direct casein testing with InfraRd instruments, Technical Report to Agriculture and Food Laboratory Services Division).
- The same technical report recommended a simple equation to estimate casein more accurately as: Casein = (0.830 x protein) – 0.199. Then, of course this calculated value can be used for C in the Van Slyke formula. This equation importantly means that the casein number increases (slightly) with total protein. In other words, if we feed or breed cattle to produce milk with higher protein content, the relative proportion of casein in the protein increases and the proportion of whey protein decreases.
- The third solution is to use a milk analyzer calibrated to estimate casein as well as total protein. A precaution here is to ensure that the milk analyzer is regularly checked to ensure proper calibration.
- A second difficulty is that the formula fails to consider important variables such as variation in salt content and whey solids, particularly whey protein. This gets further complicated considering that pasteurization conditions determine how much whey protein is denatured and subsequently included in the cheese causing protein recovery and cheese yield to increase.
- Finally, the equation is specific to Cheddar, although it is often used for American style cheeses such as Colby and Monterey Jack. Accordingly, other formulae have been developed for other varieties of cheese. See Emmons et al. J.Dairy Sci. 73 (1990):1365-139; and, other recommended references).