Skim powder, composition

Product intrinsic attributes 1. Perishability and rapid deterioration in quality of milk and milk products with time and temperature. 2. Variability in milk and miUc products due to variation in fat and SNF content of cow and buffalo milk variation due to difference in cattle feed, fodder, genetic composition, breed, etc. 3. Seasonality lack of milk production in lean season due to lack of green fodder lack of milk production due to lack of drinking water for the cattle in lean season use of skimmed powder for production of milk and milk products in lean season high production of milk and mUk products in flush season... 

Four reconstituted milks were prepared by blending hydrated skim milk powder (35g/L) with four different emulsions (35g/L) differing by composition of the fat-water interface. Whole reconstituted milks were coded MP (milk proteins), BCAS ( 6-casein), and BLG5 (j6-lactoglobulin 5 g/L). 

Figure 13.1 is a flow diagram for the traditional processes for the manufacture of WMP, SMP, and BMP. An alternate process for the manufacture of WMP that involves blending highly heated cream and low heat-treated skim milk concentrate prior to drying, was described by Hols and Van Mil (1991). The recommended standards for WMP, SMP and BMP, as well as average composition of the powders, are shown in Table 13.1. 

Table 13.1. Recommended (ADPI, 2002) and actual Mean Composition of Skim Milk, Buttermilk and Whole Milk Powders (USDA, 1999)...

Twomey, M., Keogh, M.K., O Kennedy, B.T., Mulvihill, D.M. 2000. Effect of milk composition on selected properties of spray-dried high-fat and skim-milk powders. Irish J. Agric Food Res. 39, 79-94. 

Figure 19.13 depicts the meltdown behaviour of ice creams with different compositions. In experimental situations, 50% of the skim milk powder was replaced by whey protein in the form of either conventional whey protein or WPP (Koxholt et al. 2001). 

The product composition, i.e. the proportion between protein and lactose in the final product, may be controlled from the central control room. Thus the factory is able to produce protein powder with compositions varying from approx. 35% protein of total solids to approx. 85% protein of total solids. The first mentioned is a product with a composition like cheap skim-milk powder, whereas the latter is an expensive product used for baby-food and dietetic food. 

At the same time, Baer et al. (1) determined water content, protein, and fat in commercial skim milk powders and in the laboratory prepared from nonfat dry milk (NFDM) samples. They used 44 commercial low-, medium-, and high-heat samples and 38 laboratory prepared samples, to increase the composition range. Data indicated that prediction of moisture content by NIR was linear throughout the tested range (2.85-9.70%). They reported that usable predictions of composition could be obtained even when physical factors such as particle size were not constant. They also concluded that NIR calibrations at individual dairy plants probably result in lower SEP values because of the greater physical homogeneity of samples. 

For fat calibrations of skim milk powder it was impossible to get samples with enough variation in fat content. Owing to the standardization of this product the calibration set is not ideal. All calculated correlation coefficients are above 0.98 with an SEC of 0.09 to 0.31 % (Table 20.8). The latter 0.31 % is found for a combined calibration of skim milk powder and buttermilk powder. The standard error is about twice as high than the separate calibrations, which is caused by an increased variation in the sample composition. On the other hand, samples of milk powder with nonmilk fat give a wide variation of fat content. The SEC value is 0.30. By making two calibrations the SEC value decreases to 0.14 on average. The values of SEP and R obtained for the prediction sets are comparable with those of the calibration sets. 

SECs of0.38,0.35, and 0.34% were calculated for the determination of lactose in skim milk powder, buttermilk powder, and milk powder with nonmilk fat (Table 20.9). These are acceptable in relation to the standard deviation of the reference method of about 0.25%. The SEC for the combined calibration of skim and buttermilk powder and for denatured milk powder are too high for accurate analysis. Similar results were observed for the test set. This is caused by the variation in physical and chemical composition of the samples and by the few specific NIR bands of lactose in relation to the overall NIR spectrum. Therefore many filters are needed for the calibration. About ten filters gave significant... 

The performance of reverse osmosis in concentrating milk is limited by the osmotic pressure and most commercial modules have operating pressure limits of 30 to 40 bars, which limits the concentration of milk to a factor of three to four. In the production of skimmed or whole milk powder, the milk is usually concentrated to 45 to 50% total solids before spray drying. Thus reverse osmosis cannot substitute entirely for conventional evaporation, rather it is used as a pre-concentration step before evaporation, to reduce operating costs or to increase capacity of existing plant. The relative energy consumption of reverse osmosis and thermal concentration methods in the concentration of milk differ by an order of magnitude, and with thin film composite membranes the cost of reverse osmosis is lower still. 

---