STUDEES» 0N UNSTAN'HZING AND PROPERTIES OF BNSTA-NTJZED DRY MiLKS‘ 'fhssis far ‘31:» Dwrea oi M. S.- MlCHéGAN Wit—5’5 UNWERSETY Ken-Ec-hi Mari 1964 THES'S LIBRARY Michigan State University ABSTRACT STUDIES ON INSTANTIZING AND PROPERTIES OF INSTANTIZED DRY MILKS by Ken-ichi Mori The purpose of the study was to determine the effect of various product related factors on the characteristics of the instantized dry milks, especially dispersibility and density. The cost of agglomer- ating operation also was investigated. Skimmilk with high and low heat treatments or with various percentages of milk fat was concentrated and spray dried in a horizontal spray drier under controlled conditions. These dry milks were instantized in a Blaw-Knox unit. The particles were wetted with steam, subjected to hot air redrying and conditioned on a belt which delivered the product to the Sifter. After sifting, the product was packaged in moisture proof containers and held at room temperature for analysis by common procedures. In a comparison of instantized high heat and low heat nonfat dry milk the dispersibility was not significantly different (48. O and 47. 6 grams). Average density of 18 trials was higher and moisture increase due to instantizing was 0. 37 percent higher in high heat samples. High heat nonfat dry milk shattered more easily as indicated by size of particles separated by the Rotap sieve test. KEN -ICHI MORI An 87 percent larger amount of particles passed through 100- and ZOO-mesh sieves. Nonfat dry milk tempered to 400, 700 or 900 F. before entry into the agglomerating chamber gave the following average results on 12 trials of instantized samples, respectively: (a) The dispersibilities of 47.4, 48.3 and 47.7 grams were not significantly different. (b) The packed bulk density of 0. 466, 0. 490 and 0. 477 gram per ml. was not significantly different. (c) The moisture increase during agglomeration and redrying was 1.72, 1.38 and 1.16. These are significantly different. Instantized dry milks with l, 2, 5 or 10 percent milk fat varied in dispersibility between 44. 4 and 48. 0 grams (no significant dif- ference), but dry milk with 26 percent fat was 31. 9 grams which was a significantly lower dispersibility. Moisture increase from instantizing was 0. 97 percent for 1 percent fat dry milk and incon- sistently varied from 1. 84 to 2.12 percent for the dry milks with 2, 5, 10 and 26 percent fat. The packed bulk density average decrease of 5 trials was 0.268, 0.237, 0.223, 0.192 and 0.136 gram per ml. for the samples with l, 2, 5, 10 and 26 percent fat. Particles separated into groups according to size by sieve retention on 14-, 20-, 28-, 35-, 48-, 60-, 65- and lOO-mesh screens did not show a significant difference in dispersibility. KEN -IC HI MORI However, particles retained on a ZOO-mesh screen and in the pan had a slightly lower dispersibility. Packed bulk density ranged from 0.223 to 0.613 gram per ml. as the particle size decreased when separated by retention on 14-me sh to the ZOO-mesh sieves or in the pan. A combination of sieve and Coulter electronic counter methods was required to classify particles of agglomerated dry milks. A substantial variation in the percentage of particle sizes was evident in the results of six commercial brands of instantized nonfat dry milk obtained from the East Lansing, Michigan, grocery stores. The projected cost of instantizing nonfat dry milk in the Michi- gan State University Dairy Plant was 2. 1422 cents per pound of which 0. 7230 was fixed and 1. 4192 was variable costs. A commer- cial dry milk plant with an instantizer had fixed cost of 0. 7517 cents per pound, variable of 1. 6776 and thus total costs of 2. 42.93 cents. STUDIES ON INSTANTIZING AND PROPERTIES OF INSTANTIZED DRY MILKS By Ken-ichi Mori A THESIS Submitted to the College of Agriculture Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science 1964 ACKNOWLEDGEMENTS The author expresses his sincere gratitude and appreciation to Dr. T. I. Hedrick, Professor, Department of Food Science, for his encouragement, inspiration, timely guidance and constructive criticism throughout this study. Thanks are also expressed to Dr. C. M. Stine and Mr. A. L. Rippen, Associate Professors, Department of Food Science, and Michigan State University Dairy Plant personnel including office secretaries and graduate students for their ever helping advice to the author. Sincere gratitude is extended to Dr. H. J. Raphael and Dr. C. M. Stine for their services as advisory committee members. ii TABLE OF CONTENTS INTRODUCTION ...................................... LITERATURE REVIEW ................................. I. Production of instant dry milk ..................... A. After dehydration ............................ 1. Agglomeration by wetting ................. 2. Coating and other methods ................. B. Single stage process .......................... 1. Large particle ........................... 2. Single stage agglomeration .. .............. C. Foam drying ................................. 1. Foam mat drying ........................ 2. Foam spray drying ....................... 11. Factors affecting instantizing ...................... III. Change of properties of nonfat dry milk by instantizing ...................................... A. Dispersibility ............................... B. Flavor ....................................... C. Color ....................................... D. Hygroscopicity .............................. E. Nutritive value ............................... IV. Tests and standards of instant nonfat dry milk ...... 10 11 l3 14 15 17 18 19 19 22 23 24 25 26 V. Usage of instant nonfat dry milk ................... VI. Costs ................................ . ........... VII. Packaging ........................................ EXPERIMENTAL PROCEDURES ................ . ........ 1. Preparation of nonfat dry milk samples ............. II. Preparation of dry milk sarriples with various milk fat contents ...................................... III. Instantizing process ............................... IV. Design of experiments ............................. V. Methods of testing the dry milks ................... VI. Cost analysis of instantizing operation ............. RESULTS .............................................. I. Effects of high heat and low heat nonfat dry milk and the temperature of the nonfat dry milk per s_e_for instantizing on the properties of the instantized product ......................................... II. The influence of milk fat in dry milks on the properties of the instantized products .............. III. Particle size analysis ............ . ................ A. Particle size analysis of instantized nonfat dry milk ........................................ B Combination sieve and counter methods ......... C. Properties of instant nonfat dry milk of different particle size ......................... IV. Cost analysis of instantizing operation .............. iv 28 30 30 31 32 33 34 38 39 39 44 51 51 52 53 55 A. PlantA B. PlantB DISCUSSION ........ SUMMARY ......... LITERATURE CITED OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Page 55 60 64 69 71 Table 10. LIST OF TA BLES The effects of high heat and low heat nonfat dry milk and the temperature of nonfat dry milk on the properties of instantized product ................. Summation of analyses of variance for dispersibility, moisture increase and bulk density ............... Correlation coefficients among properties of instantized nonfat dry milk ....................... Influence of milk fat in the dry milk on the properties of instantized dry milk .............. . . Comparison of properties of instantized dry milks with various milk fat contents .................. . Correlation coefficients among properties of instantized dry milk with various milk fat contents . . Classification of size and the resulting properties of commercial instant nonfat dry milks ............ Particle.size analysis of Brand D of instant nonfat dry milk by combination of sieve and counter methods ........................................ Dispersibility and bulk densities of instant nonfat dry milk of various particle sizes ............ . . . . Shattering test results of low and high heat nonfat dry milk after instantizing .............. . ....... vi Page 40 42 43 46 48 49 52 53 54 55 LIST OF FIGURES Figure Page 1. Regression line between bulk densities, loose and packed, in instantized nonfat dry milk ........... 45 2. Regression line between bulk densities, loose and packed, in instantized fat containing dry milks . . . . 50 vii INTRODUCTION Since the first instantized nonfat dry milk was marketed in 1954, the consumption of "instant" nonfat dry milk has increased very rapidly. Data giving the total instantized nonfat dry milk pro- duction were not available. However, practically all nonfat dry milk for home use is instantized. The reasons may have been attributed to its apparently quick dispersibility and solubility, and also to its ready availability. It has been widely used not only for beverage purposes, but also for cookery. Industrial uses have been limited probably due to the higher cost. The amount of nonfat dry milk for home use was 141 . 7 million pounds in 1954, and 255. 3 million pounds in 1963. Home use to the total nongovernmental use was 19. 7 percent in 1954, and it increased to 28. 5 percent in 1963 (3) . Dry whole milk and dry milk with 12 or 8 percent fat also have been available. Sales have been relatively small because of stability problem, cooked or chalky flavor and presumably lack of dispersibility. In spite of the large amount of instantized nonfat dry milk that is currently produced, there is much that is not known about processing factors which influence the dispersibility and other characteristics of the final instantized product. Published informa- tion is lacking. Whatever is known has been restricted to individual company personnel who may have conducted studies. A disadvantage has been the increased cost of the agglomerat- ing process. Furthermore the decrease in density results in added packaging expenses. An objective of this study was to investigate various common product factors that affect agglomeration of dry milks and their influence on the characteristics of the resulting product. Of great- est interest was the effect of fat content, heat treatment, dry milk temperature for instantizing and particle size. Another objective was to ascertain the cost of agglomerating nonfat dry milk. The term "instantizeH refers to altering powder particles so that they disperse readily in water during reconstitution. There are various degrees of effectiveness of instantized products. All prod- ucts called "instant" are not always strictly qualified. Agglomer- ation means, in this study, clumping or clustering of particles of powder into large aggregates. Since the single pass process has been less satisfactory for product dispersibility, agglomerating process has been widely employed. This causes two terminologies, instantizing and agglomeration, to be synonymous in trade usage. Solubility usually refers to the ability of particles to dissolve homogeneously into water, forming a solution. Dispersibility im- plies the ability of particles to distribute as a colloidal state in water. The dispersed substance remains suspended in the liquid. Since milk is a complex of many components, some are soluble such as lactose, and others are dispersible such as proteins. Despite the difference, trade usage of terms, solubility and dispersibility, has been interchangeable. In this manuscript, these two terms are used synonymously. Sinkability refers to the ability of the particle to penetrate the water surface and disperse or settle to the bottom of the quiescent container. LITERATURE REVIEW I. Production of instant dry milk A search of the literature indicated that the manufacture of instant dry milk, especially nonfat dry milk, has been developed mainly in the United States. There are many processing techniques . They could be classified into three major groups: A. after dehydra- tion (two-stage process), B. single stage process, and C. foam drying. A. After dehydration 1. Agglomeration by wetting Most processings were based on the principle of wetting and redrying. Dry milk with the normal small particles was produced by the conventional spray drier. The instantizing equipment, which might or might not be attached to the drier, formed clumps or clusters from many small particles. Nonfat dry milk was moistened by water or steam under specific conditions. The moistened parti- cles stuck together to form aggregates attaining fluffy, porous, sponge -like structure. These aggregates were redried to the de- sired moisture level by hot air. Several patents have been granted to inventors of equipment and methods under this classification. Although the principle was the same, there were many differences in machine design, such as method of wetting and redrying, maximum moisture content of wetted dry milk and the claimed change of physical and chemical state of lactose and protein. Peebles (44) has been considered the pioneer in the develop- ment of instant dry milk. Product from his process was first introduced extensively throughout the United States in 1954. Ac- cording to the description of apparatus by Peebles (43) and Havighorst (22), the dry milk was dispersed as it entered the processing chamber. In the chamber, the dry milk came into in- timate contact with an air stream containing finely atomized water and steam. The moistened particles stuck together to form aggre- gates. Their moisture content was adjusted between 10 to 20 per- cent. The moisture content was reduced to the desired level by means of heated air within a shaker-drier. The inventor claimed that substantial amounts of lactose changed into the crystalline form. The Cherry-Burrell system was described by Carlson eta}; (9) . The dry milk was agglomerated in a horizontal chamber with moist air moving at high velocity. The moisture content increased 6 to 10 percent. Wetted particles were redried in a vertical tube of increasing diameter. Warm particles were cooled on a shaker- table, then sized between two stainless steel rolls. Griffin ( 15) was granted a patent for the Blaw-Knox method. This machine was cone shaped and very compact. Two balanced straight -line tubes with jets of stream were directed downward against each other at an angle of about 40 degrees from a vertical plane. Dry milk was passed in a free falling dispersion along this plane into the zone of collision of the twoijets. The particles were moistened and dropped through a turbulent heated air drying stream onto a belt. Redried particles were conveyed by an inclined belt, on which cooling occurred to a gyrating Sifter for sizing. Lauder and Hodson (33) developed a modified method for instantizing products. A stream of dry milk containing less than 4 percent moisture was exposed to steam, causing surface moistening and agglomeration. The moisture content of wetted powder might increase to 9 percent, but preferably to 5.5 percent. The aggre- gates were immediately brought in contact with a hot air stream. The inventors claimed that the instantized product was characterized by a relatively high beta-lactose and a low alpha-lactose content which was thought to improve the dispersibility of the product. An U.S. patent has been given to Bissell (5) for another modifi- cation of instantizing process. Dry milk was fed onto a moving belt from which it fell as a curtain into the path of a row of steam jets. The aggregates fell on the second belt and were carried into a dry- ing chamber. The inventor claimed that a substantial portion of lactose in the dry milk, which was in the beta-anhydrous form, had been changed to the alpha-hydrate form in the instantized dry milk. Scott (50) received a patent for a novel method of agglomerat- ing wetted milk particles. Dry milk wetted in a hydration chamber with a fine spray of water or milk to a moisture content of 10 to 14 percent was then conveyed to a hammer mill within 15 to 20 seconds. During this time, according to patent claims, protein swelled and lactose was hydrated and absorbed its required water of crystalli- zation. The agglomerates were redried on porous belts which traveled through drying ovens. In the process invented by Shields (52) , nonfat dry milk was moistened by water to approximately 15 percent and held for a suf- ficient period to convert a substantial portion of the lactose into the alpha -hydrate form. The product was dispersed in a stream of air and mixed with normal nonfat dry milk particles in such proportions that the additional particles occupied about 25 percent of the total solids in the final product. The inventor claimed the resulting clumps were highly soluble in cold water. At the present time, instantizing has been mostly confined to nonfat dry milk. Hall and Hedrick ( 17) , however, reported that a small amount of instantized dry whole milk, dry milk with 12 per- cent fat and dry buttermilk were being marketed in the United States, and an 8 percent fat dry milk in Canada. A patent which provided specifically for the manufacturing of an instant fat containing dry milk product was granted to Peebles (45) . This process comprised the same steps used for instantizing nonfat dry milk (44) , followed by spraying a butterfat -water emulsion over the porous agglomerates of the instant nonfat dry milk. Ab- sorption of the butterfat was promoted by heating the fat and apply- ing it in atomized form, and by having the dry milk temperature above the melting point of the fat. The amount of fat added to the in- stantized nonfat dry milk could be adjusted to as much as 70 percent. Kennedy (28) reported in his patent the use of a protein modi- fying enzyme such as rennin or papain in fat containing milk con- centrate having a solid content from 5 to 70 percent and a fat con— tent from 12 to 65 percent dry basis. The mix was kept at the enzyme active temperature. The viscosity of the mix decreased first, then increased. Enzyme action was stopped by rapid heating to an enzyme inactivating temperature when the viscosity had risen from a minimum to at least its original value, but not to a point of greatly accelerated increase. Then mix was dried, producing easily reconstitutable fat containing dry milk. Another patent specifically referring to fat containing dry dairy product was granted to McIntire and Loo (38) . Their method com- prised the homogeneous mixing of lecithin with a fat containing dry dairy product. Then, individual particles were moistened to an extent sufficient to render them sticky and caused these sticky par- ticles to adhere together in the form of random shaped aggregates of a size substantially greater than the size of individual particles. Finally excess moisture was removed. 2. Coating and other methods Sjollema (54) described a method of coating nonfat dry milk or whole dry milk with a surface ~active agent such as oleic acid or soya lecithin to produce a product with instant properties. Warm dry milk was mixed with the liquid surface active agent in an amount of 2 percent by weight for about 5 minutes, then cooled. Inventor claimed that the finished product showed excellent dispersing action in cold water. Budding (7) also received a British patent for using surface active agents. His process consisted of mixing the dry milk with 0.2 to 4. 0 percent of a water-free liquid surface active agent. Suggested surface active agents were 1 percent oleic acid for non- fat dry milk, 2 percent of soya lecithin at 700 C. (1580 F.) for dry whole milk, 4 percent of a mixture of 1 part of glyceryl monolaurate and 2 parts of olive oil for dry ice cream mix, and 1 percent of a 20 percent solution of egg lecithin in petroleum ether for dry baby food. Niro Atomizers, Ltd. (42) developed an instantizing unit mounted under the conical drying chamber. Dry milk was dis- charged into the unit with a moisture content of approximately 8 to 10 percent. The instantizer was provided with special perforated plates which were divided into three sections. The plates were 10 vibrated, resulting in a forward movement of the powder particles toward the point of discharge as air passed upwards through the perforations. The first section served as an agglomerating com- partment giving the product time to cluster. The second section served for the purpose of redrying the product to the required moisture content. In the last section, cold air was used to cool the instantized product. The atomizer of the drying chamber was pro- vided with double outlets. One set was for concentrate and the other for fines. The fine particles were exhausted back through the atomizer for adherence to the concentrate spray; thus, some pre- agglomeration occurred in the spray drying chamber. Recently an attempt has been made at TEKNIKA Inc. (47) to utilize sonic energy in agglomerating dry milk. Dry milk was intro- duced into stainless steel sonic column, in which intense sound was generated. Dry milk was agglomerated by the use of sonic energy and a slight amount of steam. The agglomerates formed blocks which entrained a large amount of air. The exposure to sonic energy was in terms of milliseconds. The block was either cut or ground to the size of particle desired. Since the exposure time was short, the inventor claimed that the final product did not have an overcooked flavor. B. Single stage process The single stage process does not require additional reprocessing ll equipment. This process could be classified into two groups: 1. large particle dry milk that depends on size and density for reconstitution properties and 2. agglomeration of particles within the drying chamber. Some equipment can perform both to a limited extent. 1. Large particle In order to dry a larger particle, longer drying time or higher drying temperatures were required. Since heat damage of protein places a limit on the temperatures that can be used, a longer drying time was often employed. A tower-type drier, which allowed the particle to remain suspended in the drying air for longer periods of time, was more ideal for this purpose. Coulter and Townley (12) were granted a patent for the Coulter spray drier which could be operated to produce a large particle in— stant dry milk by varying conditions of drying such as large nozzle orifice, lower feeding pressure and a drying temperature adjustment to affect the required drying. Sanna (48) designed a tower-type spray drier suitable for single stage instantizing. This tower was unique in the method of heating. Two or more heat zones were predetermined, depending upon the moisture content of the drying particles. The drier was tall with a capacity for more than one pressure nozzle delivering the concen- trated milk near the top of the drying chamber. The falling particles 12 passed through at least two heating zones. In the first zone, the heat was more intense. The heating source in the first zone was electromagnetic wave energy as in the microwave range. This in- cluded several magnetron tubes arranged at the top of the tower above the spray nozzles. As the particles fell, they passed through an arrangement of baffles which prevented heating energy in the first zone from passing into the next zone. In the second zone, radiant energy was supplied by a series of infrared lamps. If nec- essary, additional heating zones could be added. The inventor also suggested that the whole system could be held under a partial vacuum to aid in drying products that are easily heat damaged. The single stage process by Rice (46) involved evaporating milk to total solids of 40 to 45 percent, heating the concentrate to 1650 F. for 20 seconds, and cooling to 80° to 950 F. in a cooling vat down to 400 to 450 F. About 1.5 gallons of concentrate containing partly crystallized lactose from a previous batch were added to each 1, 000 pounds of concentrate in the vat. The mixture was agitated to produce a high viscosity. The concentrate was introduced into a spray drier having a nozzle with a relatively large orifice at low pressure and at a temperature of concentrate from 1400 to 1450 F. Inlet drying air temperature was sufficient to obtain an outlet air of 2040 to 2100 F. Dried particles of relatively large size, but quite dense, were obtained. l3 2. Single stage agglomeration The University of Wisconsin drying tower is probably one of the most versatile driers used for milk today. This tower is suitable for the production of large particle instant and/or single stage agglomeration (4) . Agglomerates were produced by using a suitable arrangement of five nozzles in the top of the drying chamber. Four of them were arranged horizontally, radiating out 12 inches from the circular common wall and 90 degrees apart. The fifth nozzle was placed 36 inches below the center of the other nozzles. The effect was to subject a mist of relatively dry particles from upper nozzles to a mist of relatively wet particles from the lower nozzle. This resulted in an agglomeration of the particles. Amundson (4) reported that under such an arrangement, moistures ranged from 6 to 7 percent and that further drying was required outside the main drying chamber. Process described by Spiess and Sullivan (56) involved pro- jecting one stream of an inert gas carrying the powder particles and another stream of an inert gas carrying an atomized wetting liquid into an agglomeration zone at high speed so that the two streams impinged on each other at an angle greater than 90 degrees, prefer- ably 180 degrees. The chamber was designed to maintain these powder particles and liquid in suspension for a sufficient period of time to form agglomerates. 14 Method by Sharp and Kempf (51) was a single operation without rewetting. The process was comprised of three drying stages; in the initial spray drying stage concentrate containing a predetermined quantity of lactose seed crystals was sprayed through a hot drying air stream. The moisture content of the particle was reduced to 10 to 18 percent at a time when both crystalline lactose and noncrystal- line lactose were present in the particle. Intermingling of the par- ticles at this stage caused agglomerates to be formed by virtue of the tackiness of the noncrystalline lactose. These agglomerates then passed through the second and third drying stages where the moisture was reduced to the desired content.“ Alfa Laval Co. (1) was granted a patent on a unique drier. The process relied on the recirculation of fine particles through the dry- ing chamber to increase the particle size and, thus, improve the Sinkability during reconstitution. In the drier, a suction element was installed above dry milk discharge which evacuated the fines and rediverted them into the atomizer. The atomizer had special open- ings for dispersing fines above the normal liquid discharge. C. Foam drying Foam drying was developed to minimize flavor defects and im- prove reconstitution of dry whole milk. Foam drying can be divided into two categories: 1. foam mat drying and 2. foam spray drying. 15 1. Foam mat drying This process is also called film drying or puff drying. Sinnamon eta}; (53) reported on the manufacture of dry whole milk that was dried under high vacuum and low temperature with a sponge -like structure: milk was homogenized at 1450 F. and 2,500 p. s.i. , and pasteurized at 1620 F. for 16 seconds. It was concen- trated to 47 to 50 percent total solids at 850 to 1000 F. in a high vacuum falling film evaporator, then heated to 1350 F. and homog- enized at 4, 000 and 500 p. s.i. Before the concentrate was homog- enized, nitrogen was entrained in it. The gas-concentrate then flowed over drying pans and was cooled to 550 F. The concentrate expanded as the pressure was reduced. A collapse of the sponge- like expanded structure in the early drying stages was prevented by maintaining quite low temperatures (near freezing point) and allow- ing some evaporation to take place before applying heat and drying. Tamsma gal. (59, 60) further studied this type of drying. Milk was standardized to a solids -not-fat to fat ratio of 2.75:1 and preheated in a hot water jacketed vat under continuous agitation at 1450 F. for 30 minutes. The milk was concentrated at 1150 F. to 160 Baumé and standardized to 50 percent total solids. The con- centrate was homogenized at 1350 F. and nitrogen was incorporated at the same time by adding approximately 20 ml. of nitrogen (at room temperature and l atmospheric pressure) per liter of l6 concentrate through a fine capillary in the line immediately prior to the homogenizer. The concentrate was cooled to 500 F., dispensed onto a vacuum shelf of a drier and dried fo’r 3 hours at below 1 mm. pressure. The temperature of the product was kept below 1200 F. Dry foams were comminuted through a 20-mesh screen. A patent was granted to Winder and Kielsmeier (63) for foam mat drying. Pasteurized, homogenized milk concentrate was cooled down to approximately 500 F. when substantially all of the fat was solidified. The concentrate was dried by electrical heating while attempting to maintain the fat in the solid state. The temperature of milk concentrate during the drying operation should be maintained below 950 F. until the solid content of the dried particle was at least 70 percent, and then, below the value specified by the equation: Y = 47 — l5. 4 log 10X where Y is the temperature in centigrade and X is the time in minutes, until the solids in the particle reached about 87 percent. This method, however, was not suitable for commercial milk drying as it was a batch method that required 80 minutes drying time. The Chain Belt Company manufactured suitable equipment for the drying of whole milk. This machine was described by Fixari E]; 3.1.: (14) . It was referred to as a vacuum dehydrator and has been used for dehydrating heat sensitive materials. The vacuum dehydrator consisted of a horizontal vacuum chamber, in which an 17 endless stainless steel belt passed over a heating drum at one end and a cooling drum at the other end. The drying material was sprayed on the belt which moves over or under‘a series of radiant heaters, over the heating drum, past another series of radiant heaters, each of which could be individually controlled for heating intensity, and then over the refrigerated cooling drum to complete the cycle. The cooled, dried product was scraped off by a blade and fell into a tube. An air lock permitted periodical removal of the product. The dehydrator has been acceptable for the drying of instant coffee and citrus juices. This dehydrator permitted a cons- tinuous drying operation. Drying time for 'milk would be only a few minutes, possible three or four. Morgan gal. (40) described a modified technique. Milk was. concentrated to about 36 percent solids, edible stabilizer was added, the concentrate cooled, whipped to produce a foam which was then spread on a drying tray or belt, dried by heated air at 1300 to 1900 F. , screened and compressed. The inventors claimed that the dried products made by this process had a unique, fine structure and were easily reconstituted in cold water. 2. Foam spray drying After milk was concentrated to 45 to 60 percent solids, gas such as nitrogen or air was incorporated by high pressure between the pressure pump and the spray nozzle by means of a mixing 18 chamber in the line. The resulting spray dried particles were very porous with a low density, but were easily dispersed in water. Hanrahan and Webb (18) reported the above process was suit- able for drying of whey with high acidity. Hanrahan gal. (19) also reported on the production of foam spray dried whole milk. Pro- cedures were almost the same as for cottage cheese whey except nitrogen was used. The flow rate of nitrogen into the mixing device was determined by a flow meter and manually controlled by a needle valve. The high pressure pump was equipped with a spring—loaded bypass valve which allowed the utilization of various nitrogen- concentrate ratios, maintaining a constant pressure on the spray head. The resulting particles, although of light density and large size, had improved dispersibility on reconstitution. II. Factors affecting instantizing Several factors affect instantizing such as volume and velocity of air current, amount and dryness of steam used, size of chamber, etc. Takano (58) theorized from the electron-microscopic photo- graphs that regular nonfat dry milk particles with crust (formed by low drying temperatures and longer time) on the surface were dif- ficult to moisten. High drying temperature and short drying time tended to form porous particles. Air bubbles within the particles influenced the density of nonfat dry milk. He stated that if the 19 particles were light, the particle traveled too fast and the frequency of the collision decreased. If too dense, the particles moved slowly, then faster air velocity and more air would be needed to get suffi- cient movement. He concluded that there should be a certain range of density of regular nonfat dry milk according to the design of the instantizer . The bulk density depended upon the shape and inner structure of the powder particles (32) . The spherical particles of spray dried milks permitted a closer packing than flaky shape of roller dried milk. The apparent density of spray dried nonfat dry milk ranged from 0. 5 to 0. 6 gram per ml. The bulk density of spray dried milks varied according to the type of plant and degree of precondensation of milk. Spray dried milk with a minimum of occluded air has a bulk density as high as 0.8 gram per ml. III. Change of properties of nonfat dry milk by instantizing A. Dispersibility The reasons nonfat dry milk is improved in dispersibility and wettability by agglomeration are several in number. Literature re- view indicated that the improvement could be due to the increase of the weight of agglomerated particle. The water absorbing capacity caused by porous structure of the agglomerate increases diSpersibility. Bockian e1: a_1_. (6) made a study of factors affecting dispersibility 20 of nonfat dry milk before and after instantizing. Dispersibility time of particles with various sizes was observed. Dispersibility time was defined as the time lapse between the initial contact of the par- ticle with the water and the complete submergence of the particles beneath the water surface. In this sense, dispersibility meant wet- tability rather than its usual connotation. They found a general cor- relation between particle size and wettability. Protein nitrogen distribution, chloride, total alpha- and beta-lactose, sodium, potassium and calcium were investigated. Slight changes in the lactose distribution and in the protein distributions were considered. insufficient to account for the increased dispersibility of instantized dry milks. They concluded that the wettability of instant dry milk resulted mainly from the distribution of salts, lactose and proteins, and also particle size. The successive washings of instant dry milk indicated that a reorientation of salts and lactose toward the particle surface occurred. They also concluded that the less solu- ble components such as calcium and protein were oriented toward the center of particles. The larger size and shape of the particles kept the individual particles farther apart as they reached the sur— face of the water. These heavier individual particles allowed more rapid breaking through surface tension of the water. Schulz (49) stated that powder particles with less than 50 mi- crons had in general very poor wettability. The larger the particle 21 size, the more wettable it was. He suggestedparticles should be 100 to 250 microns for purely soluble materials such as sugar and detergent. But, dry milk, because of its colloidal structure, needed to contact water as much as possible. He suggested particle size of 30 to 100 microns for dry milk with instant characteristics. Loftus (34) compared commercial instant dry milks by micro- scopic study. He found that the particles of the most excellent dispersibility were larger than those of others and contained strong birefringent of many minute lactose crystals uniformly distributed throughout the particle. Harper _e_t_a_l_. (21) investigated the reconstitution rate of in- stantized nonfat dry milks. They used various times of contact of the dry milks with water and also analyzed the amount of water retained by the particles. Instant nonfat dry milk increased its water retention and usually decreased the rate of the solids passing into solution. They explained that contact between dry milk and water caused at least two reactions, one leading to the dispersion of the milk solids, and the other to the destabilization and insolu- bility of the milk proteins . Starey (57) noticed a slight increase of temperature when instant nonfat dry milk was reconstituted in distilled water at 200 C. Instant dry milk of United States origin, other dry milks and lactose monohydrate were used. There was a general tendency for the 22 change of temperature to vary inversely with the expected degree of hydration. He reported that the samples of instant dry milk showed rapid dispersibility but appeared to be incompletely hydrated. Howard e_t_a_1_. (25) investigated the surface areas of dry milks produced by various drying methods. The specific surface area, estimated permeametrically, of whole and nonfat dry milk was up to 9, 200 cm2 /gm. for spray dry milks and 2, 400 to 5, 600 cmz/gm. for instantized and foam dry milks. Bockian 919i: (6) found that a fairly large amount of anhydrous, amorphous lactose in nonfat dry milk and a ratio of alpha to beta crystallization of 40:60. The amorphous lactose absorbed water and became viscous during instantizing. Lactose was crystallized in instantized nonfat dry milk with the ratio of alpha to beta revers- ing to 60:40. Amorphous lactose dissolved quickly over the surface of water and increased the surface tension of water, this hindered the rest of the particles from being wetted, lactose in crystalline form such as alpha lactose dissolved more slowly, thus helping the wetting of the rest of the dry milk. B. Flavor Keeney and Bassette (27) mentioned that instant nonfat dry milk exhibited a more intense cereal-like flavor in reference to their observations of chemical changes. Mann and McIntire (35, 36) found that the flavor acceptability 23 was almost the same for noninstant and for instant nonfat dry milk with a slight decrease during storage. They also performed flavor acceptability tests for paired samples under different storage con- ditions. Hedonic scores for instant and noninstant nonfat dry milks were initially 5. 6 and 5. 7 respectively, and were 4. 7 and 4. 8 after storage at 1200 F. Triangular taste test indicated no significant difference either before or after storage at 1000 F. and 1200 F. They concluded that the instantizing process did not adversely affect the flavor of nonfat dry milk. The flavor of the instantized nonfat dry milk was affected by preheating temperature of the skimmilk according to .McDivitt gal. (37) . Instant nonfat dry milk from milks forewarmed at 1650 F. and 1750 F. were superior to those heated higher or lower in temperatures. Preheating to 1550 F. resulted in a product with the maximum scores for off flavors. C. Color Bockian eta. (6) observed some browning reaction in the instant dry milk. They thought the carbonyl compounds, the re- sultants of casein and lactose reaction were more rapidly dispersed than casein itself. Keeney and Bassette (26, 27) found that the instantizing proc- eSs promoted the development of browning-type reactions by their analysis of 5 hydroxymethylfurfural, reductone contents and 24 indophenol reducing values immediately after reconstitution of in- stant nonfat dry milk. Kumetat ital. (31) also reported that the instantizing process favored the development of the browning reaction. This was con- cluded by estimating ferricyanide reducing values of the protein. The extent of the browning reaction appeared to correlate with the amount of wetting and the necessary subsequent heating to redry. Mann and McIntire (35) made a comparative investigation of flavor, acid ferricyanide reducing values, thiobarbituric acid value and color of instant and noninstant nonfat dry milks during storage. The color of noninstant was: slightly lighter than that of instant milk. Average initial acid ferricyanide reducing value was 0. 471 and 0.232. and the thiobarbituric acid value 0. 191 and 0. 085 for instant and noninstant, respectively. Mann and McIntire (36) extended their study on paired samples under various storage conditions. The average thiobarbituric acid values for the instant nonfat samples were 0.210 before and 0.244 after storage at 1000 F. The average ferricyanide reducing values were similar to those obtained with the thiobarbituric acid method. D. Hygroscopicity Studies in this aspect have been scarce. Sone 33%. (55) in- vestigated saturated vapor pressure curves of instant nonfat dry milk at various temperatures. These curves did not show the 25 hysteresis. This would account for the fact that capillary texture with porous structure was deteriorating. Their research showed that instantized dry milk was less hygroscopic than regular dry milk. The reasons were not explained completely, butthe inter- ference of hygroscopicity by the presence of crystalline lactose and/ or decrease of undenatured portion of protein were offered as two pos sibilites . E. Nutritive value Hodson (23) reported the investigations of nutritional compari- son of paired samples. One was noninstantized and the other was instantized. The results showed that there were no significant dif- ferences in thiamin, riboflavin, niacin, pantothenic acid and vitamin B6 contents between instantized nonfat dry milk and fresh milk; nor did the conversion of a low heat nonfat dry milk into an instantized product by a commercial process affect the content of these vitamins. Only the ascorbic acid content of instant dry milk tended to be lower than that of the regular nonfat dry milk. Rat growth tests indicated that the protein quality in instant dry milk was similar to that in low heat nonfat dry milk or pasteurized milk. The next year, Hodson and Miller (24) investigated the nutritive value of stored samples of instant nonfat dry milk. They stated that the rat growth tests showed no decrease in the nutritive value of the protein of samples of instant nonfat dry milk stored for periods of 26 two years. No loss of lysine was observed in samples stored in containers impervious to moisture. Samples stored in a cardboard container where the moisture had increased from 3. 8 to 5. 8 percent showed a slight loss of lysine, but the rat growth tests indicated no loss in its nutritive value. IV. Tests and standards of instant nonfat dry milk The instantizing of nonfat dry milk does not improve the quality of resulting product. Therefore, the manufacturer has to strictly specify the quality of nonfat dry milk to be used. In general, these requirements are: (a) low or medium heat nonfat dry milk with un- denatured whey protein nitrogen between 2 to 5 mg. per gm. , (b) large particle size, and (c) the U.S. Extra Grade. The American Dry Milk Institute and the U.S. Department of Agriculture have established standards for Extra Grade instant nonfat dry milk (2, 61). V. Usage of instant nonfat dry milk Because of its superior dispersibility, usage of instantized dry milk in many kinds of food would be very possible. The disadvan- tages would be the more expensive price and greater bulk than regu- lar nonfat dry milk. These have not been serious disadvantages for use in the home. Consequently, practically all the more than 250 million pounds used at the retail level have been the instant type. One report of usage of instant nonfat dry milk in commercial 27 scale cottage cheese production has been made by Kosikowski (30). Cottage cheese was made from reconstituted skimmilk of instant nonfat dry milk. He stated that cottage cheese of good or excellent flavor was obtained. Because of the higher moisture retention (com- pared with regular low heat dry milk) during cooking, the body characteristics tended to result in a pasty body. Yields of uncreamed cottage cheese curd, adjusting to an 80 percent moisture content, averaged l. 64 and 1. 72 pounds per pound of dry milk for low heat and instant nonfat dry milk samples, respectively. VI. Costs Hall and Hedrick (17) reported that from past experience, in order to achieve realistic operating costs, a plant had to produce at least 16, 000 pounds of instant nonfat dry milk per day. Some plants produced 30, 000 pounds a day (with a 2, 000 pounds per hour stand- ard unit). With a nominal investment for building, some plants were observing an operating cost of 0. 5 to 0. 75 Cent per pound of finished product for an efficient operation. The estimated total cost of instantiz- ing nonfat dry milk in several plants ranged from 1. 5 to 2. 0 cents per pound in 1958. The total processing cost was 6.8 cents per pound. The cost for instant nonfat dry milk to consumer in 1958 was approximately 47.2 cents per pound, with the allocation as follows: to producer 6. 9; to processor 13. 7; to distributor 36.1; to retailer 38.5; and to consumer 47.2 cents. 28 The operation cost in one commercial company in Japan was ob- tained (64). The cost of regular nonfat dry milk was 29 cents per pound and instantized nonfat dry milk was 38 cents per pound. The difference of 9 cents was considered to be instantizing cost (includ- ing packaging). The breakdown of the costs of instant nonfat dry milk in percentage was as follows: raw material (regular nonfat dry milk) 77.7; packaging materials 15.. 0; fuel and utilities 0.2; labor 1. 5; depreciation 3.5; indirect costs 2.1 percent. These were based on annual production in 1962. The amount of instant nonfat dry milk produced was approximately 7. 9 million pounds in a year, 25, 000 pounds per day. VII. Packaging Packaging in a quantity to reconstitute into one quart is not as popular as it was for regular nonfat dry milk, due to the greater cost because of the lower density. Myrick (41) stated that a p0pu- lar size for instant type was the 0. 6 pound package which makes 3 quarts of nonfat milk at a cost to the housewife of from 23 to 30 cents. Larger packages of 1. 6 pounds for 8 quarts, and 2. 4 pounds for 12 quarts of reconstituted product have been in demand. For commercial and industrial usage, the powder is placed in fiber drums holding about 100 pounds or in bags holding 25 or 50 pounds. The carton type of retail packages are constructed of bleached, rigid paperboard. They have a separate liner consisting of a 29 polyethylene laminated paper. Aluminum foil laminated to paper overwaxed with a mixture of polyethylene and wax is another popu- lar liner or overwrap. EXPERIMENTAL PROCEDURES 1. Preparation of nonfat dry milk samples Most samples of the nonfat dry milks were manufactured in Michigan State University Dairy Plant using equipment with com- mercial characteristics. Samples of instant nonfat dry milk on the market were purchased randomly from supermarkets in the East Lansing area. Samples of five different brands with the intention of each representing a different processing system were obtained. Skimmilk was separated from the milk of a common source, pas- teurized at 1450 F. for 30 minutes, and divided into two portions. One portion of skimmilk was processed for low heat and the other portion was manufactured to obtain high heat treatment. The skim- milk concentrates were dried using two pressure nozzles, in a tear- drop frame horizontal drier with an auger in the bottom. Two sets of nonfat dry milk from low and high heat treatments were manu- factured for the instantizing trials. An attempt was made to dry the milks to 3. 0 percent moisture. Processing details were: Low he at ' High heat (a) Skimmilk preheat o 0 temperature 145 F. 195 F. (b) Hotwell o o . temperature 145 F. for 1 min. 190 F. for 20 min. 30 31 (c) Evaporator steam chest 155° ’5 5° F. 165° t 5° F. o o Vapor chamber 115 F. 115 F. (£1) Total solids of concentrate 38 - 42% 38 - 42% (e) Concentrate preheat o o o 0 temperature 125 - 145 F. 155 - 165 F. (f) Pressure to spray nozzle 2,500 - 2,700 p.s.i. 2,700 - 3,000 p.s.i. (g) Inlet air temperature 2900*; 50 F. 2950 t 50 F. (h) Outlet air 0 o o 0 temperature 200 t 5 F. 205 t 5 F. (i) Nozzle diameter 0. 0292 inch 0.0292 inch (j) Core No. 20* No. 20* To study the effect of dry milk temperature for instantizing, the samples were held at 900, 700 and 400 F. until temperature equilibrium occurred. 11. Preparation of dry milk samples with various milk fat contents Preparation of dry milks with different fat percentages was car- ried out according to the low heat treatment pattern. First, milk was separated into cream and skimmilk. Separated skimmilk was pasteurized and concentrated in accordance with the conditions in nonfat dry milk sample preparation. *Spray Systems Co . 32 Standardization of fat and solids ~not-fat was accomplished after concentration. The required amount of cream was added to the concentrated skimmilk. The mixture was adjusted to approximate— ly 40 percent total solids, heated to 1400 F. and homogenized. Standardized concentrates were spray dried under similar condi- tionsdescribed for nonfat dry milk, except inlet temperature was lowered as the percentage of fat in concentrate increased. Dry milks with 2, 5, 10 and 26 percent fat were manufactured. Nonfat and 2 percent fat dry milks were sifted through 24—mesh screen and 5, 10 and 26 percent fat dry milks through lZ-mesh screen. All dry milks were packaged in 50-pound multi-wall Kraft paper bags with 2—mil polyethylene inner liner. Dry products were stored at room temperature until used for the trial unless otherwise stated. III. Instantizing process Instantizing was conducted in the Blaw-Knox Model 350 Instan- tizer. The instantizer was warmed until outlet redrying air tempera- ture reached 1500 F. Dry milk samples of 25 to 30 pounds were used for each trial. These were dumped into the feed hopper at the top of the instantizer and processing started soon thereafter. The control of processing was performed in accordance with the standard operation: (a) Dry milk feed rate (rheostat dial setting) 60 - 70 33 (b) Pressure of steam injection 3 p. s.i. (c) Inlet redrying air temperature 2950 - 3050 F. (d) Outlet redrying air temperature 1550 - 1650 F. Difficulty was encountered in obtaining a uniform flow of sam- ples with 26 percent milk fat from the hopper to the vibratory trough. Occasionally manual assistance was necessary to obtain a uniform discharge into the vibratory trough for distributing into the agglom- eration chamber. Preliminary trials indicated that 3 p. s.i. steam pressure was not sufficient for agglomerating dry milks with high fat content, such as 5, 10 and 26 percent. The steam pressure was raised to 4 and 5 p. s.i. Five p. s. 1. resulted in excessively high moisture content. Finally, 4 p. s.i. was chosen for 5, 10 and 26 percent fat dry milks to be more comparable to 3 p. s.i. for l and 2 percent fat dry milks. Inlet and outlet redry air temperatures we re. maintained at the same levels for all trials. During a 3- or 4-hour run, a considerable accumulation of milk solids was observed on the metal lining in the agglomeration section of the unit. IV. Design of experiments The first set of experiments. was designed to investigate the effects of temperature of the dry milkEE sion instantizing process and the properties of the instantized product. High heat and low heat 34 nonfat dry milks were studied for the same purposes. The next exper- iments were designed to study the effects of fat content in the dry milk on the agglomeration process and the properties of the result- ing product. Experiments were designed in a way to be able to conduct the analysis of variance. The experiment on heat treatments and tem- perature of dry milk ahead of agglomeration was designed as a ran- domized block containing a two -way factorial. Analysis of variance ( 13) was conducted to find the differences caused by the various factors. In order to determine the difference caused by a change of steam pressure in the trials with dry milks having milk fat the t-test (13) was used. Guenther (16) presented Tukey's method for multi- ple comparison which was used to compare the difference between the average for each condition in the experiments with dry milk having various amounts of milk fat. Correlation and regression (8) among properties of instantized products were investigated. The CDC 3600 computer was used for a portion of the statistical analysis. V. Methods of testing the dry milks The following tests were conducted on the dry milks: American Dry Milk Institute methods (2) were used for moisture, scorched particle score, dispersibility and solubility. The Cenco method (39) also was used for moisture determination. Bulk density in loose and 35 packed form was determined by Harland's method (20). For the particle size analysis of dry milks, two methods were utilized; one was the Coulter electronic counter for small particles; the other was the Rotap sieve method for the large particles. The Coulter counter (11) consisted of a sample stand, an am— plifying and counting unit, and a small vacuum pump. A beaker con- taining the sample suspended in a 4 percent ammonium thiocyanate in isopropyl alcohol (vehicle) was placed on the sample holder in such a manner that a tube with a 280-micron or 400-micron aper- ture was immersed in it. The tube was filled with the vehicle and was connected to a vacuum system which provided for the flow of a measured volume of vehicle, for instance 500 or 2, 000 microliter, through the aperture on the tube. Electrodes were located within and without the tube and a constant voltage maintained between them. The vehicle flowing through the aperture exhibited a resistance which was changed when a particle passed the sensing zone. This change in resistance was amplified and counted. The size of par- ticles was reflected by the magnitude of change in resistance which was manifested as pulse height on an oscilloscope. The number of particles passing through the aperture was counted electronically. An adjustable threshold setting permitted the counting of only those pulses which exceed a given height on the oscilloscope. In this man- ner, the minimum size of particles counted was regulated. To 36 increase the range of pulse height and thus the range of particle size measured, it was possible to decrease the resistance across the electrodes, and to increase the amplification gain. The volume of vehicle passing through the counting aperture was regulated by the vacuum system provided with a traveling mercury column which activated starting and stopping switches preset for known volume displacements . The apparatus was standardized by determining the half count of an uniform size particle sample of known diameter in microns, such as ragweed pollen. The settings of amplification gain, threshold, resist- ance across electrodes were adjusted in order to calculate a calibra- tion factor (k) for each aperture tube. This factor which corresponded to the known mean diameter of the standard was in turn used to cal; culate settings to give other particle diameters. Counting of particles was then made in an unknown sample at these settings, usually starting with the largest particles. Coinci- dence of particles paSsing through the aperture at high counts was compensated by use of a coincidence factor which was a function of the number of particles actually counted, size of aperture and vol- ume of vehicle passing through the aperiture. The adjusted counts were corrected to particle volume factor for each interval in diam- eter. The summation of these factors comprised the total volume of particles in a given volume of vehicle. From these data, the 37 percentage weight distribution was calculated, assuming a uniform density for all particles. The Rotap Te sting Sieve Shaker was used with 8-inch diameter sieves of six different mesh sizes. The designations as U.S. sieve series were Nos. 40, 50, 60, 70, 100 and 200 which corresponded to 35, 48, 60, 65, 100 and 200 meshes, respectively (62) . A 100- gram sample of instantized dry milk was placed on the tOp sieve. The system was closed and samples were submitted to shaking for 5 minutes with horizontal cycle of 300 and tapping rate of 150 per minute. The dry milk retained on each sieveand bottom pan was weighed and percentage of the total was calculated. The retained particles on each sieve were considered to be relatively uniform in size. The sample on each sieve was examined for bulk density and dispersibil- ity. A shattering test was developed because none have been reported in literature. After preliminary study and alteration the following test proved useful: a 100-gram sample was weighed into a Sealright container. It was of 1 quart size and was a resin coated paperboard cylinder. The diameter was 3 1/4 inches and the height was 6 1/2 inches. The containers were sealed with adhesive tape at the joint of cover and body, and wrapped in a polyethylene bag as a precaution against sample loss. The samples were subjected to shattering by 38 rotation in a wooden butter churn at 35 revolutions per minute for 20 minutes. Inside churn diameter was 18 inches. After subjection to the shattering test the samples were classified for size by Rotap sieve method. VI. Cost analysis of instantizing operation The processing cost analysis of instantizing dry milk was con- ducted in the Michigan State University Dairy Plant (Plant A) and a commercial plant (Plant B). In this study costs were limited to those which occurred only for the instantizing operation. The vari- ous items were divided into the fixed or variable costs. Factors considered were: (a) labor cost, (b) utilities, such as steam, elec- tricity and fuel, (c) depreciation, interest, tax, insurance and maintenance of machine and building, (d) miscellaneous supplies, (e) indirect expenses and so forth. RESULTS I. Effects of high heat and low heat nonfat dry milk and the tem- perature of the nonfat dry milk pf}: s_e_for instantizing on the properties of the instantized product (The regular low heat nonfat dry milk had a moisture of 3. 15 per- cent and 0. 515 and 0.770 gram per ml. loose and packed density. The regular high heat nonfat dry milk analysis was moisture 3. 05 percent, loose density 0. 525 and packed density 0.769 gram per ml. Table 1 presents the analysis of dispersibility, moisture and bulk densities of six instantizing trials. The tests were conducted in duplicate. Dispersibility of instantized high heat and low heat nonfat dry milk was similar. The average of 18 trials for high heat nonfat dry milk was 48. 01 grams and for 18 trials of low heat was 47. 61 grams. Instantized high heat nonfat dry milk had higher loose and packed bulk density than low heat nonfat dry milk. The difference in average bulk densities in the loose form was 0. 020 gram per m1. and in the packed form was 0. 028 gram per ml. The moisture increase in high heat nonfat dry milk after agglom- eration and redrying averaged l. 60 percent for 18 trials. The low heat nonfat dry milk increased 1.23 percent. Nonfat dry milk was tempered to 400, 700 and 900 F. and then instantized by the standard procedure. None of these three 39 40 TABLE l--The effects of high heat and low heat nonfat dry milk and the temperature of nonfat dry milk on the properties of instantized product v'T" Trial Pre- Temper- Disper- Moisture Bulk density heat ature for sibility increase treat- I instan- Loose Packed (num- ment tizing ber) (deg. F.) (gm.) (percent) (gm./ml.) (gm./ml.) Low 90 47.35 0.55 0.373 0.510 Low 70 48.78 0. 95 0.338 0.472 1 Low 40 47. 10 1.10 0. 320 0. 463 High 90 48.90 1.35 0.417 0.526 , High 70 48.82 0.95 0.384 0.521 High 40 44. 97 1. 90 0. 362 0. 521 Low 90 49.40 1.30 0.342 0. 500 Low 70 47.00 1.20 0.320 0.446 2 Low 40 47.10 0.95 0.325 0.442 High 90 48.27 1.60 0. 336 0.471 High 70 50.15 1.25 0.370 0.503 High 40 47.49 1.45 0.343 0.495 Low 90 48.17 .,1.60 i 0.324 0.446 Low 70 46.95 1.05 0.324 0.421 3 Low 40 48.40 1.65 0.308 0.446 High 90 49.07 1.35 0.314 0.438 High 70 48.19 2.80 0.354 0.521 High 40 46.67 2.15 0.312 0.446 Low 90 ‘ 44.33 1.30 0.331 0.471 Low 70 45.32 1.05 0.305 0.442 4 Low 40 47.40 1.65 0.312 0.454 High 90 46.12 1.25 0.333 '0.520 High 70 46.62 1.85 0.358 0.510 High 40 48.07 2.10 0.316 0.458 Low 90 49.83 0.35 0.296 0.417 Low 70 48.78 1.40 0.362 0. 537 5 Low 40 47.38 1.65 0.333 0.458 High 90 48.23 1.50 0.324 0.490 High 70 49.45 1.35 0. 348 0.490 High 40 47.61 1. 55 0.336 0.480 Low 90 46.70 0.90 , 0.291 0.454 Low 70 49.97 1.05 0.347 0.526 6 Low 40 47.06 2.50 0.325 0.454 High 90 46.56 0.95 0.350 0.532 High 70 49.33 1.60 0.342 0.490 High 40 49.63 1.95 0.338 0.476 41 TAB LE 1 --C ontinue d m Trial Pre- Temper- Disper- Moisture Bulk density heat ature for sibility increase treat- instan- Loose Packed (num- ment tizing V ' ber) (deg. F.) (gm.) (percent) (gm./m1.) (gm./ml.) Average low 90 47. 63 1. 00 0. 326 0. 466 Average low 70 47.80 1.12 0.333 0.474 Average low 40 47. 41 1. 58 0. 321 0. 453 Average high 90 47. 86 l. 33 0. 346 0. 488 Average high 70 48.76 t 1.63 0.359 0.506 Average high 40 47.41 1.85 0.335 0.479 temperatures (Table 1) seemed to affect the dispersibility signifi- cantly. The average of 12 trials were 47.41, 48.28 and 47.74 grams, respectively. Nonfat dry milk tempered to 400 F. ranged in moisture increase after agglomeration and redrying from 0. 95 to 2. 50 percent. The average was 1.72 percent. Under similar conditions a temperature of 700 F. resulted in a moisture increase that averaged l. 38 percent. Nonfat dry milk tempered to 900 F. showed the least moisture in- crease. The range was 0. 35 to 1. 60 percent and averaged 1.16 percent. The decrease in packed bulk density among 12 trials ranged from 0.248 gram per ml. to 0. 328 gram and averaged 0. 304 gram for the instantized nonfat dry milk tempered to 400 F. At 700 F. the results ranged from 0.233 gram per ml. to 0. 349 gram and averaged 0. 280 gram. At 900 F. the range was 0.237 gram per ml. to 0.353 gram. 42 The average was 0.293 gram. Solubility index of all the samples was less than 0.1 m1. Scorched particle scores were 15.0 mg. (Disc B) or less. The results of analyses of variance for each property are shown in Table 2. TABLE 2—-Summation of analyses of variance for dispersibility, moisture increase and bulk density m F values Factor D. F. . Dispers- Moisture Bulk den31ty value ibility increase Loose Packed Preheat treatment (A) 1 <1.00 9.13** 10.03** 8.60** Temperature for instantizing (B) 2 1.44 6.75** 2. 91 l. 94 AxB 2 <1.00 1: _ _ _ >1< 2 — _ - - _ — __ - - .. 5 _ _ - _ .. - 10 .. _ 26 Note: * Significant at 5 percent level - Not significant 49 The correlation and regression among properties were calcu- lated and results are shown in Table 6. Only the correlation between loose and packed densities was significant at 1 percent level. No significant correlations were found among other combinations. TABLE 6--Correlation coefficients among properties of instantized dry milk with various milk fat contents m Dis er 'bil't Bulk density Moisture p 51 1 y Loose Packed increase Dispersibility __ -0.0779 -0.2437 -0. 1419 Bulk density Loose __ 0. 8589** -0. 0639 Bulk density Packed 0. 0449 Moisture ** Significant at 1 percent level Regression equation was Y = 1.2056X + 0.1016, where Y was packed bulk density in grams per m1. and X was loose bulk density in grams per m1. Regression line is shown in Figure 2. The effect of instantized dry milk with 5, 10 and 26 percent milk fat caused by raising steam pressure 1 p.s.i. was examined. The moisture increase after agglomeration and redrying was greater with higher steam pressures. Bulk densities, loose and packed form, did not vary significantly. The dispersibility of dry milk with 26 percent milk fat increased significantly, from 31 . 90 grams to Packed bulk density (gm./m1.) 0. 560- 0.540- 0. 520~ 0. 500_ 0. 480 -- 0.460~ 0.440- 0. 4201’ 0.400‘r Fig. 2. 50 Standard error of estimate / /' 0.030 gm./ml. / I / l I I I l I I 0.300 0.320 0.340 0.360 0.380 0.400 0.420 f Loose bulk density (gm./ml.) Regression line between bulk densities, loose and packed, in instantized fat containing dry milks. 51 37. 41 grams when the steam pressure was changed from 4 p.s.i. to 5p. s.i. Both of these dispersibility values are below the standard of 44.00 grams. Five p. s.i. caused too high a moisture increase. 111. Particle size analysis A. Particle size analysis of instantized nonfat dry milk Six different brands of commercial samples including Michigan State University product were classified by the Rotap sieve method. The classification of size and properties of each sample is shown in Table 7. Data in Table 7 are the average of duplicate determinations. The moisture content of all samples was less than 4. 0 percent. Loose bulk density of samples was from 0.255 to 0. 450 gram per m1. Packed bulk density also varied from 0. 344 to 0. 526 gram per ml. Dispersibility of all samples was above 44. 00 grams, which was the minimum value designated by American Dry Milk Institute for this factor to qualify for Instant Type, Extra Grade. Sample A had a large percentage on the sieve No. 200 and the same applied to sample E. Samples B and C showed very similar results. Both contained large percentage of coarse particles that remained on sieve No. 40. They had a second peak that was of relatively small particle size (sieve No. 100). Samples D and F were similar to samples B and C. The difference was that samples D and F contained a higher percent of coarse particles and less 52 of fine particles than samples B and C. All samples had the least percentage that remained on the sieve No. 70. TABLE 7--C1assification of size and the resulting properties of commercial instant nonfat dry milks m Sieve Brand (num- Mesh A B c D . E F ber) Classification (percent) 40 35 3.5 28.2 23.8 39.7 1.9 43.3 50 48 8.5 21.6 18.6 9.6 4.0 18.3 60 60 4.9 11.3 10.4 7.3 6.3 6.1 70 65 5.5 5 9 5.7 6.7 7.9 3.9 100 100 14.4 13.3 16.6 9.7 25.9 11.7 200 200 39.5 14.3 16.5 15.9 38.3 10.6 Pan 23.7 5.4 8.4 11.1 15.7 6.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 Dispersibility (gm.) 45.45 46.70 49.00 47.58 47.53 47.07 Moisture (percent) 3.66 3.30 3.80 3.45 3.55 3.30 Bulk density Loose (gm./m1.) 0.316 0.291 0.311 0.284 0.450 0.255 Packed(gm./m1.)0.446 0.373 0.376 0.376 0.526 0.344 B. Combination sieve and counter methods Since the 400-micron aperture could count and classify only particles with a diameter of 200 to 250 microns or less, the larger particles were sifted out by Rotap for 5 minutes with U.S. sieves Nos. 16, 20, 30, 40, 50 and 60. The particles which passed through 53 TABLE 8-Particle size analysis of Brand D of instant nonfat dry milk by combination of sieve and counter methods Sieve number 16 20 30 40 50 60 Opening (mesh) 14 20 28 35 48 60 Classification (percent) 16.0 10.8 9.5 7.3 8.7 4.2 Diameter by counter method (micron) 210 184 146 116 92 .74 59 Classification 7.0 9.5 10.0 8.0 3.9 1.8 0.1 Diameter by counter method 47 39 34 29 23 18 14 Classification 1.0 0.3 0.1 0.7 0.1 0.7 0.3 sieve No. 60 were subjected to the Coulter counter analysis. The results are shown in Table 8. Each value is the average of dupli- cate tests. C. Properties of instant nonfat dry milk of different particle size Instantized nonfat dry milk on each sieve was analyzed for the loose and packed bulk densities and dispersibility. Results are shown in Table 9. Each value is the average of duplicate tests of the sample retained on the designated sieve. The loose and packed bulk densities of instant nonfat dry milk increased as the particle size decreased. Dispersibility of particles on each sieve between 14 and 100 mesh fluctuated between 46. 28 and 54 TABLE 9-~Dispersibility and bulk densities of instant nonfat dry milk of various particle sizes m Bulk density 5:33,? Mesh Disppgrriigility Loose Packed (gm./ml.) (gm./m1.) 16 14 48.56 0. 187 0.223 20 20 47.45 0.192 0.237 30 28 49.15 0.211 0.257 40 35 46.28 0.222 0.302 50 48 46.40 0.257 0.338 60 60 49.45 0.284 0.350 70 65 48.53 0.310 0.379 100 100 47.19 0.357 0.441 200 200 44.73 0.394 0.483 Pan --- 44.44 0.500 0.613 ,— 7 7 49. 45 grams with a lack of a consistent pattern in relation to size. The dispersibilities of the particles held on ZOO-mesh sieve or col- lected in the pan were a trifle lower, 44. 73 and 44.44 grams, respectively. The nonfat dry milks with low and high heat treatments were subjected to a shattering procedure and particles were classified for size. The results are shown in Table 10. Each value in Table 10 is the average of tests in triplicate. Results of low heat nonfat dry milk indicated that the total amount of particles retained on No. 100 sieve or larger in the con- trol sample was 41. 8 percent, but the percentage decreased to 27.4 55 TABLE lOwShattering test results of low and high heat nonfat dry milk after instantizing . Low heat High heat 83:6 Mesh nonfat dry milk . nonfat dry milki ' ($133531) 52:23:23? (5:31:36 5254:2214 40 35 10.3 7.0 1.7 1.3 50 48 9.3 6.0 5.3 2.7 60 60 5. 0 2.7 3. 0 2.3 70 65 4.0 3.0 3.0 2.0 100 100 13.2 8.7 8.7 7.0 200 200 29.4 30.1 32.3 30.2 Pan --- 28.8 42.5 46.0 54.5 Total 100.0 100.0 100.0 100.0 during the shattering procedure. The high heat control samples without the shattering procedure had a similar percent of fines as the low heat samples with the shattering procedure. Apparently some breakup of agglomerated particles occurred during the screening step of the instantizing operation. The shattering pro- cedure had only a minor effect on breakup of agglomerates of high heat samples because of the high percent of fines in the control. IV. Cost analysis of instantizing operation A. Plant A The cost of operation was projected on the basis of the follow- ing conditions: (a) (b) (C) (d) (e) 56 8-hour working day with 21. 67 days in a month and 260 days in a year. Time schedule of operation 8:00 - 9:00 a.m. setting up and preparation 9:00 - 12:00 a.m. production 12:00 - 1:00 p.m. lunch 1:00 - 3:30 p.m. production 3:30 - 5:00 p.m. cleaning up. In other words, 2. 5 hours were involved for auxiliary operations. Net production hours were 5. 5. Entire operation including setting up and cleaning up of the instantizer was done by one operator. Rated production capacity was 300 pounds per hour, con- sequently, l, 650 pounds per day. The depreciation of the instantizer and the building was calculated by straight-line method for 10 and 30 years, respectively. Interest, taxes and insurance, and mainten- ance of equipment were added as 2. 5. 2.0 and 3.0 percent of total investments, respectively. The maintenance of building was considered to be 1. 5 percent per year. (f) Allocation of indirect costs such as management, office, sales, laboratory, etc. , was estimated on the basis that instantizing required 5 percent of total dairy plant drying Operations . 57 (g) Requirements of utilities and supplies were calculated from the manufacturer's data and augmented by experience: steam-~6. 7 pounds for hot water for cleaning per day, 40 pounds for instantizer per hour, water-~100 gallons per day, e1ectricity-—43. 6390 KWH--efficiency was assumed to be 85 percent, all purpose detergent--0. 5 pound per day, and acid cleaner--0.2 gallon per day. (h) Prices of equipment, building, etc.: instantizer total price was $10, 705. 00, ten percent was added as installation costs. Conse- quently, the initial investment was $11, 775. 50, building cost was $15.00 per square foot. The space where the instantizer had been installed was 384.4 square feet, labor cost of $21.40 per day was used (10), utilities and supplies: the cost of a KWH was 1. 8 cents, water was $0.15 per 100 cubic feet or 750 gallons, steam at 100 p. s.i. cost $2.25 per 1, 000 pounds, all purpose detergent was $22. 95 per cwt. , acid cleaner was $1. 80 per gallon, 58 indirect costs such as management, office, laboratory, etc. . were taken from Plant A records. (i) Breakdown of individual cost per day was as follows: Fixed cost Instantizer and accessories Depreciation $ 4. 529038 Interest 1.132260 Tax and insurance 0. 905808 Maintenance 1. 358711 Subtotal $ 7. 925817 Building Depreciation $ 0. 739200 Inte rest 0. 554423 Tax and insurance 0. 443538 Maintenance 0. 332654 Subtotal 35 2.069815 Indirect costs Management $ 0. 989505 Office 0. 403375 Laboratory 0. 207660 Maintenance 0. 333000 Subtotal $ 1. 933540 Fixed cost total $11.929202 59 Fixed cost per pound of instant dry milk was 0.7230 cent. Variable cost Utilities and supplies Electricity $ 0. 785498 Water 0. 020000 Steam for machine 0. 585000 for warm water 0.151200 Detergent and cleaner 0. 474750 Subtotal $ 2. 016448 Labor $21. 400000 Variable cost total $23. 416448 Variable cost per pound of instantized dry milk was $0. 014192 and the fixed cost $0. 007230. The combined fixed and variable costs were $35. 35 or $0. 021422 per pound of instantized nonfat dry milk. (j) Package cost Bulk density of nonfat dry milk decreased approxi - mately 100 percent during instantizing. Therefore, the space occupied by instantized product was twice as much as noninstantized product. This caused the package costs to approximately double. If a regular 50-pound bag could accommodate only 25 pounds of instantized product, the package cost per pound was $0. 009576 based on the original cost of $0.23915 per bag. B. 60 Same conditions could be extended to retail packagings. A small operation used the cylinder shaped container of wax coated paperboard with push-in metal lid. It held 5 pounds of regular or 2 pounds of instant type nonfat dry milk. The cost of the container was $0. 12525. The pack- age cost was 2. 505 cents for regular and 6.2625 cents per pound for the instant nonfat dry milk. Plant B Information on instantizing costs for nonfat dry milk was ob- tained from the plant records that were maintained by a computer system for March 30 to April 25, 1964: (a) (b) (C) (d) The data indicated that instantizing time was 97. 25 hours per week or 19. 45 hours per day. Five working days were considered a week and 260 days constituted a year. The instantizing equipment was operated by two persons each shift. Production capacity was 650 pounds per hour and 12, 643 pounds per day which was an average of a 3-month operation. The depreciation of instantizing equipment and building was calculated by straight-line method for 15 and 30 years, respectively. Interest was computed at 2. 5 percent and tax and insurance at 2. 0 percent of total investment. The maintenance cost was calculated to be 3. 0 percent of (e) (f) (g) 61 equipment and l. 5 percent of initial investment of the building. Costs of utilities and supplies were obtained from the bills by allocating 0. 7458 KWH per HP. A total of 68 HP was required by the motors. The efficiency of a motor was assumed to be 85 percent. Water used for processing averaged 8 gallons per hour and 1, 000 gallons for cleaning which occurred once a week. Light, heat and fuel were estimated to be 70 percent of the monthly operation bills . The information on indirect cost included a portion of man- agement, office, sales and laboratory expenses. Prices of equipment, building, etc. The price of instantizing equipment was $49, 692 and installation cost of $12, 000 was in addition giving a total of $61, 692. Building spaces considered in this study were process- ing room with l, 850 square feet, warehouse with l, 500 and fan room with 300. Processing room was two story in height. The cost per square foot was $12. 00, $3. 00 and $10.00 respectively. Labor cost averaged $3. 00 per hour including fringe benefits. The cost of a KWH was $0. 012. Water was $0.15 per 62 100 cubic feet or 750 gallons. Supplies and miscel- laneous items were $23. 07 per day. (h) Breakdown of costs on daily basis. Fixed cost Instantizing equipment Depreciation $15. 818615 Interest 5. 931923 Tax and insurance 4. 745538 Maintenance 7. 118308 Subtotal $33. 614384 Building Depreciation $ 3. 807615 Interest 2. 855762 Tax and insurance 2. 284615 Maintenance 1. 713462 Subtotal $10. 661454 Indirect costs $50. 761422 Total fixed cost $95. 037260 Variable cost Utilities and supplies Electrhjty $13.925577 Water 0. 023707 Light 13.211814 63 Heat and fuel $ 17. 863443 Supplies 23. 073373 Subtotal $ 68. 097914 Labor $144. 000000 Variable cost total $212. 097914 Fixed and variable cost total $307.135074 Variable cost per pound of instant dry milk was $0. 016776 and the fixed cost was $0. 007517. The combined fixed and variable cost was $0. 024293 per pound. DISCUSSION The dispersibility of instant nonfat dry milk was not signifi- cantly different between the low heat and the high heat types. How- ever, the dispersibility test was not capable of indicating small differences. Undoubtedly, some factors other than the heat treatment of the skimmilk before drying within the range of good practice have an important influence on dispersibility of the result- ing instantized nonfat dry milk. Apparently the high heat instantized nonfat dry milk had fragile agglomerates. The sieve classification indicated a much larger amount of fines. However, these were not sufficiently small to affect the dispersibility test. The analysis of variance denoted that the moisture increased more when instantizing high heat than low heat nonfat dry milk. The difference can be explained by the fact that high heat nonfat dry milk has a greater moisture absorption and adsorption capacity and con- trolled redrying air had no additional capacity to remove it. The exact nature of water binding capacity is not known. Lots of nonfat dry milk were tempered to 400, 700 or 900 F. and then instantized. The analysis of variance indicated a significant increase in moisture retention of the instantized nonfat dry milk from the 400 F. trial compared to 700 F. trials and the 700 F. trials com- pared to the 900 F. trials. This trend might be anticipated on the basis of greater condensation on the particles from the humid air 64 65 currents in the agglomerating chamber. Since the difference aver- aged 0.34 percent between 400 F. and 700 F. trials and 0.16 percent between 700 and 900 F. trials the keeping quality of instant nonfat dry milk would be affected to a small extent. Therefore, to reduce moisture variation of the instantized nonfat dry milk the temperature of the powderperie should be controlled. Another possibility would be to adjust steam flow into the agglomerating chamber. The air movement in the chamber is balanced for a predetermined rate of product infeed so that increasing the amount of nonfat dry milk is not a recommended correction. Despite careful attempts to control the instantizing operations significant differences were observed among the moisture increases of various samples. From a practical standpoint the moisture of instant nonfat dry milk should be limited to not more than 4. 5 per- cent. This is the standard for Extra Grade, Instant Nonfat Dry Milk published in 1963 by the U.S. Department of Agriculture (61) . Correlation analysis did not reveal a relationship between moisture increase after redrying and dispersibility of instantized nonfat dry milk. The dispersibility of instantized nonfat dry milk was independ- ent within limits of the bulk density. Within the classification of particle sizes studied, sieves with 14 to 200 mesh plus pan, the dispersibility did not show a correlation until particles passed 66 through the 100» and were retained on the ZOO-mesh sieve. The product in the pan (not collected on the 200~mesh sieve) was only slightly lower in dispersibility. There may be several reasons the difference was not pronounced. The product in the pan may have contained a minimum of the exceedingly small particles. These are known to be detrimental to dispersibility. A representative sample contained 9 percent particles of 92 microns or smaller (fines) in diameter. Plants may add up to 15 percent fines to instant nonfat dry milk for commercial sale. The dispersibility method may not have been sufficiently refined for consistently detecting small dif- ferences. The term, dispersibility, has been applied to different phenomena in reconstitution of nonfat dry milk. For example, Bockian £31. (6) reported a general correlation existed between particle size and dispersibility. A scrutiny of their dispersibility method indicated that these authors were actually measuring wettability. An analysis of the effect of milk fat content of instantized dry - milks on dispersibility denotes a lack of significant difference among dry milks with milk fat contents of 1, 2, 5 and 10 percent although the average of all trials is slightly less for the dry milk with 10 percent than the other three dry milks . In fact, more than half of the results were below the minimum of 44 grams. Dry whole milk (26 percent milk fat) was significantly lower in 67 dispersibility (31. 9 grams) than the average of the samples with l, 2, 5 and 10 percent milk fat that were from 44. 4 to 48. 0 grams. The milk fat in large amounts such as dry whole milk undoubted- ly retards wetting of the particles during agglomeration. The use of more moist steam during agglomeration was an attempt to overcome the wetting problem. Theoretically milk fat on the surface of the particle would not be expected to have as much adherence during the collision of particles as wetted casein or lactose. The lyophobic property of milk fat also would decrease its dispersion tendency in aqueous solutions. The form of the fat also would have an effect. The free fat should have a larger adverse influence than fat within its natural membrane covering. A direct influence of increasing amounts of milk fat in the dry milks on bulk density was noticed. The specific gravity of milk fat is definitely lower than the remaining components of milk; however, this fact was offset by the reduced agglomeration of dry milk with the high milk fat contents. The net effect was a higher bulk density of the dry milk with higher percent milk fat. The difference be- tween 1 and 2 percent milk fat in the dry milk was not detectable by bulk density. The Coulter counter has definite limitations for classifying the particle size of agglomerated dry milks. Separation by the sieve method is necessary for the larger agglomerates; then the Coulter 68 counter can be used for the remainder which must be within the counting range of the aperture tube. Replications are desirable to assure reliability of results because of the very small quantity of sample used for the Coulter method. Another consideration is the complexity of the calculations of particle size from the data. The use of a computer minimized the required time for the calculations. Kinsman (29) reported that a 12-point curve (data from 12 observa- tions) required 2 minutes with a computer, 7 to 10 minutes with memory type calculator, 10 to 15 minutes with standard calculator and 20 to 30 minutes with slide rule. Cost analysis of two instantizing operations showed that these expenses are important. The projected cost of 2.14 cents per pound for Plant A and 2. 43 cents per pound for commercial Plant B repre- sent 15.3 and 17. 4 percent increases over 14 cents for regular non— fat dry milk. In addition, the packaging cost would double approxi- mately for the agglomerated dry milk because of the decrease in bulk density. The largest item of expense was labor. It constituted more than 50 percent of the instantizing expenses. Large capacity operations should be able to utilize labor more efficiently so the cost per pound should be lower. SUMMARY The dispersibility of high heat and low heat nonfat dry milk after instantizing was not significantly different. The difference in bulk density by the loose and packed methods was small, but sig- nificantly greater for the high heat nonfat dry milk. This may have been due to the larger amount of small particles. Instantized high heat nonfat dry milk had an 87 percent larger content of small par- ticles (those passing through 100- or ZOO-mesh sieves); conversely less larger particles were retained on 35- to 65-mesh sieves indi- cating a greater breakup of agglomerates during sifting or the sieve test than low heat nonfat dry milk. The high heat nonfat dry milk retained more moisture from redrying after agglomeration than the low heat product. The average difference was 0. 37 percent. Nonfat dry milk tempered to 400, 700 or 900 F. for instantizing showed no significant difference in dispersibility. The same applied to bulk densities. The average moisture increase after agglomerat- ing and redrying was 1. 72 percent, 1. 38 and 1.16 for 400, 700 and 900 F. nonfat dry milk, respectively. The dispersibility of instantized dry milks with l, 2, 5 or 10 percent milk fat varied (44. 4 to 48. 0 grams), but all averaged above the minimum standard of 44. 0 grams. Dispersibility of dry milk with 26 percent milk fat averaged 31 . 90 grams. After agglomera- tion and redrying the moisture increase of nonfat dry milk ( 1 percent 69 70 milk fat) was 0. 97 percent. Dry milks with 2, 5, 10 and 26 percent milk fat varied from 1. 84 to 2.12 percent increase. The increase was not consistent with the milk fat content increase. The packed bulk density decreased 0.268, 0.237, 0.223, 0.192 and 0.136 gram per ml. with the milk fat content of 1, 2, 5, 10 and 26 per- cent in the instantized dry milk. A large variation was observed in the particle sizes of six brands of instantized nonfat dry milk on the local retail market. The sieve method of separating particles with screen mesh of 14, 20, 28, 35, 48, 60, 65 and 100 did not give samples with a signifi- cant difference in dispersibility. Particles retained on the 200- mesh sieve and in the pan had a slightly lower dispersibility. The loose bulk density increased from 0.187 to 0. 500 and packed bulk density from 0.223 to 0.613 gram per ml. with the above range of particle sizes. A combination of the sieve method and Coulter counter was required to classify particle sizes of agglomerated nonfat dry milk. Plant A with an instantizer having a capacity of 300 pounds of instantized product per hour had a fixed cost of 0. 7230 and variable cost of 1.4192 for a total instantizing cost of 2.1422 cents per pound. Plant B's capacity was 650 pounds per hour. The fixed cost was 0.7517 cent and the variable cost was 1. 6776 cents. The total instantizing cost was 2. 4293 cents per pound. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (ll) (12) LITERAT URE CITED Alfa Laval Co. (1961) . Metliod for the manufacture of coarse powder and of soluble agglomerates from fine powder and suitable equipment. French Pat. 1, 255, 231. (Dairy Sci. Abstr. 24(9):2501) American Dry Milk Institute, .Inc. (1962) . Standards fir Grades for £132 Dry Milk Industry Including Methods of Analysis. Bul. 916. Chicago, 1. American Dry Milk Institute, Inc. (1954) , (1963) . Census _o_f Dry Milk Distribution and Production Trends. Chicago, 1. Amundson, C. H. (1960) . A1 study of the effect of certain processing variables on Some spray dried milk products. Ph.D. Thesis, University of Wisconsin. Bissel, R. H. (1960) . Process for increasing the solubility of powdered milk. U.S. Pat. 2, 949, 363. Bockian, A. H., G. F. Stewart and A. L. Tappel (1957) . Factors affecting the dispersibility of "instantly dis- solving" dry milks. Food Res. 22:69-75. Budding, H. (1961) . A process for the preparation of a rapidly moistening and rapidly sinking milk powder. British Pat. 859, 133. Buffa, E. S. (1962). Modern Production Management. John Wiley and Sons, Inc., New York. 125-134. Carlson, E., R. A. Weir and J. V. Ziemba (1956). New way to instantize milk solids. Food Eng. 28(10):62—63, 194. Clayton, J. E. (1964). E. B. McLain Co. Personal cor- respondence, Mar. 9. Coulter Electronics Industrial Division (1958) . Instruction Manual Coulter Counter Industrial Model A. Chicago, 14. Coulter, S. T. and V. H. Townley (1959). Method and apparatus for spray drying. U.S. Pat. 2, 887, 390. 71 (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) 72 Dixon, W. J. and F. J. Massey Jr. (1957) . Introduction t_q Statistical Analysis. McGraw-Hill Book Co. , Inc., New Ybrk. 119-124. Fixari, F., W. Conley and G. Bard (1959) . Continuous high vacuum drying techniques. Food Tech. 13:217-220. Griffin, H. L. (1959) . Agglomeration process and apparatus. U.S. Pat. 2, 893, 871. Guenther, W. C. (1964). Analysis 9_fVariance. Prentice- Hall Inc., Englewood Cliffs, New Jersey. 54-57. Hall, C. W. and T. I. Hedrick (1961) . The manufacture of instant milk powder. Dairy Eng. 78( 1) :7-11. Hanrahan, F. P. and B. H. Webb (1961) . USDA develops foam spray drying. Food Eng. 33(8):37-38. Hanrahan, F. P., A. Tamsma, K. K. Fox and M. J. Pallansch ( 1962). Production and properties of spray-dried whole milk foam. Jour. Dairy Sci. 45:27-31. Harland, H. A. (1962) . Maple Island Inc., Stillwater, Minnesota. Personal correspondence. Harper, M. K., K. K. Fox, V. H. Holsinger and M. J. Pallansch (1960) . The reconstitution rate of milk pow- ders after wetting. Abstr. Pap. Amer. Chem. Soc., Sept. Havighorst, C. R. (1954) . New process produces instant milk powder. Food Eng. 26(10):?4-77. Hodson, A. Z. (1956) . Nutritive value of instant nonfat dry milk. Food Tech. 10(5) :221-224. Hodson, A. Z. and C. B. Miller (1957). Nutritive value of the proteins of stored instant nonfat dry milk. Food Tech. 11(2):89-92. Howard, N., K. K. Fox, V. H. Holsinger and M. J. Pallansch (1961) . The specific surface areas of milk powders pro- duced by various drying methods. Abstr. Pap. 56th Ann. Meeting ADSA. Jour. Dairy Sci. 44( 6):1155-1156. (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) 73 Keeney, M. and R. Bassette (1957a) . Flavor and chemical properties of nonfat dry milk. Abstr. Pap. 52nd Ann. Meeting ADSA. Jour. Dairy Sci. 40(6) :605 pp. Keeney, M. and R. Bassette (1957b). Rapid chemical changes in reconstituted dry milk. Sci. 126:511—512. Kennedy, J. G. and E. R. Spence (1960). Easily reconsti- tuted milk powder. U.S. Pat. 2, 928, 742. Kinsman, S. (1962) . Measurement of particle size and distribution. Proc. of 10th Natl. Dairy Eng. Conf. 81--88. Kosikowski, F. V. (1958) . Cottage cheese from low heat, "Low-Temp” and instant spray-dried milks. Milchwissen- schaft. '13(8):351 pp. (Dairy Sci.Abstr. 20:918) Kumetat, K., R. Beeby and A. K. Morris (1956) . A note on the reducing groups of "instant" milk powders. Aust. Jour. Dairy Tech. 11(3):113-114. King, N. (1954) . The physical structure of milk powder. Part 11. Dairy Industries 19(2):131—-134. Lauder, E. A. and A. Z. Hodson (1958) ., Instantly soluble milk powder and process for making same. U.S. Pat. 2, 832, 686. Loftus, H. G. (1955). "Instant" milk powders. Aust. Jour. Dairy Tech. 10(1):8-9. Mann, J. B. and J. M. McIntire (1960). Comparison of flavor and some other chemical properties of noninstantized and instantized nonfat dry milk. Abstr. Pap. 55th Ann. Meeting ADSA. Jour. Dairy Sci. 43(6):844 pp. Mann, J. B. and J. M. McIntire (1961). Comparison of flavor and some other chemical properties of instantized and noninstantized nonfat dry milk. Jour. Dairy Sci. 44(5) :851-855. McDivitt, M., A. M. Swanson and L. V. Witter (1960). The effect of forewarming treatments on the flavor of nonfat dry milk. Abstr. Pap. 55th Ann. Meeting ADSA. Jour. Dairy Sci. 43( 6):844 pp. (38) (39) (40) (41) (42) (43) (44) (45) (46) (47) (48) (49) (50) (51) 74 McIntire, J. M. and C. C. L00 (1960). dairy product and method of manufacture. 2, 941, 886. Fat-containing dried U.S. Pat. Milk Industry Foundation ( 1959) . Laboratory Manual, Meth- ods of Analysis of Milk and Its Products, 3rd ed. W‘ash- ingt5r1'6, D.C., 296-297. Morgan, A. I. Jr., L. F. Gumette, J. M. Randall and R. P. Graham (1959) . Technique for improving instants. Food Eng. 31(9):86-87. Myrick, N. (1956) . 150 million pounds of instant milk. Amer. Milk Rev. 18(7):24, 26-28 and 104. Niro Atomizers, Ltd. ( 1960) . production of instant milk powder. Copenhagen. Niro Atomizer equipment for Company catalogue, Peebles, D. D. (1956) . The development of instant milk. Food Tech. 10(2):64-65. Peebles, D. D. (1958) . Dried milk product and method of making same. U.S. Pat. 2, 835, 586. Peebles, D. D. (1959) . Milk manufacturing method and product. U.S. Pat. 2, 911,300. Rice, M. A. (1961) . Method of preparing instant dry milk. U.S. Pat. 2, 994, 612.. Rich, S. R. (1963). Personal correspondence. TEKNIKA Inc. , Hartford 5, Conn. Aug. 5. Sanna, F. L. (1960) . Method for drying food materials. U.S. Pat. 2, 953, 457. Schulz, M. E. (1963) . Zerstaubungstrocknen und Instantisi- eren. Hans Carl, Niirnberg. 14-18. Scott, E. C. (1959) . Method and apparatus for producing granulated food. U.S. Pat. 2, 900, 256. Sharp, P. F. and C. A. Kempf (1960) . Method for the pre- paration of powdered milk products. U.S. Pat. 2,921,857. (52) (53) (54) (55) (56) (57) (58) (59) (60) (61) (62) (63) (64) 75 Shields, J. B. (1962) . Method of manufacturing an instantized product. U.S. Pat. 3,057,727. Sinnamon, H. I., N. C. Aceto, R. K. Eskew and E. F. Schoppert (1957) . Dry whole milk. I. A new physical form. Jour. Dairy Sci. 40:1036~1045. Sjollema, A. (1960) . Process for modifying powdered milk products. . U.S. Pat. 2, 953, 458. Sone, T. and M. Makino (1961) . Snow Brand Milk Products Co., Ltd. , Tokyo, Japan. Personal correspondence. Spiess, N. E. and N. E. Sullivan (1962) . Powder agglom~ crating method and apparatus. U.S. Pat. 3, 042, 526. Starey, A. W. (1956) . Note on the condition of lactose in milk powder. Aust. Jour. Dairy Tech. 11(1):11 pp. Takano, T. (1961) . Instant milk. "Instant food" symposium by Nosan Kako Gijutsu Kenky-ukai (in Japanese). Tamsma, A., L. F. Edmondson and H. E. Vettel (1959) . Free fat in foam-dried milk. Jour. Dairy Sci. 42:242 pp. Tamsma, A., M. J. Pallansch and T. J. Mucha (1959). A study of factors related to the storage stability of foam- dried whole milk. Abstr. Pap. 54th Ann. Meeting ADSA. Jour. Dairy Sci. 42:910 pp. U.S. Department of Agriculture (1963) . United States stand— ards for instant nonfat dry milk. Federal Register, March 26. W. S. Tyler Co. (1964) . Standard Testing Sieves and Labyo- ratory Sieve Equipment. Latest Revised Specification. Bul. 608. Winder, W. C. and E. W. Kielsmeier (1959). Process of dry milk. U.S. Pat. 2,911,301. Yamamoto, Y. (1963) . Snow Brand Milk Products Co. , Ltd., Tokyo, Japan. Personal correspondence. n (3511411 11 $1.- 0 Elm 9 if.“ . I ”1111111111(1111131111111'Es