FIELD ^* OF THE INVENTION

This invention relates to processes for making food products by selective concentration of milk, food products obtained by such processes, and foods made from such food products.

BACKGROUND OF THE INVENTION

A. Definitions of Terms

"Milk" means the lacteal secretion obtained by the milking of one or more females of a mammalian species, such as cow, sheep, goat, water buffalo, or camel.* Broadly speaking, such milk is comprised of casein (a phospho-protein) and soluble proteins, lactose, minerals, butterfat ^" (milkfat), and water. The amount of these constituents in the milk may be adjusted by the addition of, or the removal of all or a portion of, any of these constituents. The term "milk" includes lacteal secretion whose content has been adjusted.

Milk obtained by milking one or more cows is referred to as "cows milk". Cows milk, whose composition has not been adjusted is referred to herein as "whole milk". It is comprised of casein, whey proteins, lactose, minerals, butterfat (milkfat), and water.

OMPI

The composition of "cows milk" can be adjusted by the removal of a portion of or all of any of the constituents of whole milk, or by adding thereto additional amounts of such constituents. The term "skim milk" is applied to cows milk from which suf¬ ficient milkfat has been removed to reduce its milkfat content to less than 0.5 percent by weight. The term "lowf t milk" (or "part-skim milk") is applied to cows milk from which sufficient milkfat has been removed to reduce its milkfat content to the range from about 0.5 about 2.0 percent ^* by weight.

The additional constituents are generally added to cows milk in the form of cream, concentrated milk, dry whole milk, skim milk, or nonfat dry milk. "Cream" means the liquid, separated from cows milk, having a high butterfat content, generally from about 18 to 36 percent by weight. "Concentrated milk" is the liquid obtained by partial removal of water from whole milk. Generally, the milkfat (butterfat) content of concentrated milk is not less than 7.5 weight percent and the milk solids content is not less than 25.5 weight percent. "Dry whole milk" is whole milk having a re¬ duced amount of water. It generally contains not more than five percent by weight of moisture on a milk solids not fat basis. "Nonfat dry milk" is the product ob¬ tained by the removal of water only from skim milk. Generally, its water content is not more than five weight percent and its milkfat content is not more than 1.5 weight percent. Thus, the term "cows milk" includes, among others, whole milk, low fat milk, (part-skim milk), skim milk, reconstituted milk, recombined milk, and whole milk whose content has been adjusted.

The term "whey proteins" means cows milk pro- teins that do not precipitate in conventional cheese

making processes. The primary ^' whey proteins are lac- talbumins and lactoglobulins. Other whey proteins that are present in significantly smaller concentrations include euglobulin, pseudoglobulin, and- i munoglobulins. 5 In the conventional manufacture of cheese, milk, is processed to form a coagulum, which is further processed to produce a semi-solid mass called "cheese curd" (or "curd") and a liquid (whey). The curd contains casein, a small amount of lactose, most of the butterfat, 10 minerals, and water. The whey contains whey proteins, most of the lactose, some of the butterfat, minerals, and water. The curd may be worked (e.g., stirred and/or combined with certain flavor and taste producing ingredi¬ ents, and/or ripened using bacteria to produce different 15 varieties of "natural cheese."

One or more varieties of curd or natural cheese can be comminuted and mixed with an emulsifying agent to form, with the aid of heat, as homogeneous plastic mass called "processed cheese". (Examples of processed 20 cheese include: "process cheese"), "process cheese food", "process cheese spread", and "process cheese product" . The various types of processed cheeses are obtained depending on the processing conditions, the specific varieties of curds or natural cheeses used, 25 and the additional ingredients added during the proces¬ sing. "• , "Imitation cheese" is a food made in semblance of any natural cheese variety, processed cheese, or other foods made of natural or processed cheese, in 30 which casein, caseinates, and/or safe and suitable non- milk ingredients, such as vegetable proteins and vegetable oil, replace all or part of the nutritive milk components normally found in the food being simulated.

Herein "buffer capacity" is the resistance of ^• 35 a system to pH change and is expressed in milliequivalents

of hydrochloric acid absorbed per 100 grams of sample. Buffer capacity is measured by diluting 2 g of the milk, or the equivalent amount of process fluid with 100ml of distilled water and titrating with .05 N hydrochloric acid to pH 5.1 until that pH remains constant for at least, twenty seconds. "Lactose" is expressed as grams of anhydrous lactose per 100 grams of sample.

B. The Problem This Invention Addresses Natural cheese of the types used as bases for making processed cheese products, for example, Cheddar, stirred curd, Colby, Gouda, or Swiss, are made by coagu¬ lating milk, ripened with suitable lactic acid producing bacteria to a suitable acidity, with appropriate milk clotting enzymes, cutting the coagulum and cooking the resulting curd in its whey. The whey is drained from the curd, whereupon the curd may be cheddared or stirred while additional acid is produced by fermentation of lactose to lactic acid in the curd. The curd may or may not be washed with water. If cheddared, the curd is milled, whereupon it is salted and pressed into rec¬ tangular blocks or packed into barrels for maturing. The addition of lactic acid producing bacteria, ripening of the milk, cutting the curd, cooking the curd, stirring the curd, cheddaring the curd and salting the curd are all controlled to yield a product, in which the residual lactose and the unused buffer capacity of the curd are balanced so that complete fermentation of the residual lactose in the curd to lactic acid will result in a cheese with the proper minimum pH for the particular market requirements, usually between 4.9 and 5.5.

The conventional processes for making natural and processed cheeses utilize only casein. Whey proteins remain dissolve in the whey and are discharged as a by-product of the process. The whey proteins comprise

OMPI // WIPO _ _,

about 14 to 24 weight percent of whole or skim milk's proteins and the nutritional value of the whey proteins is at least comparable to the nutritional value of casein. Accordingly, the loss of the whey proteins in the con- ventional cheese making processes limits the potential yields of such processes. The utilization of even a portion of the whey proteins in the manufacture of natural and processed cheeses is of great commercial importance.

C. Prior Attempts to Recover Whey Proteins For above-stated reasons, efforts have been made to design a process that would allow full utilization of the whey proteins in cheese making.

One approach is to recover whey proteins from the whey by drying, condensing, ultrafiltration, or reverse osmosis of the whey. The recovered whey proteins are then recombined with the cheese. Such processes are described for example in F.V. Kosikowski, Cheese and Fermented Foods, Edward Brothers, Inc., Ann Arbor, Michigan, U.S.A., 2 ed. 1977, pp. 451-458. A problem with this approach is that some countries have laws which prohibit, for most types of natural cheeses, recombining separately recovered whey proteins with the curd. For example, in the United States whey proteins can be reincorporated into cheese curd only in the making of certain natural cheeses, such as skim milk cheese.

An additional problem associated with this approach is that the recovered whey constituents lack the physical and chemical characteristics required for the making of natural cheese. For example, the dried whey proteins can be sprinkled into the cheese curd. However, only a limited amount of whey proteins can be added to cheese curd in this manner without changing its desired properties. Accordingly, the whey proteins

OMPI /fa WIPO . ,

recovered from whey are not used to any significant extent in commercial processes for making natural cheese. For these same reasons whey proteins recovered in this manner are not used to any significant extent 5 in the commercial manufacture of process cheeses. More¬ over, it is commercially feasible to add only limited amounts of such recovered whey proteins to process cheese spreads, process cheese foods, process cheese products, or imitation cheeses.

10 Another approach for utilization of whey pro¬ teins in cheese making is to co-precipitate them with the casein. One process for obtaining co-precipitates of casein and whey proteins from milk is disclosed in - U.S. Patent No. 3,535,304 and in the corresponding

15 Australian Patent No. 403,065 (hereinafter referred to as the Muller patents) . The Muller patents specify that the product resulting from this process should find acceptance in some forms of baby food, ice cream, coffee whiteners, small goods, biscuits, bread, break-

20 fast cereals, and canned processed foods. The final product of the Muller process lacks the "functionality" for making processed and imitation cheeses, i.e., it does not have the functional characteristics that are required for making such cheese.

25 Other known processes for the production of co-precipitates of casein and whey proteins also result

^■* .- in products that, although useful for some applications, cannot be used as starting materials for making cheese. A further approach for increasing the yields

30 of cheese making processes by utilization of the whey proteins is by ultrafiltration of milk. It has pre¬ viously been proposed to produce products suitable for conversion into cheese by altering the composition of whole or skim milk utilizing ultrafiltration or reverse

OMPI

I Λ

osmosis. The milk contacted with a membrane which per¬ mits the passage of water, lactose and some minerals, but prevents the passage of casein, the whey protein, butterfat, and some minerals. The selective concentra- tion of milk results in the formation of a retentate which contains whey proteins. When the retentate is coagulated by acid or rennet, it forms a curd which contains the whey proteins. If this coagulum is then subject to syneresis some of the whey proteins are lost in the expressed whey. One method for producing cheese by ultrafiltration of milk is disclosed in U.S. Patent No. 4,205,090 (Maubois et al, ) and in Australian Patent Specification No. 477,399. According to this method, whole or skim milk is concentrated by ultrafiltration to about one fifth of its volume to give a product, sometimes called a "liquid pre-cheese", which is then manufactured by conventional means to give cheese. This method, however, is only suitable for making soft cheeses of the Camembert or Reblochon type and possibly some semi-hard cheeses. It cannot be used to produce harder cheese of the Cheddar, Colby, or stirred-curd type, because the water content of the pre-cheese is too high and the ratio of buffer capacity to lactose will not give the desired final pH in the product. More recently, C-.A. Ernstrom, B.J. Sutherland and G.W. Jameson described in an article entitled "Cheese Base for Processing: A High Yield Product from Whole Milk by Ultrafiltration" and published in Journal of Dairy Science, Vol. 63,228-234, (1980), a process whereby the moisture content and pH of the final product can be controlled to any desired levels. This process provides a product, cheese base, which has a moisture content and pH appropriate for use in place of conventionally made natural cheeses in the production of process cheeses. In the Ernstrom et. al process, whole milk of normal pH

or acidified to pH 5.7 is concentrated by ultrafiltra¬ tion to 40% original milk weight and diafiltered at constant volume until a desired ratio of lactose to buffer capacity is established. Then the retentate is further concentrated by ultrafiltration to 20% of the original milk weight. The retentat.es are then inocu¬ lated with cheese starter and incubated to completely - ^* ferment the residual lactose. Precise control of final pH is achieved by controlling the level of lactose through diafiltration.

The fermented retentate is converted in a batch manner to cheese base in a swept-surface vacuum pan evaporator. The cheese base can be used to replace the unripened natural cheese component of processed cheese as it has the same pH and gross composition as Cheddar cheese. The use of the batch evaporator is necessitated b the fact that the retentate upon fermen¬ tation forms a curd or coagulum. Such a product cannot be readily processed in any continuous flow evaporator. Accordingly, the overall process of Ernstrom et. al. is essentially a batch process.

Thus the prior attempts have been unsuccessful in devising a process for economical utilization of whey proteins in the manufacture of products with a composition of cheese. There is therefore an -unsatisfied, long-felt need for a high-yield, efficient process for making such products. The present invention is concerned with improvements to the Ernstrom et. al. process just described, particularly in relation to the adaptation of the process to commercial use.

SUMMARY OF THE INVENTION

This invention provides a process for an ef¬ ficient production of cheese or cheese base containing substantially all the casein and whey proteins which

OMPI

were originally present in the ^' milk. The process permits an efficient evaporation of water from the fermented concentrate by maintaining the fermented concentrate in the liquid state prior to the evaporation step. The 5 process avoids the possibility of losing whey proteins by totally eliminating syneresis.

The process of the present invention comprises four phases: (1) selective concentration of milk to form a concentrate; (2) increasing the ionic strength 10 of the concentrate so as to maintain it in the liquid phase during fermenting, (3) fermenting the concentrate; and (4) removing water from the fermented concentrate to produce cheese or cheese base containing substantially all the casein and whey proteins originally present in 15 the milk.

The process of the present invention produces unique intermediate products: a unique concentrate and fermented concentrate which do not coagulate even at low pH values (i.e., 4.9-5.5). 20 It has been found that preheating the fermented concentrate prior to water removal in phase four provides unexpected results, namely: (1) preventing phase separation (i.e. oiling off) during water removal at high percent solids; (2) reducing burn-on during water removal; and 25 (3) reducing the viscosity of the resultant product.

- , BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 is a schematic diagram of a semi- continuous process apparatus of this invention.

FIGURE 2 is a.schematic diagram of a batch 30 process apparatus of this invention.

DETAILED DESCRIPTION OF THE PRO ^~ CESS

A. General Description of the Process The process of the present invention permits an efficient production of cheese and cheese base from 5 milk. The efficiency of the process stems from increasing the ionic strength of the concentrate prior to or during fermentation but before the formation of any coagulum. The increase in the ionic strength maintains the concen¬ trate in the liquid state during and after the fermenta- 10 tion. This allows an easy removal of water from the fermented concentrate so that the desired moisture level can be achieved in a commercially feasible manner. The increase of the ionic strength prevents the formation of a coagulum and subsequent syneresis and thereby 15 assures that the whey proteins remain in the concentrate and are included in cheese or cheese base. The process of the present invention can be subdivided into four phases: (1) selectively concentrating milk to form a concentrate; (2) increasing the ionic 20 strength of the concentrate so as to maintain it in the liquid phase during and after fermentation; (3) fermenting the concentrate; and (4) removing water from the fermented liquid concentrate until a desired moisture level is reached.

_*

25 B. Raw Materials for the Process

^■ ' , The process of the present invention can be carried out on any milk. The milk used in the process of the present invention may be pasteurized or homogenized before the completion of the selective concentration 30 phase.

During the process of the present invention neither protein nor butterfat is normally removed from the concentrate. Accordingly, if a specific ratio of protein to butterfat is desired in the final product

such ratio is preferably established before the concen¬ tration phase. For example, a product which will sub¬ stitute for a natural Cheddar cheese (having 50.5 to 51.5% fat in dry matter) will require a protein/fat ratio of 0.63 to 0.64 when measured as casein/fat ratio by the Walker Casein Test. The adjustment of the pro¬ tein to fat ratio can be achieved by either adding or removing one of these components.

In situations where it is preferred to ultra- filter skim or part-skim milk, protein to fat ratio of the final product may be adjusted by an appropriate incorporation of milk fat in the form of cream, anhydrous milk fat, or butter, at any stage after ultrafiltration.

C. The Selective Concentration of Milk to Form A Concentrate

The purpose of this phase is to achieve the desired levels of moisture and lactose in the concen¬ trate. The precise moisture level depends on the de¬ sired moisture in the final product and the cost of removal of the moisture by ultra iltration as opposed to evaporation. While the process can be operated in a concentration ratio of from 1:1 to 9:1, a concentration ratio of from 3:1 to 6:1 compared to the original milk . is preferred. The reason for adjusting the level of lactose is that the ratio of buffer capacity to lactose deter¬ mines the pH of the final product. In other words, the pH desired in the final product can be achieved by con¬ trolling the ratio of buffer capacity to lactose concen- tration during the selective concentration phase.

The ratio of buffer capacity to lactose is, in turn, controlled by the manner of preparation of the process fluid, i.e., in the preferred process by a com¬ bination of the extent of diafiltration and the degree ^■

• fE cr OMPI

of ultrafiltration (concentration) of the retentate following diafiltration.

The desired buffer capacity to lactose ratio can be achieved in any desired manner. It is presently 5 preferred to use for that purpose a combined ultrafiltra¬ tion and diafiltration treatment. The preferred manner of carrying out this treatment comprises:

(1) concentrating whole milk by ultrafiltra¬ tion until about 40% to 60% by weight of the milk

10 has been removed as the permeate fraction;

(2) subjecting the retentate fraction to diafiltration, preferably continued until the amount of water added is from 15 to 91% of the weight of the original milk; and

15 (3) resuming and continuing ultrafiltration until the retentate fraction amounts to about 20% to 30% of the weight of the original milk. The extent of diafiltration and ultrafiltration steps are adjusted so as to achieve the desired buffer capacity 20 to lactose ratio. For example, a ratio of buffer capacity to lactose of 5.7 at the end of the diafiltration step and 16.5 at the end of the second ultrafiltration step will yield a final product with pH 5.2.

D. The Increase of the Ionic Strength of 25 the Concentrate

- .. In this phase the ionic strength of the concen¬ trate is adjusted to substantially eliminate formation of an acid coagulum during fermentation. Such adjustment results in a fermented product which is in the liquid

30 state and can be processed by continuous flow evaporators, for example, a swept surface evaporator.

It is convenient to make the necessary adjustment to the ionic strength of the process fluid by the addition of salt, e.g., sodium chloride, and more preferably by

-g fUEAlT

OMPI

^■ **

the addition of an amount of sodium chloride sufficient to give the required sodium chloride concentration in the final product.

The addition of sodium chloride, however, is 5 not the only way in which the ionic strength of the process fluid can be adjusted and any other suitable technique may be used so long as it achieves the desired ^~ •object, that is, the avoidance of the formation of an acid coagulum during fermentation. The ionic strength 10 can be adjusted by the addition of any other non-toxic salt that substantially ionizes in water and does not significantly affect the pH of the concentrate. Examples of such salts include potassium chloride, buffered phos¬ phates, citrates, and mixtures of such salts. 15 The amount of the salt added must be sufficient to prevent coagulation during fermentation but insuf¬ ficient to either significantly slow down the activity of the cultures or to exceed the level of salt allowed in the final product. The amount of salt added is gener- 20 ally less than about four percent by weight of con¬ centrate. If sodium chloride is the salt, preferably about 0 ^" .5 to 4 percent by weight of the concentrate is added. For sodium chloride the most preferred range is from about 0.5 to 1.5 percent by weight of the concen- 25 trate.

The increase of the ionic strength of the - , concentrate must be effected before the formation of any coagulum.

E. The Fermentation of the Concentrate 30 The concentrate is inoculated with lactic acid producing agents, such as bacteria, and incubated until the lactose content of the .process fluid is consumed or until the pH reaches the desired level.

The presently preferred agents are Streptococcus lactis. Streptococcus cremoris, or a combination of these, but any desired starter cultures can be employed in this phase. The fermentation is preferably conducted at the optimum incubation temperature for the starter culture used. Generally the optimum incubation temperature is in the range from about 25 to 35 C.

F. The Removal of Moisture The moisture can be removed from the fermented concentrate in any convenient manner. It is presently preferred to remove the moisture by evaporation until the moisture content of the product reaches the desired level. Especially preferred are continuous swept surface evaporators, such as Models HS 0050, HS 200, and. HS

1200, sold by Luwa Corporation, Charlotte, N.C., U.S.A.

The fermented concentrate is preheated to a selected temperature in the range of 40 C to 100 C prior to evaporation. The particular preheat temperature selected will depend principally upon the present total solids sought in the resultant product after evaporation. The preheat temperature should be sufficient to enable evaporation to the total percent solids desired without phase separation in the form of the ^' appearance of free fat (oiling off) but insufficient to cause burning, discolorization, or caramelization. The preheat tempera¬ ture effective for prevention of oiling off will vary with the feed rate of the concentrate and the preheat time duration and may also vary with seasonal character- iεtics of the milk being processed.

G. The Product of the Process of this Invention The present invention produces unique intermediate products: a concentrate and a fermentate, having increased

ionic strength by virtue of a ήon-toxic salt that sub¬ stantially ionizes in water. The amount of such salt should be sufficient to prevent any coagulation throughout processing but insufficient to either significantly slow down the activity of the cultures or to exceed the level of salt allowed in the final product.

The compositions of the intermediate concentrate and fermentate vary depending on the starting materials, the amount of lactose removed and the concentration achieved. The following are the compositions for a retentate, salted retentate and fermentate obtained from whole milk. The retentate was obtained by a five¬ fold concentration of whole milk by a combination of ultra- and diafiltration.

Composition Salted Fermented By Weight Retentate Retentate Salted Retentate

Total Solids 38-43 38-44 38-44

Fat in Dry

Matter 50-55 50-55 50-55

Fat 18-23 18-23 17-22

Protein 13-19. 13-19 13-19

Lactose 1.6-1.8 1.6-1.8 Trace

Ash 1.5-2 1.5-2 1.5-2

Sodium Chloride None •5-4% •5-4%

PH 6.6-6.8 6.6-6.8 4.9-5.4

The final products made by the process of ^• this invention vary in the solid contents, pH, and butter¬ fat content depending on the starting materials and the processing conditions.

It is important to ncrte, however, that the process of the present invention can be used to make on a commercial scale a final product which (1) includes substantially all of the whey proteins originally present in the milk and (2) has the composition of hard cheese, such as Cheddar type.

In the preferred embodiment of the invention, i.e., when making cheese base with the composition of Cheddar type cheeses, the pH range in the final product is from 4.9 to 5.5 preferably from 5.1 to 5.2, and the moisture content is in the range from about 33 to 46 weight percent, preferably about 36 weight percent.

The conventional Cheddar cheese process results in the recovery of about 93% of the fat, about 95% of the casein and 7% of the whey proteins from milk. The process of the present invention recovers essentially all the fat, all the casein and about 90% of the whey proteins. This results in a yield increase of about 13-18% over the conventional cheese process. The exact percentage increase of yield will depend on the moisture level in the final product and the composition of the original milk.

Embodiments of the invention are illustrated by the following specific examples. The examples refer particularly to production of cheese-like products of the Cheddar, Colby, or stirred-curd type, and the extents and relative amounts of ultrafiltration and diafiltration given are based on the particular equipment used. The use of other kinds of ultrafiltration equipment, the use of different membranes or the requirements for other cheese types, may alter these specifications for achieving the proper ratio of lactose to buffer capacity.

DETAILED DESCRIPTION OF AN EMBODIMENT OF THE PRESENT INVENTION

The process of the present invention will now be described in connection with a semi-continuous process for making cheese base of the composition of Cheddar cheese, schematically shown in FIG. 1.

A. Standardization and Heating of Milk Referring now to FIG. 1, the whole milk is introduced from storage (not shown) to milk tanks 10 and 20 via lines 12 and 17, respectively. The whole milk in the tanks 10 and 20 is not pasteurized or homo¬ genized. The milk is lightly agitated using stirrers 18 and 19.

Cream from a vessel 21 is added to the milk in the tanks 10 and 20 via lines 22, 26 and 28 to stan¬ dardize the milk. The amount of cream is manually regu¬ lated using a valve 29 to bring the weight ratio of casein to butterfat to about 0.63. The amount of casein is determined by the conventional Walker test described in Jour. Ind. Eng. Chem. Vol. 6, No. 2, 1914 and the article by T. E. Gilmore and W. V. Price published in The Butter, Cheese, and Milk Products Journal Vol. 44, No. 3, 1953. The butterfat is determined by the Babcock method described in Standard Methods for the Examination of Dairy Products (Interdisciplinary Books

& Periodicals for the Professional and the Layman, 14th ed. 1978) p. 236-239.

It should be noted that the casein to butter¬ fat ratio changes depending on the desired final product and that instead of adding cream the milk could be stan¬ dardized by removing some skim milk from the whole milk.

The standardized milk is pumped by a centrifugal pump 34 via a line 35 past either a valve 37 or a valve 39 into a conventional plate heat exchanger 45. The heat exchanger 45 heats the milk to a temperature high

enough to obtain efficient performance of the ultrafil¬ tration apparatus but low enough to avoid denaturing the whey proteins. Generally the temperature of milk is in the range from about 49 to 60 C, preferably to 50 ^W C.

B. Ultrafiltration of the Milk

The heated milk is pumped via a line 47 to a surge tank 48. From the surge tank 48, the milk is pumped by a pump 50 via a line 49, a flow meter 52 and a line 53 into a first ultrafiltration module 57. The ultrafiltration module 57 can be of any suitable type. Presently preferred is a two-spiral-membrane module sold by Bell Byant PTY Ltd. , of Kensington, Victoria, Australia. This module uses a membrane of ABCOR, Inc. , Wilmington, Massachusetts, U.S.A.

In the module 57 the milk is separated into a permeate which leaves the module 57 via a line 58 and a retentate (concentrate). A part of the retentate is recycled via a line 59 and a heat exchanger 60 using a centrifugal pump 61 and a part of the retentate is passed via a line 62 to a second ultrafiltration module 63. The heat exchanger 60 is provided to cool to the pre¬ selected ultrafiltration temperature the milk heated as the result of treatment in the ultrafiltration module 57. The ratio of the recycle through line 59 to the through-flow through line 62 varies depending on the operating conditions but is generally in range from about 10:1 to about 20:1.

The retentate entering the module 63 is separated into a second stage permeate and a second stage retentate. The second stage permeate is removed via a line 64. A part of the second stage retentate is recycled via a line 65 and a heat exchanger 66 by a centrifugal pump 67. The heat exchanger 66 cools to the preselected ultrafiltration temperature. The

second stage retentate is heated as the result of treat¬ ment in the module 63. The recycle to through-flow ratio is in the same range as that of the first stage ultrafiltration. 5 The second stage retentate is passed via a line 68 to a first diafiltration module 70. The module 70 is substantially -identical in construction and opera- , tion to the ultrafiltration modules 57 and 63. The only difference is that water is supplied to the module 10 70 via a line 69 at the same rate as the permeate is removed therefrom via a line 72. The diafiltration results in the removal of additional lactose from the second stage ultrafiltration retentate. The retentate is recycled by a centrifugal pump 71 via a line '73 and 15 a heat exchanger 75 in the same manner as in the ultra¬ filtration modules.

The retentate is passed from the diafiltration module 70 via a line 77 into a second diafiltration module 78. The retentate of the module ^' 78 is recycled 20 via a line 79, and a heat exchanger 83 by a centrifugal pump 85. Water is added via a line 89 to the retentate at the same rate the permeate is removed from the module 78 via a line 90. The retentate is then passed via a line 87 to a final ultrafiltration module 88. 25 In the final ultrafiltration module 88, the retentate is recycled via a line 91 and through a heat - .. exchanger 92 by a centrifugal pump 93. The permeate is removed from the module 88 via a line 95. The permeate from modules 57, 63, 70, 78 and 88 are combined in a 30 line 97 and removed from the system via a line 100.

The water introduced into the system via a line 105, is pumped by a pump 110 into lines 115 and 120. A valve 125 controls the flow rate via line 69 and a valve 130 controls the flow rate via line 89. 35 The flow rate of water through line 105 is adjusted by ■

a conventional ratio controller 113 which receives an input from the flow meter 52 via conduit 114 and a flow meter 111 via a conduit 112.

The retentate from the final ultrafiltration 5 module 88 is passed via a line 135 to a conventional refractometer 140. The refractometer 140 determines the level of solids in the retentate. A signal in¬ dicative of the level of solids, generated by the re¬ fractometer 140 is passed to a conventional controller 10 145 via a conduit 147. The controller 145 then adjusts control valve 150 which varies the flow rate through a line 135 thereby controlling the residence time and the amount of solid of the retentate. The retentate then flows via a line 155 into a balance tank 160.

15 C. Pasteurization and Cooling

From the balance tank 160 the retentate is passed by a centrifugal pump 165 via a line 168 through a conventional plate heat exchanger 170 which includes a heating section 175 and a cooling section 180. In

20 the heating section 175 the retentate is pasteurized by being subjected to the temperature of 72 C for at least fifteen seconds or such other temperatures and times that lead to pasteurization of milk without de¬ naturing the whey proteins. The pasteurization

25 eliminates cultures which could affect the fermentation

^**• .. of the retentate.

The pasteurized retentate is cooled in the cooling section 180 to the optimum incubation temperature which is generally about 30 C.

30 D. Adjusting of Ionic Strength of the Retentate

From the heat exchanger 170, the retentate is passed into a line 185 and therein mixed with a slurry

of sodium chloride in water which is pumped from a tank 186 via a line 187 by a metering pump 190. In the al¬ ternative, sodium chloride can be added in crystalline form. The addition of sodium chloride is controlled by monitoring the conductivity of the retentate in a con¬ ductivity meter 200. The signal generated by the con¬ ductivity meter is fed to a controller 205 which controls the pump 190. The amount of sodium chloride must be sufficient to prevent the coagulation of the retentate during and after fermentation but must be low enough not to significantly retard the fermentation and not to exceed the allowable levels of sodium chloride in the final product. Generally the amount of added sodium chloride is in the range from about 0.5 to 4.2 weight percent of the retentate.

E. Fermentation

The retentate is next passed via a line 209 to one or more jacketed fermentation vessels 210, 220 and 230. The flow into these fermentation vessels is controlled by valves 235, 240 and 245, respectively.

The pH in the fermentation vessels 210, 220 and 230 is monitored by probes 250, 255 and 260, respectively, and the pH monitor 265. The content of the vessels 210, 220 and 230 can be stirred by agitators 267, 269 and 271, respectively. A starter culture such as Streptococcus cremoris is manually added to the concentrate in vessels 210, 220 and 230 and the fermentation is allowed to proceed until all lactose is used up or the pH reaches the desired level. The fermented retentate, whose vis¬ cosity is about the same as that of heavy cream, is then passed from one or more fermentation tanks 210, 220 or 230 to a surge (balance) tank 290. The fermented retentate from the tank 210 is pumped by a pump 290 via

lines 295 and 297 by opening a "valve 300. The fermented retentate from the tank 220 is pumped by a pump 310 via lines 312, 315 and 297 by opening a valve 317. The fermented retentate from the tank 230 is pumped by a pump 320 via lines 322, 325 and 297 by opening a valve 327.

F. Evaporation

A positive displacement metering pump 350 pumps the fermented retentate from the surge tank 290 via a line 352 to a conventional plate heat exchanger 355.

The fermented retentate is preheated by the heat exchanger 355 for a time period which is dependent upon feed rate. The preheat temperature is set to a level sufficient to prevent phase separation in the form of free fat (oiling off) during evaporation. The specific preheat temperature level selected will depend upon the percent total solids sought in the final product obtained from the evaporator. The preheat temperature level required increases with increase in the percent total solids in the final product and with increase in the retentate feed rate through the heat exchanger 355. For extended runs on a commercial basis, a swept or scraped surface type heat exchanger is preferred.

In addition to preventing oiling off during evaporation, the preheating step prior to evaporation has been found to reduce burn-on on the interior evapo¬ rator wall surfaces and to reduce the criticality of the evaporator rotor design because of a lower increase in viscosity during evaporation relative to the viscosity increase observed when preheating is either absent or not optimal. Lower viscosity of the ^' final cheese base has a further beneficial effect in that it lessens the load on the discharge system.

- J E ;

OMPI /fa

The heated fermented ^" retentate is then fed via a line 357 into the evaporator 360 where it is sub¬ jected to a vacuum and heat in order to evaporate water. The vacuum is provided via line 362 attached to a vacuum 5 . source (not shown). The heating is accomplished by steam passing via a line 365 and valve 367 into a jacket of the evaporator 360. The condensed water is removed from the jacket of the evaporator 360 via a line 370. The heat and vacuum effect the evaporation of water

10 which rapidly increases the viscosity of the fermented retentate. The evaporated fermented retentate is forced to discharge via a line 375 and it is passed through a refractometer 377 to determine the solids level. From the refractometer the final product is discharged via a

15 line 379. Generally the desired solids level is about 65 weight percent. The presently preferred evaporators are series HS evaporators sold by Luwa Corporation, Charlotte, N.C., U.S.A., modified by enlarging the dis¬ charge section of the evaporator to permit easier flow

20 of the dewatered, high viscosity final product.

The properties of the final product are such that it can be used for making processed cheeses in place of conventional cheddar cheese.

DESCRIPTION OF ANOTHER EMBODIMENT OF THE PRESENT 25 INVENTION

- . The process of the present invention can also be carried out in a batch system. The process steps are performed in a manner described in connection with the semi-continuous system except that an ultrafiltration

30 plant and a batch evaporator are used. The equipment for the batch process of the present invention is schematically depicted in FIG. 2.

Referring now to FIG. 2, whole milk is fed via lines 430 and 440 past a valve 445 into a jacketed

tank 500 where it is standardized and heated to about 50 C. Then the milk is pumped via a line 510 by a pump 515 to an ultra/diafiltration module 520. The permeate from the module 520 is discharged via line 525 5 and a flow indicator 530 into a measuring container 535.

The retentate is recycled via line 541 through a heat exchanger 545 back into the tank 500. The heat exchanger 545 cools the retentate to the preselected

10 ultrafiltration or diafiltration temperature which is generally about 50 C.

The amount of permeate collected in the con¬ tainer 535 is monitored. When the amount of permeate reaches a predetermined level such as sixty percent of

15 the volume, cold water is fed via lines 433 and 440 and past the valve 445 into the tank 500 at the same rate as the permeate is discharged from the module 520 via the line 525. The diafiltration continues until a predetermined ratio of lactose to buffer capacity (5.7)

20 is achieved. Then the valve 445 is closed and the ultra¬ filtration is resumed and continued until the amount of permeate reaches a predetermined volume. For the cheese base of the composition of cheddar cheese, such volume is about 4/5 of the initial volume plus the amount of

25 water added for diafiltration.

The retentate is then discharged via a line

- .. 537 past a valve 539 into a vessel 550 where 1% of sodium chloride by the weight of the retentate is introduced.

30 A starter culture (Streptococcus cremoris type E8) is then manually added to the retentate in the vessel 550 and the content of the vessel 550 allowed to ferment until the pH is lowered to a predetermined level.

The fermented retentate is pumped by a pump 555 via a line 560 into a swept surface type of evapo¬ rator 565 where sufficient amount of moisture is re¬ moved from the fermented retentate to achieve the de- sired solids level in the final product. For a cheese base having a composition of cheddar cheese the desired solids level is about 65 weight percent. The final product is collected in a container 600.

The line 560 preferably includes a flow through heat exchanger (not shown) like heat exchanger 355 in FIG. 1 for preheating the fermented retentate prior to evaporation in the manner previously described. Al¬ ternatively, the fermented retentate may be preheated in a atch manner.

EXAMPLE 1

100 kilograms of pasteurized whole milk was heated to 50°C and subjected to membrane ultrafiltra¬ tion in a Paterson-Candy International (PCI) 8 foot (240 cm), series flow, ultrafiltration module, with T6B membranes. The pressure at the inlet of the module was 90 psi and the pressure at the outlet of the module was 20 psi. The milk was thus separated into a retentate fraction which contains essentially all the fat together with over 96% of the protein and insoluble salts, and a ^* permeate fraction which contains mostly water, lactose, and soluble salts. Ultrafiltration was continued until approximately 60 kg of permeate had been removed.

At that point a stream of water at 50°C was introduced into the feed tank at the same rate permeate was expelled from the ultrafiltration membrane. This process is known as diafiltration at constant volume

-δUREA

OMPI Λ-AT10

since the volume of retentate is kept constant by con¬ trolling the input of diafiltration water. When ap¬ proximately 35 kg of diafiltration water had been added, the water was turned off. The ultrafiltration was then resumed until the amount of retentate remaining was approximately 20% of the weight of the original milk. The retentate at this point contained an appropriate ratio of buffer capacity to lactose so that when the lactose was fer- ented to lactic acid, the pH of the product was be¬ tween 5.1 and 5.2. The appropriate ratio was achieved by means of diafiltration and final ultrafiltration. The retentate was removed from the ultra¬ filtration equipment and salt was added in an amount such that when the final moisture content of the pro¬ duct was 36% the salt in the water phase equaled 4.5%. The retentate was then cooled to 25°C and inoculated with a culture of Streptococcus lactis C6 and incubated at 25°C until the desired pH of 5.2 was achieved. The fermented retentate was then introduced into a swept surface vacuum evaporator. The operating conditions of the vacuum evaporator were selected so that the vapor temperature in the evaporator was in the range 45 to 55°C. The moisture was removed therein by evaporation until the final product containing about 36% moisture was achieved. The composition of the retentate after the second ultrafiltration step but before fermentation and that of the final product were as follows:

Retentate Retentate before after fermentation evaporation

(weight percent) (weight percent)

Total solids 40.65 64.0

Fat 21.36 33.6

Protein 15.9 25.0

Lactose 1.0

Calcium 0.48 0.76

Orthophosphate 0.77 1.22

Sodium chloride 1.6

Lactic acid 1.6

The pH, buffer capacities, and buffer capacity/la ratio, were as follows:

Retentate Retentate before after fermentation evaporation

(weight percent) (weight percent)

PH 6.82 5.20

Buffer capacity 19.5 meg HCl/lOOg

Buffer capacity/lactose 19.5

The final product was pressed into rectangular blocks and sealed in gas-proof wrappers until required for processing. It was found suitable for conversion to processed cheese by the techniques used for the pro¬ cessing of conventionally-made cheese.

EXAMPLE 2

100 kilograms of pasteurized whole milk was acidified with concentrated hydrochloric acid to a pH of 5.7. The acidified milk was then heated to 50°C and ultrafiltered in the module and in the manner described in Example 1 until the weight of permeate removed was equal to 60% of the weight of the original milk.

-$ EA

OMPI

.A ^, wi ^p o N lO

The retentate was the ^~ n diafiltered at con¬ stant volume as described, above until 65 kg of dia¬ filtration water was added. The water was then turned off and ultrafiltration was resumed and continued until the weight of the retentate was reduced to 20% of the weight of the original milk. The fermentation, evapora¬ tion, and packaging steps were then carried out as de¬ scribed in Example 1. The final product was suitable for conversion to processed cheese. The calcium and phosphate content of the final product made in this Example were different from those of the final product made in Example 1. The properties of the processed cheese made from the final product of this Example were not markedly changed from those of the processed cheese made from the final product of Example 1.

The composition of retentate after the second ultrafiltration step but before fermentation and that of the final product were as follows:

Retentate Retentate before after fermentation evaporation

(weight percent) (weight percent)

Total solids 36.9 64.0 Fat 19.6 33.6

Protein 15.6 25.6 Lactose 0.6 Calcium 0.26 0.45

Orthophosphate 0.49 0.85 Sodium chloride 1.6 Lactic acid 1.04

The pH, buffer capacity, and buffer capa¬ city/lactose ratio were as follows:

OMPI WIPO

Retentate Retentate before after fermentation evaporation

(weight percent) (weight percent)

PH 5.87 5.10

Buffer capacity 8.30

Buffer capacity/lactose 13.8

The yields of final products in Examples 1 ^* and 2 were 17.8% and 16.4%, respectively, greater than the yields of cheese which could have been made from the same quantity of milk by conventional cheesemaking technology. The precise yield increase is dependent on the protein/fat ratio of milk. The yield increases result from the retention of whey proteins in the final product. Since whey proteins are not removed during the process and since the protein composition of the products must reflect that of the original milk, ap¬ proximately one-fifth of the protein content of these products must be whey protein. The relative amounts of the various milk components vary (for example, with season, breed of cattle, etc.) and these variations will be reflected in the relative amounts of the com¬ ponents in concentrates prepared in other times and places and in the yields obtained by the process. The general procedures utilized in Examples 1 and ^' 2 may be modified. For example, using the same general procedures of Examples 1 and 2 rennet and/or any other safe and suitable proteolytic or lipolytic enzyme may be added to the retentate prior to evapora- tion or at the same time as the lactic acid producing bacteria, provided that the quantity is insufficient to produce a coagulum. The enzyme treatment may make the product more suitable for conversion to processed cheese in some applications. Furthermore, using the general procedures utilized in Example 1 (with or without the rennet or

enzyme) the appropriate ratio ό"f buffer capacity to lactose concentration may be achieved by diafiltering until 48 kg of water have been added, then ultrafilter- ing until the retentate is 30% of the weight of the original milk. This achieves the same ratio of lactose to buffer capacity but permits optimization of the se¬ lected amount of moisture removed by ultrafiltration and evaporation.

In addition, using the general procedures utilized in Example 2 (with or without the rennet or enzyme) the appropriate ratio of buffer capacity to lactose concentration may be achieved by diafiltering until 91 kg of water have been added, then ultrafilter- ing until the weight of the retentate is 30% of the weight of the original milk. This procedure changes the water removal process by variation of the relative extents of ultrafiltration and diafiltration, while retaining the same lactose to buffer capacity ratio in the final product. The following examples 3 and 4 are examples of the process of this invention in semi-continuous form particularly illustrating the step and effects of preheating the fermented retentate prior to evapora¬ tion.

EXAMPLE 3 Whole milk, 2500 liters, was adjusted to a casein/fat ratio of 0.63 by the addition of 19.4 liters of cream containing 42% butterfat. The standardized milk was pasteurized at 72°C for 15 seconds and was then ultrafiltered and diafiltered continuously through an Abcor model HFK130 membrane system containing 4 stages with 2 membranes per stage. The feed rate during ultra- filtration/diafiltration was 590 liters/hour. Diafiltra¬ tion water was fed into stages 2, 3, and 4 at 40% of the initial milk volume. The operating temperature of

- ϋ ZEA c OMPI WIPO

the membrane system was maintained at 52°C and the pres¬ sures maintained were 90 psi at the inlet and 30 psi at the outlet. Retentate was withdrawn from the continuous system at a rate of 126 liters/hour. 5 The 4.7 X retentate obtained, 530 liters, was pasteurized at 72°C for 15 seconds and cooled to 25°C. Salt, in the form of dry powder, was added to the pas¬ teurized retentate to a level of 1.6% water phase. The salted retentate was then innoculated with Streptococcus 10 cremoris and Streptococcus lactis mixed starter culture in the form of a frozen concentrate at a level of 0.01% and incubated at 25°C, until a desired pH of 5.1 was obtained.

The fermeted retentate at 25°C was passed 15 through an in-line Votator model 3/4 BW VOT scraped surface heat exchanger preheater. In the initial part of the run, the preheater was not turned on. Subse¬ quently the preheater was turned on and the fermented retentate was preheated for approximately 22 seconds of 20 flow through time duration to temperatures in the range of from 53°C to 73°C. The fermented retentate was fed continuously to a Luwa model NL0050 swept surface vacuum evaporator at rates ranging from 198 pounds (89.81 kg) per hour to 217.8 pounds (98.79 kg) per hour. The rotor 25 speed of the evaporator was maintained as 1200 rpm and the vacuum at 700 mm(Hg) . During evaporation the jacket - . temperature of the evaporator was maintained at 127°C at the inlet and 125°C at the outlet by circulatory hot oil. The vapor temperature during evaporation was 34°C, 30 and the product discharge temperature was 36°C. The feed rate to the evaporator was used to control the solids level of the final product in the range of from 58.51% to 65.95%. The final product was packaged in barrier bags and sealed for storage.

In this example, the'following observations were made during the run relevant to the effects of preheating on phase separation:

% Total Approx. Solids

Preheat Feed Rate Preheat in Final Characteristics

Time Temp. Ibs./hr Time Product of Final Product

Startup 11:01

11:20 None 217.8 None 58.51 No phase separat

11:30 None 207.9 None 59.24 No phase separat

11-.37 None 198 None 60.85 Phase separation

11:55 53°C . 207.9 72 sec. 61.89 Phase separation

12:00 54°C 198 72 sec. 62.35 Phase separation

12:15 65°C 207.9 72 sec. 63.86 No phase separat

12:23 66°C 198 72 sec. 64.24 No phase separat

12:32 66°C 193.1 72 sec. 65.85 No phase separat

12:45 73°C 217.8 72 sec. 62.13 No phase separat

1:00 73°C 207.9 72 sec. 63.46 No phase separat

1:20 73°C 207.9 72 sec. 64.18 No phase separat

1:40 73°C 207.9 72 sec. 64.13 No phase separat

1:55 73°C 207.9 72 sec. 65.95 No phase separat

2:15 73°C 207.9 72 sec. 64.09 No phase separat

2:30 73°C 207.9 72 sec. 63.95 No phase separat

2:37 Run ended.

In this example 2500 liters of milk were processed into 520 liters of 4.7 X retentate which produced 663 lbs (300.73 kg) of final product. This represented a 12.05% recovery (yield) from milk. Composition of the retentate before fermentation and of the final product were as follows:

Retehtate Final Product wt. J wt. %

Total Solids 36.1 64.2

Fat 19.2 33.2

Protein 14.4 25.2

Lactose 1.3 0.2

Calcium 0.40 0.70

Orthophosphate 0.72 1.36

Sodium Chloride 1.6

Lactic Acid 1.9

PH 6.85 5.05

Buffer Capacity 17.3 (meg HCL/lOOg)

Buffer Capacity/Lactose 13.3

It was further observed from this example that preheating the fermented retentate sufficiently to prevent phase separation had a tendency to reduce the viscosity of the final product, and reduce burn-on on the interior wall surface of the evaporator.

EXAMPLE 4

Whole milk, 5460 liters, was adjusted to a casein/fat ratio of 0.63 by the addition of 45 liters of cream containing 40% butterfat. The standardized milk was pasteurized at 72°C for 15 seconds and was then ultrafiltered and diafiltered continuously through an Abcor model HFK130 membrane system containing 4 stages with 2 membranes per stage. The feed rate during ultra- filtration/diafiltration was 570 liters/hour. Diafiltra¬ tion water was fed into stages 2, 3, and 4 at 36% of the initial milk volume. The operating temperature of the membrane system was maintained at 52°C and the pres¬ sures maintained were 90 psi at the inlet and 30 psi at

the outlet. Retentate was withdrawn from the continuous '.^system at a rate of 116 -liters/hour.

The 4.9 X retentate obtained, 1114 liters, was pasteurized at 72°C for 15 seconds and cooled to

5 25°C. Salt, in the form of dry powder, was added to the pasteurized retentate to a level of 1.6% water phase. The salted retentate was then innoculated with Strep¬ tococcus cremoris and Streptococcus lactis mixed starter culture in the form of a frozen concentrate at a level

10 of 0.01% and incubated at 25°C, until a desired pH of 4.95 was obtained.

The fermented retentate at 25°C was preheated in-line in a Votator model 3/4 BW VOT swept surface preheater in a range from 68° to 75°C for approximately

15 70 seconds and fed to the Luwa model NL0050 swept sur¬ face vacuum evaporator at rates ranging from 287 to 396 lbs/hour (130.18 to 179.62 kg/hour). The rotor speed of the evaporator was maintained at 1200 rpm and the vacuum at 720 mm(Hg). During evaporation the jacket

20 temperature of the evaporator was maintained at 110°C at the inlet and 109°C at the outlet by steam at a pres¬ sure of 50 KPA. The vapor temperature during evapora¬ tion was 34.4°C, and the product discharge temperature was 38.6°C. The feed rate to the evaporator was used

25 to control the solids level of the final product in the range of from 58.3% to 66.4%. The final product was

^■ * .- packaged in barrier bags and sealed for storage.

In this example 5460 liters of milk were pro¬ cessed into 1114 liters of 4.9 X retentate which pro-

30 duced 1,406 lbs (637.75 kg) of final product. This represented a 11.71% recovery (yield) from milk. Com¬ position of the retentate before fermentation and of the final product were as follows:

Retentate Final Product wt . % wt. %

Total Solids 37.3 66.1

Fat 19.4 33.5

Protein 14.9 25.7-

Lactose 1.5 0.1

Calcium 0.44 0.77

Orthophosphate 0.79 1.36

Sodium Chloride 1.74

Lactic Acid 2.1 pH 6.8 4.9

Buffer Capacity 18.1 (meg HCL/lOOg)

Buffer Capacity/Lactose 12.1

In this example, no phase separation occurred in the evaporator in a total running time of 7 1/4 hours with percent total solids in the final product as high as 66.4%. It was again observed that preheating the fermented retentate sufficiently to prevent phase separa¬ tion had a tendency to reduce viscosity of the final product and to reduce burn-on on the interior wall surface of the evaporator.

It should be noted that a percent total solids in a final product sample as high as 71.5% has been achieved without phase separation with a preheat tempera- ' ture of 72 C at flow-through preheat time duration of 70 seconds.

Those skilled in the art will appreciate that the invention described herein is susceptible to varia¬ tions and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications which fall within its spirit and scope.