mm? o; vmm~s m papame ENTALLAfiGE‘éS GER? ‘E'HEE ACID SEGREE 05‘ 33A? EN MELK Thesis {at {he Dag-m 35 M. S. ME'CHEGAEQ STA'E‘EE umwmsm' Louis leksy’ 5?;5 é at“- TN" EFFECT OF VARIANTS IN PIPELINE INSTALLATIONS ON THE ACID DEGREE OF FAT IN MILK By LOUIS £93ch A THESIS Submitted to the School of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Dairy 1956 at“ W—IOvSL ACKNOWLEDGEMENTS The author wishes to thank Professor J. M. Jensen, under whose direction the project was develOped, and for his assistance in the preparation of the manuscript. The writer is also indebted to Dr. G. M. Trout for his words of encouragement and advice. Appreciation is further expressed to both men for their assistance in evaluating the quality of the milk samples. ii TABLE OF CONTENTS Page INTRODUCTION 1 REVIEW OF LITERATURE . . . . . . . . . . 2 Description of Rancid Flavor 2 Direct Cause of Rancidity 2 Lipase in General . . . . . . . . . . . . . . 3 Nature of Milk Lipase h Causes of Lipolysis 5 Spontaneous Rancidity 5 Individuality of cows 6 Stage of lactation 6 Season and feed 8 Breed and milk production of cows 8 Activating factors 9 Prevention and inhibition 9 Induced Rancidity . . . . . . . . . . . . . . 10 Mechanical agitation by shaking . . . . . , 10 Activation by cooling-warming-cooling . . 11 Mechanical activation by handling of raw milk . . . . . . . . . . . . . . 11 Factors of Induced Rancidity . . . . . . . . . 15 The surface of the fat globule . . . . . . 15 The melting point of fat . . . . . . . . . 16 iii Fat:plasma ratio Inactivation of lipase by formaldehyde Preventing Activation During Handling Tests Used for Measuring Hydrolysis of Butterfat . . . C°ganoleptic examination Surface tension measures _ents Chemical measurements of fat soluble aCIdS O O O O O O O I 0 EXPERIMENTAL PROCEDURE Description of Pipeline Installations Line "A" Line "A1" . . . . Line "B H Line "Bl" Lille ”CH Line "D" . . . . . . . . . . . . . Sampling Procedure . . . . Analytical Methods Preparation of Reagents Preparation of purified absolute alcohol Preparation of ethanolic KOH "Skellysolve F" . . . . . Organoleptic Examination Surface Tension Measurement EXPERIMENTAL . . . . . . . . . . . . . . . . . iv Effect of Installation "A" on F.F.A.O Of Milk . . . . . . . . . . . . . . . . . . . 32 Effect of Line ”A" on F.F.A.O of Milk Fat When Milk is Admitted in Volume at Distant End of Line . . . . . . . . . . . 33 Influence of Breed on F.F.A.° in Line "A". . . 39 Effect of Altering ine "A" to "A1" on F.F.A.O of Milk . . . . . . . . . . . . . . . 39 Effect of Milker Installation "B" on F.F.A.° of Milk . . . . . . . . . . . . . . . 42 Effect of Altering Line "B" to "B1". . . . . . 4N Lipolysis of Individual Cow Milk Samples . . . 45 Effect of Line "C" on F.F.A.O of Milk . . . . #7 Effect of Length of Milk Line Without Risers on F.F.A.O of Milk as Observed on Installation "D" . . . . . . . . . . . . . . . 43 Seasonal Differences in F.F.A.° of Milk . . . #9 Relationship between Rancid Flavor and F.F.A.0 of Milk . . . . . . . . . . . . . . . 50 Relationship between Rancid Flavor and Surface Tension Measurement . . . . . . . . . 51 DISCUSSION . . . . . . . . . . . . . . . . . . . . 53 SUMMARY . . . . . . . . . . . . . . . . . . . . 60 BIBLIOGRAPHY . . . . .-. . . . . . . . . . . . 62 Table II. III. IV. VI. VII. VIII. IX. .XI. LIST OF TABLES Influence of Length of Line and of Risers on F.F.A.O of Milk in "A" Holstein Breed Influence of Length of Line and of Risers on F.F.A.° of Milk in "A" Brown Swiss Breed Influence of Length of Line and of Risers on F.F.A.O of Milk in "A" Ayrshire Breed . . . . . . Influence of Length of Line and of Risers on F.F.A.O of Milk in "A" Guernsey Breed . . . . . . Influence of Length of Line and of Risers on F.F.A.3 of Milk in "A" Jersey Breed . . . . . . . Height Pipeline Height Pipeline Height Pipeline Height Pipeline Height Pipeline Effect of Distance and Risers on F.F.A.O Admitting 10 Gallon Volumes of Milk into Distant end of Line "A" . . . . Breed Influence on F.F.A.° When Milk was Admitted into Line "A" in 10 Gallon Volumes Influence of Line "A1" on F.F.A. Various Distances of Flow Risers on F.F.A. O with Influence of Length of Line and Height of "B H and "BI H of Milk in Pipelines Relationships Between Spontaneously Rancid Flavor, Surface Tension, and F. Milk from Individual Cows . . . F.A.° of F.F.A.O in Milk from Milking Parlor Weigh Jars and Subsequent Bulk Tank Samples of Line "C" . . . . . vi Lo Lo 38 no 41 43 #8 fl h: g., ‘2: Table XII. XIII. XIV. XV. F. F. A. 0 of Milk from Stanchion Barn Pipeline "D" No Risers . . . . . . . . . . . Seasonal Differences in F.F.A.O from Various Milking Installations . . . Relationship Between Rancid Flavor and F. F. A. 0of Milk . . . . . . . . . Relationship Between Rincid Flavor and Surface Tension of Milk . . . vii Page 49 50 51 52 Figure Duh)?“ 4;:- LIST Pipeline Installation Pipeline Installation Pipeline Installation Pipeline Installation OF FIGURES "A H AI "B H "B1 H viii Page 29 30 31 INTRODUCTION The taste quality of milk is one of its most distin- guishing attributes. Taste of fresh milk may be damaged severely and rapidly by decomposition of butterfat to fatty acids with resulting rancidity. Rancidity of milk is not a new problem. Trouble caused by rancidity heretofore has been attributed mainly to the physical conditions affecting cows during the later part of their lactation. Usually not more than one or two cows in a herd were affected. But with the advent of pipe- line milking and bulk tank cooling, a considerable number of such instances have plagued producers of large volumes of milk. Investigations made elsewhere have uncovered a number of conditions that stimulate the development of rancidity in raw milk. Considerable data however appear to be contradictory, and solutions frequently must be found on the basis of a particular set of conditions. This study was made for the purpose of determining the conditions in pipeline milking which produce damage to milk, causing rancidity, and of finding practical solutions of installing and operating pipelines with minimum effects on lipolysis. REVIEW OF LITERATURE Description of Rancid Flavor Rancid flavor, as the term is used in the dairy in- dustry, means a specific flavor, identifiable as the flavor of free butyric acid. Tarassuk and Jack (1949) and Herring— ton (1950) described rancid flavor as bitter, soapy, "goaty", "wintry", and sometimes resembling the flavor of coconut. In mild cases, rancidity has been identified as stale, old, and unclean. According to the eXperience of Hileman and Courtney (1935): rancid flavor also might be recognized as the peculiar burning sensation at the back of the tongue when the milk is swallowed. Herrington (195h) stated that rancid flavor is caused by the hydrolyzation of the short chain, water soluble, fatty acids, while the long chain fatty acids have little effect on flavor. Direct Cause of Rancidity The principal factors setting off lipolysis in milk Eire lipase and butterfat. Herrington (l95h) stated that it Its commonly accepted that cows' milk contains the fat- splitting enzyme lipase. He further made it clear that tnore than one lipase enzyme is involved and that butterfat «can be split apart by enzymatic hydrolysis of lipase to Iiield glycerine and a mixture of about a dozen different 2 fatty acids. Some of these fatty acids, particularly the short carbon chain groups such as butyric acid, caproic acid, and caprylic acid, have very powerful odors. There is some disagreement about the phase in which the lipolytic factor is carried. Pfeffer, Jackson, and ‘Weckel (1938) were of the opinion that the lipolytic factor is carried in the serum phase of milk, and the extent of lipolysis depends upon the fat content of milk. However, Frankel and Tarassuk (1955) contend that the lipolytic factor is in the fat phase, since the kind of fatty acids present in rancid milk is related to that of the original composition of the fat. Lipase in General A few decades ago, dairy chemists doubted that normal raw milk contained lipase. Through the investigation of the causes of bitter or rancid milk, the presence of lipase in milk has been convincingly demonstrated. Palmer (1922) reported that Harrison, in 1902, believed that the cause of bitter milk might be of bacterial origin. Later Mass (1909) and Rogers and associates (1912, 1913) demonstrated the presence of lipase in raw milk. Palmer (1922) also came to the conclusion that bitter milk of advanced lactation is caused by the secretion of the enzyme lipase, which will hydrolyze milk fat and liberate fatty acids. Furthermore he stated that the lower volatile fatty acids are respon- sible for bitter taste. Rice and Markley (1922) demonstrated 3W the presence of lipase by the increase in acidity of milk. The presence of lipase in raw cow's milk has been confirmed also by Rice (1926) and Roadhouse and Koestler (1929). Hileman and Courtney (1935) demonstrated the presence of lipase with an increase in acidity that was accompanied by bitter flavor. They ruled out the idea that milk lipase is of bacterial origin and believed it to be secreted by the cow along with the milk. Tarassuk and Jack (1949) proved that the lipase of naturally lipolytically active milk is present in the milk plasma, prior to cooling. Experimental data of Kelly (1943) indicahfi,that the tissue of the bovine mammary gland, deve10ped by pregnancy, contains a significant amount of lipase. However, Herring- ton (1950) stated that milk lipase has very little action on fat inside the udder, and, as a general rule, the lipase in milk seems inactive at the time of milking. Nature of Milk Lipase Dairy scientists presently agree that raw cow's milk contains lipase. Herrington (1954) is of the opinion that still very little is known about the lipase as such. Accord- ing to Davies (1932), milk lipase is a compound of several enzymes. Dorner and Widmer (1932) suggested the presence of two lipases, one of which is extremely heat labile and responsible for the sharp bitter taste in homogenized milk. Harrington and Krukovsky (1939a) found that formaldehyde inhibits lipolysis. They presented data to show that there are present in milk two different lipases, one of which is not affected by formaldehyde. This lipase is believed responsible for the rancid taste in homogenized milk. Harrington (1950, 1959) stated that lipase enzyme is not a single enzyme, but rather a group of enzymes, and that milk lipase is a group of highly specialized enzymes, some of which can be split into a thermostable coenzyme and a thermolabile apoenzyme. Causes of Lipolysis Lipolysis in raw milk may occur spontaneously, or it may be induced by mechanical activation. Under certain conditions, the inactive lipase present in fresh raw milk may be activated or accelerated for enzymatic hydrolysis. Herrington (1954), describing activation in general, stated that it seems probable that activation is more dependent upon changes in the state of fat than upon changes in the lipase. Dunkley and Kelley (1954b) stated that the so-called fat globule membrane helps to protect fat in milk from the lipase enzyme. If this protective membrane is disrupted or modified, the lipase is able to hydrolyze the fat, freeing the fatty acids . Spontaneous Rancidity When samples of milk are drawn from a number of cows and cooled and stored separately, some of them will develop rancidity. This kind of rancidity has been described by 6 Herrington (1950) as spontaneous. Freeden and Bowstead (1951) in a review of factors causing spontaneous rancidity, reported that spontaneous rancidity varies with the individuality of the cow, stage of lactation, season, and feed. They also reported that temperature, hours of sunshine or artificial light, and the amount of exercise that cows need, have no effect on spontaneous rancidity. As will be shown later, according to the eXperimental results of numerous workers, there is no definite relationship between breed and milk production of cows and spontaneous rancidity. Individuality of cows; Experimental data of Herrington and Krukovsky (1939b), Roahen and Sommer (1940), Krienke (19flfl), and Freeden and Bowstead (1951) indicated that there is a great deal of variation in the rate of lipolysis in the milk from different cows, during the same lactation, in spite of similar conditions of feed and season. Herring- ton and Krukovsky (1939!) found that the difference in acid- ity of fat may become pronounced only after a few hours of storage. According to Roahen and Sommer (1990) and Thomas, Harper, and Gould (1954), there is a variation of fatty acid content not only between different cows but also between successive daily milkings from the same cow. §£2§E of lactation. The lipolytic activity of milk of certain cows is accelerated by advanced lactation as reported by Sharp and DeTomasi (1932), Trout (1932a, 1932b), Bailey (1933), and Krukovsky and Sharp (1936). 7 Palmer (1922) and Hileman and Courtney (1935) believed that spontaneous rancidity in advanced lactation is due to increased lipase secretion. Fouts and Weaver (1936) found that spontaneous rancidity may occur any time within the milking, but it is most likely to appear in late lactation.- Hileman and Courtney (1935) and Krukovsky and Sharp (1938) stated that some cows in advanced lactation in winter will produce milk which will become rancid spontaneously upon cooling. Some workers do not agree that advanced lactation or advanced gestation is the most important factor of spon- taneous rancidity. Weaver (1939) believed that advanced gestation was more important than advanced lactation, while Freeden and Bowstead (1951) believed that advanced lacta- tion was the contributing factor, rather than advanced gestation. Herrington and Krukovsky (1939b), Pfeffer, Jack- son and Weckel (1938), and Roahen and Sommer (1990), testing a large number of individual cows for a long period of time, found that there was no apparent correlation between the rates of lipolysis and the stage of lactation or gestation. Also Thomas, Harper, and Gould (1954) could not observe any increase in fatty acid content with the progress of lactation. According to Kelly (1945), there is a close relation- ship between the stage of the estrous cycle of the cow and the lipase of the milk she produces. Herrington (1954) mentioned in his "Review" on lipase that eXperimental results indicated a maximum concentration of lipase on the day of oestrus and a maximum acidity of the fat on the day after oestrus. Season and feed. Several workers have presented evi- dence that there is a strong relationship between feed and season and between susceptibility of milk to spontaneous rancidity. Csiszar (1933) and Hileman and Courtney (1935) noted that lipolytic activity in milk is more pronounced in winter than in summer. According to Herrington (1950), the number of cows producing milk susceptible to spontaneous rancidity reaches a maximum in late fall and will diminish through the winter. There are certain indications that the seasonal varia- tions in susceptibility to spontaneous rancidity has some connection with summer and winter feeding. Anderson (1936), Fouts and Weaver (1936), and Freeden and Bowstead (1951) believed that spontaneous rancidity is most common in the fall or winter, due to the lack of green feed or, more specifically, of the carotene therein. Trout, Jensen and Humbert (1955) claimed that green pasture has a stabilizing effect against spontaneous rancidity during the summer. Breed and milk production of cows. Pfeffer, Jackson, and Weckel (1938) observed less lipolytic activity in milk of individual cows when the milk production was lower. How- ever, Herrington and Krukovsky (1939b) found no relationship between lipolysis and the amount of milk produced by each cow. Activating factors. Hileman and Courtney (1935) and Tarassuk and Henderson (19#2) stated that the only action needed to activate spontaneous rancidity is to cool the milk -and hold it for some time under low temperature. According to Tarassuk and Jack (19¢9), when the milk is cooled, the lipase is adsorbed on the fat globules, and the lipolysis begins immediately. Furthermore, Tarassuk and Henderson (l9k2) found that the addition of a very small amount of spontaneously rancid milk to a normal milk supply will impart the rancid flavor to the entire mixture. Prevention and inhibition. Tarassuk and Henderson (19#2) found that the development of spontaneous rancidity could be prevented by mixing naturally active lipase milk with normal milk, in a proportion of one part to four parts, within an hour after milking. They believed that in order to insure effectiveness of this method, the milk should be mixed before cooling. According to the results of several experiments by Fouts and Weaver (1936), Roadhouse and Koestler (1929), Roadhouse and Henderson (1935) and Weaver (1939), from 2 to 22 percent of the cows produced spontaneously rancid milk at any time. Dunkley and Kelley (1954b) recommended the elimination of cows which produce milk that is highly susceptible to spontaneous rancidity. 10 Induced Rancidity If the milk is subjected to certain treatments, the action of lipase upon fat is increased greatly. According to Krukovsky and Sharp (1940a) and Tarassuk and Henderson (1942), certain mechanical activation such as violent shaking of raw milk, warming precooled milk and cooling it again, and various methods of handling raw milk will cause a change in the fat globule. These investigators concluded that the change in the fat globule is responsible for mechanical activation. Tarassuk and Jack (1949) also concluded that the activation treatments of induced lipolysis lead to the disruption, partial displacement, or distortion of the natural adsorption layer on the fat globules, and lipase is able to split the fat partially, liberating fatty acids responsible for the rancid flavor. Mechanical agitation bymghaking. It has been known for a number of years that shaking of warm raw milk will induce lipolysis. Krukovsky and Sharp (1938), Roahen and Sommer (1940), and Harrington (1950) agreed that vigorous shaking of raw milk while the butterfat is even partially in the liquid state, will induce lipolysis upon cooling. They agree that vigorous shaking of warm milk partially destroys the fat globule membrane and makes the fat available for lipolysis. Holland and Herrington (1953) and Herrington (1954) believed that effectiveness of agitation depends on the fluidity of the butterfat, the rate of lipolysis increasing rapidly as the temperature is increased up to 750-850F. 11 Activation by cooling-warming-cooling; Krukovsky and Herrington (1939) found that lipolysis in cold raw milk could be activated by cooling milk to 41°F. or lower, rewarming to 86°F. and then recooling below 50°F. The length of time the milk was held at the activation temper- ature was less important than the actual temperatures employed. Mechanical activation by handling of raw milk. Some methods of handling raw milk created favorable conditions for mechanical agitation and induction of rancidity. Such methods of handling milk are: (l) the changes in temperature of separation of cream, (2) homogenization, and (3) the introduction of pipeline milkers and farm bulk tanks. (1) Cold separation of cream. Sharp and DeTomasi (1932), Roahen and Sommer (1940), and Krukovsky and Herrington (1939) found that the incidence of rancidity increased when raw milk was separated at BOO-90°F. The latter investigators believed that the cause might be warming activation. (2) Homogenization. Dorner and Widmer (1931), Halloran and Trout (1932), Trout (1933), and Gould and Trout (1936) found that homogenization of raw milk will speed up lipolysis within a few minutes. This fast increase in rancidity was attributed to a radical change in the surface of fat globules and to increased contact between fat and lipase. at“. 12 (3) Use of pipeline milkers and farm bulk tanks. Since the introduction of pipeline milkers and farm bulk tanks, the incidence of rancidity has increased throughout the main fluid milk producing areas. Research workers agree that, under certain conditions, pipeline milking systems will produce activation in the raw milk sufficiently violent to cause lipolysis. Herrington (1954), Dunkley and Kelley (1954b), and Thomas, Nielsen, and Olson (1955b) reported several cases of rancidity in raw milk handled by pipeline milkers and bulk tank installations. The same authors reported that, in several cases, the trouble was caused by the pipeline "risers," vertical sections connecting pipelines at different levels. In a riser, the milk is carried upward by a stream of air. Dunkley and Kelley (1954b) found that milk hoses which carry milk from the cow to the line could have an activating effect if excess air bubbles through them. Tarassuk and Frankel (1955) believed that air agitation will create foaming and will induce ran- cidity more than any other mechanical agitation. They explained that bubbling air through warm milk could cause a partial transfer of the fat surface material, creating conditions favorable for lipolysis. Dunkley and Kelley (1954b) conducted trials using horizontal pipe sections fitted with risers and 13 fittings, and they observed activation only when air was admitted. Air agitation in milk pipelines has been found to be influenced by a number of factors. Dunkley and Kelley (1954b) found that milk flow rate had a marked influence on the activating effect of air. They found that higher flow rates decreased agitation in the risers, thereby reducing activation. The effect of agitation is also found to be influenced by the temperature of raw milk. Holland and Herrington (1953) and Herrington (1954) believe that as milk cools, it passes through a critical temperature, where it is most susceptible to churning. Herrington is of the Opinion that the effectiveness of agitation in the role of inducing rancidity is influenced by other conditions, particularly temperature, effecting fluidity of the fat. He found sensitivity to acti- vation increased rapidly as the temperature increased to 75° or 85°F. Since the milk pump and inline filter are parts of the pipeline bulk tank system, their effect on raw milk must not be overlooked. The experimental data of Dunkley and Kelley (1954b) show that agitation of warm raw milk might occur through the continuous operation of the centrifugal pump at a flow rate below its capacity. The same authors reported that consider- able rancidity occurred when a bag or a plate-type Illl'llllllllll’llllF)’ 14 filter was used in the vertical section of the vacuum line. However no rancidity occurred when the filter was installed in the vertical discharge line from the pump. Herrington and Krukovsky (1939) found that when storing normal milk, some lipolysis will take place with increase in free fatty acid degree. However, the rate of the lipase action was retarded by fast and efficient cooling. According to Herrington (1954) and Dunkley and Kelley (1954b), activation might occur in a bulk tank if fresh warm milk is added to cold milk. Herrington (1954) found that when the temper- ature of milk in the tank exceeds 6803., lipolysis might be induced due to temperature activation. The agitator of the bulk tank has been considered as possibly having some influence on induced rancidity. According to Herrington (1954), excess agitation might occur when warm milk enters the bulk tank and only partially covers the agitator. Dunkley and Kelley (1954b) suggest that an optimum condition of stirring milk should be employed to get minimum agitation. This would be accompanied by rapid cooling of the incoming milk in order to get the least induced lipolysis. Herrington (1954) and Dunkley and Kelley (1954b) warned that with the introduction of farm tanks and tank truck transportation, there will be a growing 15 interest in alternate day pick-up from the farms. This would necessitate an increased effort to control lipase action. Herrington (1954) suggests that some raw milk supply, which will not develop rancidity during a short period of storage, might do so if held over 24 hours. None of the data show conclusive evidence that bulk tank cooling is a serious factor in induced milk rancidity. Factors of Induced Rancidity Apparently, milk responds to induced lipolysis because of the nature of several conditions that affect the fat in milk. Krukovsky and Sharp (1940a) mentioned some important factors influencing the rate of induced lipolysis, such as, the surface of the fat, the melting point of the fat, and the fat:p1asma ratio. The surface of the fat globule. Dunkley and Smith (1951) stated that the nature of the surface of the fat globule appeared to be the most important factor in determining whether milk will develop rancidity. Rimpila and Palmer (1935) and Tarassuk and Palmer (1939) theorized that a natural surface "membrane" surrounds the surface of the fat globule. They described this layer as being composed of protein, phospholipids, and ether extractable non-phospho- lipid material. Bird, Breazeale, and Bartle (1937) concluded that two distinct protective materials are adsorbed on the fat globules. 16 One is a phospholipin-protein complex lying closer to the fat surface. The second material is located on the water side of the fat globule surface, and casein is its most important constituent. The same investigators found that the protective materials do not cover the globule surface in the form of a continuous membrane, but they are held at force centers distributed unevenly on the fat globule sur- face. According to Tarassuk and Richardson (1941), the partial denaturation and distortion of the fat globule membrane protein will leave the surface of fat globule ex- posed to the action of lipase. The same authors and Krukovsky and Sharp (1940a) found that the replacement of the original fat globule membrane by some other surface active material in raw milk will cause extensive lipolysis. Herrington (1954) stated that it appeared that at least two kind of activations must exist. The original surface material may be removed irreversibly by mechanical forces, or the adsorptive and reactive properties of the fat may be changed by the phase transformations which follow cooling. The melting_point of fa£;_ Krukovsky and Sharp (1940a) found that the rate of lipolysis depends upon the melting point of fat. As the temperature required for crystalli- zation of fat was lowered, a greater increase occurred in the fatty acids. Fat?219£99 ratio. Pfeffer, Jackson, and Weckel (1938) found that greater lipolytic activity is observed in the skim phase than in any cream phase of milk. As the fat 3 t.‘:": 17 content of the cream increased, the lipolytic activity decreased. Krukovsky and Sharp (1940a) stated that: "Total lipolytic action increases with fat content up to 35-45 percent, but acidity per unit of fat and acidity per unit of plasma increases with fat content up to 8—10 percent of fat and then remains constant or decreases." Johnson and Gould (1949a) presented experimental evidence showing that the acid degree of the fat decreases with increases in the fat content of the product. Herrington and Krukovsky (1939a) found that copper, in a small concentration as low as 0.2-0.4 p.p.m., will reduce lipolysis in the presence of oxygen. According to Krukovsky and Sharp (1940b), dissolved copper will not inactivate lipase in the absence of oxygen. Gould (1941) Presented evidence showing that copper has no effect on the extent of lipolysis induced by homogenization. Herrington (1950) believed that the destructive action of copper may be due to its oxydative ability. Inactivation of lipase by formaldehyde. Herrington and Krukovsky (1939a), Roahen and Sommer (1940), and Herrington (1950) stated that formaldehyde will lessen the extent of lipolysis in normal raw milk. However, it will not inactivate all the milk lipases, since a part of the milk lipase is not affected by formaldehyde. 18 Preventing Activation During Handling Investigators have reported means of decreasing the incidence of induced lipolysis during handling of milk. Dunkley and Kelley (1954b) suggested that by studying the different sections of a pipeline during operation, one might detect the place where activation occurs. They reported that the activating effect of risers could be decreased by eliminating them or reducing their height. They also re- ported that the rate of activation in the pipeline and risers could be decreased by increasing the flow of milk and reducing air intake to the pipeline as completely as possible. The same authors found that a higher rate of milk flow could be rsecurediby using high producing cows, several milking units, and fast milking. To avoid turbu- lence and foaming in the pipelines and risers, Dunkley and Kelley (1954b) and Thomas, Nielsen, and Olson (1955b) suggested the use of a minimum amount of air necessary to remove milk, while maintaining uniform vacuum. They further advised the elimination of all air leakage and reduction of the number of fittings and elbows to the least possible number. Very little evidence is found in literature that bulk milk tank cooling has produced activation of lipolysis. However, Herrington and Krukovsky (1939a, 1942) found that very rapid cooling of the milk will retard lipolysis during storage. These workers found that refrigeration of the bulk tank must be adequate during storage to prevent rewarming 7w 111 19 of cold milk by addition of milk during a later milking. Tests Used for Measuring Hydrolysis of Butterfat Chemical and physical tests have been used in the study of fat lipolysis. Chemical tests have been the most useful, since values have been established by which the degree of lipolysis may be measured. The most commonly used methods are organoleptic exami- nation, surface tension measurements, and measurements of fat soluble acids by different kinds of titrations (acid degree values). Herrington (1950, 1954) stated that the above mentioned methods measure different things. He warned that the results must be interpreted with caution, as they will not all yield the same answer in regard to lipolysis. Organoleptic examination. Herrington (1954) maintains that rancidity detected by organoleptic examination of milk results from short chain, water soluble, fatty acids, while long chain acids have little effect on the flavor or odor of the milk. He believes that organoleptic examination is of most significance in terms of consumer reaction, although it does not yield numerical data, and the method is of low sensitivity. Surface tension measurements. Trout, Halloran, and Gould (1935) used surface tension studies to measure the changes in raw homogenized and pasteurized homogenized milk. Tarassuk and Smith (1939, 1940) found that as the hydrolytic 20 rancidity increased, the surface tension of milk decreased gradually. They claimed that the lowering of surface tension is due to fatty acids set free by hydrolysis. Tarassuk and Henderson (1942), Dunkley (1951), and Dunkley and Kelley (1954a) used surface tension measurements to detect rancidity. They found a marked relationship between surface tension changes and rancid flavor of milk. Herrington (1954) concluded that the change in surface tension is most sensi- tive to the longer water soluble acids (capric, caprylic), while butyric acid, the most important factor in the organo- leptic measurement, has little effect on the surface tension of milk. These workers also considered the possibility that monoglycerides and diglycerides were the constituents causing reduction of surface tension, which, if so, would result in depression of surface tension nearly independent of the nature of the acid liberated. Herrington believed that surface tension measurements are difficult to interpret, because a number of factors will cause variations in surface tension values when tests are made by different investigators. Chemical measurements of fat soluble acids. Several procedures are used to measure acid degree values. Hileman and Courtney (1935) and Reder (1938) titrated the milk with 0.1 normal alkali. Gould and Trout (1936) titrated 10 gram lots of churned, melted, and filtered fat with N/lO NaOH. They expressed the results in terms of free fatty acid degrees or milliliters of normal alkali required to neutralize the free fatty acids in 100 grams of fat. They expressed the opinion that direct titration of milk fat, to measure free n‘fi‘...‘ 21 fatty acids, is a far more sensitive method for detecting lipolysis than by titration of the milk. Some years later, Herrington and Krukovsky (1939a) reported extensive experi- mental results on lipase action, using the fat titration method, expressing the titration values on the same basis as used by Gould and Trout. During the 1940's, several workers attempted to develop a standard test for measuring the extent of lipolysis. Roahen and Sommer (1940) developed a method which measured only the volatile soluble fatty acids. Hollander, Rao, and Sommer (1948) developed a method which can be applied to the estimation of free fatty acids in powdered milk, cream, and other high fat dairy products. Hillig (1947) presented a method for determining the water insoluble fatty acids. This method was later modified by Hillig (1953) and Armstrong and Harper (1954). Johnson and Gould (1949a, 1949b) developed a solvent extraction method of recovering free fatty acids from rancid milk. The same workers found that the solvent extraction method was far superior to the churning method for recovery of fatty acids, inasmuch :as the water soluble fatty acids remained in the buttermilk. Johnson and Gould (1949b) also found an improved recovery of butyric, caproic, capric, and oleic acids, when the milk was adjusted to pH 2 prior to extraction. Harper, Basset, and Gould (1954) modified the solvent extraction procedure for testing homogenized milk, and Thomas, Harper, and Gould (1954), using the modified 22 solvent extraction procedure, were able to determine minor variations in free fatty acid degree of fresh milk. Workers at Minnesota, Thomas, Nielsen, and Olson (1955) used the B. D. I. detergent test to recover fat. According to the authors, this method is rapid and adaptable for field use. Frankel and Tarassuk (1955) recently published a simpli- fied extraction titration method, wherein the mixture of solvent and fat was titrated. Harper, Schwartz, and El-Hagarawy (1956) developed a rapid silica gel method for measuring total free fatty acids in milk. The authors claim that this method is more quanti- tative in recovery of free fatty acids than any of the previom methods. Tarassuk and Frankel (1954) and Frankel and Tarassuk (1955) commented that no perfect method is yet available for recovering all the free fatty acids. The authors believe that solvent extraction methods give the highest recovery and will give the highest titration values. Herrington (1954) recognized the advantage and usefulness of solvent extraction method by stating: "Measurements of fat soluble acids have the advantage that they yield numerical results which are reproducible and which reveal very small changes in the fat." He concluded: "Consequently, the effect of minor changes in handling milk can be detected easily by this method." TH” EXPERIMENTAL PROCEDURE Description of Pipeline Installations Six pipeline milker installations from which milk was sampled and tested for free fatty acid degree in this study are identified and designated as follows: Line "A". This is found in the main dairy barn at .Michigan State University. This was the installation with continuous pyrex glass pipeline along the stanchions, a ilength of 360 feet and with aggregate riser height of 85 Linches. The risers measured consecutively as follows: 214 — 7 - 14 - 25 - 25 inches. The slope of the line was 1..5 inches per 10 lineal feet. The milk was pumped contin- Luously with a diaphragm pump. .Air was admitted to the line W1 th a control device at the far end of the line and through Claw air inlets. (See Figure 1) Line "A1". This was also found at the main dairy bilzrn, with line "A" altered to decrease the length of the line to 200 feet and aggregate riser height to 24 inches. {P&1¢, line slope was decreased 0.5 to 0.8 inch per 10 feet. The milk was pumped intermittently from a full accumulation tank. Air was admitted to the line only through the claw air valve inlets. (See Figure 2) Line "B". This was found at the Coe farm barn, with CLOntinuous PYrex glass pipeline along the stanchions, a 23 Tl“: 24 length of 190 feet, with 2 risers with total riser height of 23 inches in 8 and 15 inch lengths. The milk was pumped intermittently from a full receiver tank. Air 1N1¢t5 were similar to "A1". (See Figure 3) Line "B1". The Coe farm barn line was altered to eliminate one riser and reduce riser height to 8 inches and line slope to approximately 0.5 inch per 10 linear feet. The milk was pumped intermittently from a full receiver tank. Air inlets were similar to "A1". (See Figure 4) Line "0:; The Stran steel milking parlor at Michigan State University was used. This had a one inch diameter ‘metal line, 24 feet in length with one 34 inch riser. Single cow milkings were drawn into line in volumes. No special air inlets were provided. Line "D". This was the Clinton farm barn line, a con- tinuous pyrex glass pipeline, 210 feet in length, without risers. Slope of the line is 1 inch per 10 lineal feet. The line was without specific air inlets. Milk accumulated in a receiver tank and pumped intermittently into a bulk tank. One bulk tank milk supply, produced by bucket milking, without pipeline, was tested periodically as a matter of control. This installation is designated as tank installation "E". For the purpose of simplifying presentation of material, the term free fatty acid degree is expressed as "F.F.A.°" 25 Sampling Procedure Samples of milk were examined during an eleven month period. The samples were collected in half pint bottles, iced immediately, and held for 18-24 hours at 35-4OOF., then pasteurized at 145—1500F., holding for 30 minutes. Bulk tank samples were taken immediately after milking was completed. Pipeline samples were drawn directly from line milk valves as well as from the ends of the lines. Analytical Methods The free fatty acid degree of milk was determined by the procedure of Harper, Basset, and Gould (1954), and Thomas, Harper, and Gould (1954). This procedure is out- lined. a. Fifty-five to sixty milliliters of milk was measured into a 250 milliliter Fisher Centrifuge bottle. The sample was adjusted to room temperature and to a pH of 2 with sulphuric acid. b. The acidified sample was mixed with 65 milliliters of 95 percent ethanol and shaken vigorously for 60 seconds. After a 5 minute reaction period, 40 milliliters of ethyl ether and 60 milliliters "Skellysolve F" were added, and the mixture was shaken vigorously for 30 seconds. c. The milk and solvent mixture was centrifuged at 2000 revolutions per minute for 20 minutes in a size 2, model v, International Centrifuge, after which the ether layer was siphoned into a Mojonnier fat evaporating dish. d. The solvent was removed by heating at 135°C. on the fat evaporating hot plate of a MoJonnier tester until bubbling ceased. The last traces of solvents were removed in the solids oven of a MoJonnier tester, heating for 5 minutes at 100°C. under 20 inches vacuum. e. Exactly 1 gram fat, at room temperature, was weighed into a 50 milliliter beaker on a chainomatic balance. The fat was dissolved in 10 milliliters of purified absolute alcohol and 20 milliliters of "Skellysolve F." The solvent, previous to mixing with the fat, was slightly heated and neutralized to the phenol- phthalein end point. f. The warm butterfat-solvent mixture was titrated with 0.01 Normal absolute ethanolic KOH, using 5 drops of 1 percent alcoholic phenolphthalein indicator. The endpoint was determined according to Breaseale and Bird (1938). The first definite color change, com- pared with a sample not titrated, was taken as the endpoint. The titration value was eXpressed as free fatty acid degree, defined by Gould and Trout (1936). Preparation of Reagents Preparation of purified absolute alcohol. Absolute a1 cohol was purified using Triebold's method (1949). Ab aolute ethanol was refluxed for one hour over solid KOH Isl III". I I! I ll"! 27 and aluminum metal, (using 10 grams KOH and 6 grams alumi- num metal per 1,200 milliliters of alcohol) and distilled from these reagents. Preparation of ethanolic KOH. Purified absolute ethanol was used to prepare 0.01 Normal KOH. The solution was standardized with potassium acid phthalate and restand- ardized biweekly. "Skellysolve F". The petroleum ether, having a looiling range of 30—60°C., was obtained from the Skelly (311 Company: Kansas City, Missouri. Organoleptic Examination Organoleptic examinations were made of all samples Inhile determining the fatty acid degree. According to the lsecommendation of Nelson and Trout (1951), the milk was atijusted to body temperature, 98.6°F., to volatilize the ad ors present . Surface Tension Measurement Surface tension determinations were made with an Cenco Du Nouy tensiometer. All samples prior to testing were stored und er refrigeration temperature. The milk was tempered to 20°C. when tested, as recommended by Tarassuk and Smith (1940), Tartssuk and Henderson (1942), Dunkley (1951), Costilow and Speck (1951), and Dunkley and Kelley (1954a). Determinations were made in duplicate. The tensiometer was standardized ‘hd rechecked periodically, according to the instructions (>1? the manufacturer. 11W" 28 R-l4" ‘I‘_—" M R'ESH T «r H-25" P : {>5 R‘7" J? R-riser + P-pump and milk releaser T-bulk tank l——————* R-14" Figure l. Pipeline Installation "A" (1/2"-lo') 29 R-8" V'Jb R—l4"E ”: \V 1FR’1’4" R-riser P-pump and milk releaser T-bulk tank V—two way valve \/ /\ Figure 2. Pipeline Installation "A1" (1/2"=10') on —- U 1K \V R-riser P—pump and milk releaser T-bulk tank Figure 3. Pipeline Installation "B" (1/2"=1o') R'ISH 3O 4:< R-8" > ’1 I P Q R-riser P-pump and milk releaser T-bulk tank Figure 4. Pipeline Installation "B1" (1/2"=1o') 31 EXPERIMENTAL Effect of Installation "A" on F.F.A.° of Milk A description of milker pipeline system "A" is given on page 23 and is shown by Figure 1 (page 28). This was the original system installed in the main dairy barn of Michigan State University. Frequent occurrences of rancid flavor and odors of milk produced by the system had been observed. During the study of this line, butter particles were found regularly on the milk filter medium, indicating excessive churning action. The appearance of milk flowing through the line and over the risers was marked by abundant foaming and continual turbulent action of the milk in the risers. The result of FoFvoo determinations, made at designated points in the line, is shown in Tables I through V inclusive. Data in Table I show that when the milk passed through a 330 foot section of pipeline with an aggregate riser height of 85 inches, the F.F.A.° of milk fat, in five out of seven trials, increased by 3.00 to 4.00 F.F.A.0. Only two trials gave increases of less than 1.00 F.F.A.0. Data in Table II show that when milk passed through a 300 foot section of pipeline, with aggregate riser height of 85 inches, the F.F.A.O of milk fat, in four out of five trials, generally increased by 2.00 to 3.00 F.F.A.0. One trial shows an increase less than 2.00 F.F.A.0. 32 am. hi INFLUENCE OF LENGTH OF LINE AND HEIGHT OF RISERS ON F.F.A.° OF MILK IN PIPELINE "A" TABLE I HOLSTEIN BREED F.F.A.O of Milk 33 Trial 40- Month 30' HE 107' H 273' H 330' H Total 0" V“ 14" V 35" V 85" V Inc. 1 Sept. 1.84 2.73 3.29 5.69 3.85 2 Sept. 1.68 2.66 3.38 4.88 3.20 3 Oct. 2.02 3.79 5.38 3.35 4 Nov. 2.75 3.00 3.64 .89 5 Dec. 2.12 4.84 6.19 4.07 5 Dec. 3.54 4.05 4.29 .75 7 Dec. 2.83 5.55 6.68 3.85 1 Horizontal section of pipeline 2 Vertical section of pipeline TABLE II INFLUENCE OF LENGTH OF LINE AND HEIGHT OF RISERS .ON F.F.A.O OF MILK IN PIPELINE "A" BROWN SWISS BREED 0 Trial F.F.A. of Milk N0. Month o'fiT 243' H 300' H Total 0" v2 35" v 85” v Inc. 1 Sept. 2.91 4.24 5.10 2.19 2 Oct. 1.70 3.70 4.99 3.29 3 Nov. 2.40 4.10 5.68 3.28 4 Dec. 2.03 3.37 4.46 2.43 5 Dec. 2.03 3.05 3.68 1.65 1 Horizontal section of pipeline 2 Vertical section of pipeline . ~." {Ira-2“." ‘1 H1957 fi'fi -..,. 30*” “LI-1‘- -r— 34 Data in Table III show that when the milk passed through a 200 foot section of pipeline with an aggregate riser height of 71 inches, the F.F.A.° of milk fat in one trial was 5.3. Three other trials ranged from 0.65 to 2.71. Data in Table IV show that when the milk passed through a 138 foot section of pipeline with an aggregate riser height of 64 inches, the F.F.A.° of milk fat generally increased by 1.00 to 1.50 F.F.A.°. Data in Table V show that when the milk passed through a 57 foot section of pipeline with an aggregate riser height of 50 inches, the F.F.A.° of milk fat generally increased by 1.00 to 2.00 F.F.A.°. Data presented in Tables I through V show that the F.F.A.° of the milk fat increased with the increase of distance the milk flowed in the horizontal and vertical sections of the line. In the 27 trials shown in these tables, the acid values were the lowest at the point nearest. to the milkers, but increased with the distance traveled and with the risers encountered. In all the trials, the data show that the F.F.A.° was the highest at the discharge end of the line. The acid degree of the milk drawn from all the intermediate points practically increased continuously as the distance of flow increased. 35 TABLE III INFLUENCE OF LENGTH OF LINE AND HEIGHT OF RISERS ON F.F.A.O OF MILK IN PIPELINE "A" AYRSHIRE BREED Trial F.F.A.° of Milk N0. Month 0;.H% 163' H 220' H Total 0 V 21" V 71" V Inc. 1 Sept. 1.90 2.67 7.20 5.30 2 Oct. 2.09 3.34 4.80 2.71 3 Nov. 1.38 2.31 3.45 2.07 4 Dec. 1.72 1.72 2.37 65 1 Horizontal section of pipeline 2 Vertical section of pipeline TABLE IV INFLUENCE OF LENGTH OF LINE AND HEIGHT 0F RISERS ON F.F.A.° 0F MILK IN PIPELINE "A" GUERNSEY BREED 0 Trial F.F.A. of_Milk N0. Month <3'111 81' H 138' H Total 0" v2 14" v 64" v Inc. 1 Sept. 1.25 1.68 2.89 1.64 2 Oct. .87 1.50 2.09 1.22 3 Nov. 1.28 1.48 2.19 .91 4 Dec. 1.30 1.60 2.73 1.43 5 Dec. 1.78 1.91 3.11 1.33 1 Horizontal section of pipeline 2 Vertical section of pipeline 36 TABLE V INFLUENCE OF LENGTH OF LINE AND HEIGHT OF RISERS 0N F.F.A.° OF MILK IN PIPELINE "A" JERSEY BREED Trial F.F.A.° of Milk No. Month 8': 3% 23': I; Tgfigl 1 Sept. 1.33 3.38 2.05 2 Sept. 1.16 2.69 1.53 3 Oct. .97 1.89 .92 4 Nov. 1.36 3.55 2.19 5 Nov. 1.6# 3.34 1.70 6 Dec. 2.00 4.23 2.23 1 Horizontal section of pipeline 2 Vertical section of pipeline Effect of Line "A" on "F.F.A.°" of Milk Fat When Milk is Admitted in Volume at Distant End of Line In the previous trials, milk was drawn directly into the milk line from the cows in their respective locations. That manner of sampling permitted observation on the effect of certain distance of flow and risers encountered on the milk drawn at the subsequent locations. By this means, the amount of milk in the line was not uniform and was influenced by the production or the cows being milked. Furthermore, the method did not permit a comparison of milk from different breeds, which were located in group positions around the stable. Additional trials were therefore made to observe the influence of distance of flow and height of risers on the 1115—‘a- *o .._-'_1~-_.--.'-‘ 37 F.F.A.°, when all milk was milked into milker pails and dumped into a transfer can at the distant line end, and from where milk was drawn into the line in ten ga110n volumes. Samples for testing were drawn at the various points, as shown in Table VI. Nineteen trials, each consisting of four samples, were taken at points shown by Table VI. The F.F.A.° values show that some increase in F.F.A.° occurred with respect to hori- zontal and vertical distances of flow. The average initial value of 1.75 increased, only insignificantly however, to a value of 1.76, when the horizontal distance of flow was 62 feet and no risers were encountered. With a flow dis- tance of 302 feet horizontal and 35 inches vertical, the average F.F.A.° had mounted to 2.12, yielding an average increase of 0.37 F.F.A.°. At the end of the line, with a flow distance of 361 feet horizontal and 85 inches vertical, the samples had an average F.F.A.° of 2.36, a total increase of 0.61 F.F.A.°. This was a slight increase in F.F.A.°, when compared with the values of the milk shown in Table I, where milk was drawn directly into the line as each cow was milked. The lowest increase in F.F.A.° in Table I was 0.75, and the general increase ranged from 3.2 to 4.07 F.F.A.°. When the milk was released in ten gallon volumes at the far end of the line, the amount of milk in the line increased and flowed continuously through the entire length of the line, passing over the risers smoothly and without excessive turbulance. ‘ ‘lfl V£.\‘“|r fl. mus-n “.5: i"-»“: TABLE VI EFFECT OF DISTANCE AND RISERS 0N F.F.A.° ADMITTING 10 GALLON VOLUMES OF MILK INTO DISTANT END OF LINE "A" 38 Trial P.F.A.° of milk N°' 0' H1 62' H 302' H 361' H Total 0" v2 0" v 35" v 85" v Inc. 1 1.75 1.62 1.73 2.52 0.87 2 2.50 2 35 3.00 3.18 0.68 3 2.30 1.93 2.16 2.27 - 4 1.94 1.88 2.50 2 29 0.35 5 2.31 2 50 2.78 3.31 1.00 6 2.00 2.33 2.43 2.72 0.72 _7 2.00 2. 2 2.65 2.80 0.80 8 1.66 1.80 1.73 2.16 0.50 9 2.49 2.58 3.90 5.37 2.88 10 1.58 1.65 2 29 2.10 0.52 11 1.80 1.80 2.36 2.40 0.60 12 1.2 1.43 1.70 2.01 0. 2 13 - 1.15 1.17 1.25 1.66 0.45 14 1.07 1.09 1.60 1.39 0.32 15 1.35 1.20 1.66 1.66 0.31 16 1.28 1.42 1.56 1.69 0.31 17 1.28 1.31 1.48 1.50 0.22 18 1.40 1.38 1.68 1.72 0.32 19 1.78 1.77 1.82 1.89 0.10 .Average. 1.75 1.76 2.12 2.36 0.61 1 Horizontal sectiGH of pipeliHe 2 Vertical section of pipeline 39 Influence of Breed on F.F.A.° in Line "A" The F.F.A.O in milk from cows of different breeds and the effect of flowing through line "A", with samples drawn initially and after various distances of flow is shown in Table VII. It is noted that the highest initial average, 2.18, occurred in Holstein milk. This was followed in order of F.F.A.° average by the milk of Brown Swiss, 2.06; Ayrshire, 1.88; Jersey, 1.43; and Guernsey, 1.21. The Jersey and Guernsey milk samples were consistently and considerably lower in P.F.A.O than milk from the other breeds. After flowing through pipeline and over 85 inches of risers, susceptibility to activation was found to be greatest in the Ayrshire milk with an average F.F.A.° of 3.0, followed in order by milk from Brown Swiss, 2.78; Holstein 2.65; Guernsey 1.68; and lowest Jersey with 1.70 F.F.A.°. This influence of activation by pipeline flow is also shown by the percentage F.F.A.° increase column of Table VII. The Ayrshire milk showed an increase in F.F.A.°of 37.3 percent, Brown Swiss 25.8 percent, Guernsey 22.0 percent, Holstein 17.7 percent, and Jersey 15.0 percent. Effect of Altering Line "A" to "A1" on F.F.A.° of Milk Pipeline "A" was altered to reduce the distance of flow and height of risers. Simultaneously, the degree of line slope was decreased, and the air intake limited to IP‘fi-t-T‘. cg.._'_’ WP._—+.fim TABLE VII BREED INFLUENCE 0N F.P.A.° wHEN MILK wAS ADMITTED INTo LINE "A" IN 10 GALLON VOLUMES “— 4O Breed Month P.P.A.9 of Milk Total % I: T r. - ’- *— F.F.A.o O H 02' H 302' H 301' H Inc on v2 on v 35" V 85" V ’ Holstein Sept. 1.75 1.62 1.73 2.52 Oct. 3.50 2.35 3.00 3.18 Nov. 2.30 1.93 2.16 2.27 Average 2.18 1.96 2.29 2.65 17.7 Brown Swiss Sept. 1.94 1.88 2.50 2.29 Oct. 2.31 2.50 2.78 3.31 Nov. 2.00 2.33 2.43 2.72 Dec. 2.00 2.32 2.65 2.80 Average 2.06 2.25 2.59 2.78 25.8 Ayrshire Sept. 1.66 1.80 1.73 2.16 1 Oct. 2.49 2.58 3.90 5.37 Nov. 1.58 1.65 2.29 2.10 Dec. 1.80 1.80 2.36 2.40 Average 1.88 1.90 2.57 3.00 37.3 Guernsey Sept. - 1.2 1.43 1.70 2.01 Oct. 1.15 1.17 1.25 1.66 Nov. 1.07 1.09 1.60 1.39 Dec. 1.35 1.20 1.66 1.66 Average 1.21 1.22 1.55 1.68 22.0 Jersey Sept. 1.28 1.42 1.56 1.69 Oct. . 1.28 , 1.31 1.48 1.50 Nov. 1.40 1.38 1.68 1.72 Dec. 1.78 1.77 1.82 1.89 Average 1.43 1.47 1.63 1.70 15.0 that entering is shown by Figure 2 and under "Description of Pipeline Installations," on page 23. 1 Horizontal section of line 2 Vertical section of line from the claw air—ports. The risers were limited to The description only two per line, with an aggregate height of 22 inches. Relults on this line are shown in Table VIII. 41 TABLE VIII INFLUENCE OF LINE "A1" 0N F.F.A.° WITH VARIOUS DISTANCES 0E FLOW w —-—~~--—-—~-—-—-- —-~~-—o---—-———- h; S=Ct10n F.F.A.0 of fiilk of Trial -1 I 1 Line No. 20' H; 142' H 200' H Total 0" V 8" v 22" v Inc. North l-n 1.65 1.70 1.66 0.01 2—n 1.60 2.34 1.70 0.10 3-n 1.60 1.88 1.90 0.3 4—n 1.53 1.50 1.72 0.19 South‘ 1—s 1.13 1.57 1.42 0.29 2—s 1.28 1.70 1.52 0.24 3-s 1.50 2.40 1.80 0.30 4-s 1.92 2.23 2.88 0.96 1 Horizontal section of pipeline 2 Vertical section of pipeline The data presented in Table VIII show that only a very small increase in F.F.A.O occurred as the milk flowed in line "A1." The results show that, as the milk passed through a 200 foot section of pipeline with an aggregate riser height of 22 inches, the F.F.A.0 of milk fat, in six out of eight trials, mostly increased by 0.10 to 0.30 E.P.A.0. Trial l-n of Table VIII shows no increase in acid values. Trial 4-s shows an F.F.A.° increase from 1.92 to 2.88, or 0.96 F.F.A.O change. During that trial, some air leakage occurred in the south line section, causing foaming and turbulence in the milk. In line "A1," a definite change was observed in milk flow, along with reduction of foaming and turbulence of milk. The change greatly reduced the incidence of churning as noted by the absence of butter granules on filtering media. ‘ail;‘ rifl— m 0 . 11‘ i" “-9 q‘_.‘1"1'.“ ! 42 In line "Al," the milk appeared to Cam up, then move over the risers with a low velocity and a rhythmic flow. Effect of Milker Installation "B" on F.F.A.° of Milk The description of milker installation "B” is given on page 23 and by Figure 3. This installation was similar to installation "A1," but the milk was found extremely rancid to taste and smell. The milk was produced by a large Holstein herd. Random samples on which consumers complaints were registered gave F.F.A.° values of 3.14 to 14.89. The .milk in the pipeline foamed excessively. Air leaks were discovered to exist at Joints, due to faulty milk valves, :r-e-e-_—--—- --»-—- --- n" «1! during the course of the study. The milk line sloped 1.75 inch per 10 lineal feet and to the extent that milk flowed continually toward the riser, causing it constantly to gurgle and churn at that point. The data on F.F.A.0 values on milk sampled from various line positions is shown in Table IX. Data in trials 1 to 5 of Table IX show high F.F.A.0 values on 14 of the 16 samples tested. Only two samples, both in trial 2, were below 3.00 in F.F.A.°. One of these samples had a F.F.A.0 value of 2.88 and was drawn after a 21 foot horizontal and 8 inch vertical distance of flow. The other had a F.F.A.° value of 2.65, after flowing 95 feet horizontally and 8 inches vertically. The F.F.A.0 values in three samples, drawn from the line at the closest point to the milk inlet valve (0"v, O'H), were 3.05, 4.15, and 4.39. The initial samples drawn were consistently the 43 TABLE IX INFLUENCE OF LENGTH OF LINE AND HEIGHT OF RISERS 0N F.F.A.° OF MILK IN PIPELINES "B" AND "B1" 0 +=é~ ___ Pipe Trial F.F.A. of Milk line No. 0'31 20'H 21'H 80'H 95'H 115'H 154'H 175'H Total Olive lunv 8"V 8"V 811v 2311‘, 8"V 23"V Inc . B 1 3.05 6.40 3.35 2 2.88 2.65 4.70 1.82 3 4.39 5.55 5.51 5.89 1.50 4 3.53 4.37 4.73 1.20 5 4.15 7.70 7.47 4.19 0.04 6 3.78 3.60 - 7 2.04 2.22 2.31 0.27 8 2.38 __ 2.95 3.00 2.74 0.36 O'H 20'H 21'H 115‘H 1757H Total 0"v 8"v 8"v 8"v 8"v Inc. B1 9 2.73 2.73 - 10 3.10 2.47 - 11 2.68 3.15 0.47 lowest in F.F.A.°. 1 Horizontal section of pipeline 2 Vertical section of pipeline Subsequent tests did not show consistent increases in P.P.A.° with respect to the distance that milk flowed. trial 5, where the initial F.F.A.° value was 4.15. This condition is especially apparent in After flowing 80 feet horizontally and 8 inches vertically, the milk tested 7.70 F.F.A.°, and, after flowing 154 feet horizontally and 8 inches vertically, the result was 7.47 P.P.A.°. At the end of 175 feet horizontal distance and 23 inches vertical distance, the F.F.A.° of the milk .AI’F’QO l WW‘QT'ax.s _l.__ ___ _ 44 was only 4.19. The apparent lowering was believed caused by variation in the milk, since it was not possible to en- tirely control the milk entering the line. The cause of high F.F.A.° values in milk from installation "B" was believed to be partly caused by susceptibility of the milk to spontan- eous lipolysis. Consequently, two lots of individual cow samples were taken and tested as described in the section on "Lipolysis of Individual Cow Milk Samples," page 45. Trial 5 was made after screening out 18 cows that show- \ -. T-'_"' __am ed a tendency toward spontaneous rancidity. The F.F.A.O re- sults in trial 5 showed increase in value instead of lower- ed values as had been anticipated. Air leaks were discov— 1n--.x-. .. '3- ered in the line valves at this point in the investigation, and valves in 20 positions were replaced. At the same time, workers discovered a milk line coupling without gasket, which was also repaired. Trials 6, 7, and 8 of Table IX show the F.F.A.° of milk from indicated points in line "B" after being refitted with new valves and gasket. Nine samples were tested for F.F.A.° in trials 6 through 8. With only two exceptions, the values ranged between 2.04 and 3.00. The exceptions involved milk from the same group of cows taken from the line at O'H, O"V, and 20'H, 14"V distances, with reSpective F.F.A.0 values of 3.78 and 3.60. Effect of Altering Line "B" to "B1" After line "B" was altered by removing two vertical pieces and lowering the riser height to only 8 inches and the line slope to 0.5 inch per ten feet of line, trials 9, 10, and 11 were made. Trial 9 yielded F.F.A.° values of 2.73, at distances of both 0 and 20 feet of horizontal flow. Trial 10 gave F.F.A.0 values of 3.10 and 2.47 at the respective distances of 21 and 115 feet horizontal and 8 inch vertical, representing a loss. Trial 11 at 0 and 175 feet horizontal flow distance gave respective values of 2.68 and 3.15 F.F.A.°. The milk was observed to dam up in the line, then flow without foaming. The milk flowed quietly and without apparent turbulent action over a 8 inch riser in the line. Apparently this herd of cows pro- duced milk that was more than normally susceptible to J glis_--m_- -r___ap..s‘._,—a' , lipolysis. Lipolysis of Individual Cow Milk Samples Because of the presence of unusually high F.F.A.° in milk secured from line "B", individual cow samples were taken from milker pails, directly after each cow was milked. The samples were taken, cooled, and held, simi- larly to the usual Procedure for sampling, described on page 25. Each sample was also tested for surface tension and for rancidity by organoleptic test as described under "Experimental Procedure" on page 27. Upon completion of' the surface tension test and the organoleptic test, the samples were grouped according to flavor intensity and tests were made on pooled milk lots from these groups. The results are shown in Table X. 46 TABLE X RELATIONSHIPS BETWEEN SPONTANEOUSLY RANCID FLAVOR, SURFACE TENSION, AND F.F.A.O OF MILK FROM INDIVIDUAL COWS Ave. F.F.A.0 No. Surface Tension of Combined Flavor Samples < 46 46-48 >118 Samples dynes dynes dynes < 118 > 118 ,per cm. per cm. per cm. dynes dynes No. % No. % No. % per cm. per cm. Satis- F‘ factory 84 - - 29 59 55 100 2.62 2.32 g Rancid 22 2 100 20 41 - - 3.51 - g 1U Total 106 100 49 100 55 100 ~‘.It Eighty-four samples from individual cows were deter- mined satisfactory by organoleptic test. Of those tested, 55 gave surface tension measurements of more than 48 dynes per centimeter, 29 samples tested between 46 and 48 dynes per centimeter, and none tested less than 46 dynes per centimeter. Of 22 samples that were rancid to taste, none tested higher than 48 dynes per centimeter, 20 tested 46- 48, and 2 were below 46 dynes per centimeter. With a surface tension of 46 to 48 dynes per centimeter, 59 percent of the samples were determined satisfactory by tasting, and 41 percent were rancid. Pooled milk samples, having surface tension tests of more than 48 dynes per centimeter, had an F.F.A.° value of 2.32. The combined samples of milk, with satisfactory taste and with surface tension values of less than 48 dynes per centimeter, yielded an 1+7 F.F.A.0 of 2.62. Pooled samples of rancid milk, with a surface tension less than 48 dynes per centimeter, tested 3.51 F.F.A.0. Effect of Line "C" on F.F.A.° of Milk Milk produced in milking parlor line installation "C" was tested for F.F.A.O. The milk was produced by a herd composed of Holstein and Brown Swiss cows and was drawn into weight cans. After milking and weighing, the milk was drawn by vacuum through the line to the cooling tank. One high, 2 feet 10 inch riser, and a one inch diameter steel line, with standard fittings and without gaskets, {If .rriv‘v - ..~.- \ was used, making the line somewhat different from those used in the stanchion type line. The F.F.A.° on three trials are shown in Table XI. The data presented in Table XI show a very slight increase in F.F.A.° as the milk passed through the overhead pipeline installation "C". The data of the three trials show an increase from 1.53 to 1.74 or 0.31, from 1.83 to 2.16 or 0.33, and from 2.02 to 2.18 or 0.16 F.F.A.° respec- tively. The pipeline, under constant vacuum, drew the milk from the weight Jar to the vacuum bulk tank, in one con- tinuous flow without creating foaming or turbulence. Further investigations on installation "C" were discontinued, since all three trials failed to show considerable increase in F.F.A.°. 48 TABLE XI F.F.A.° IN MILK FROM MILKING PARLOR WEIGH JARS AND SUBSEQUENT BULK TANK SAMPLES OF LINE "C" Trial No Free Fatty Acid Degree Increase -. —*' _‘ ““ in Weigh Jar Bulk Tank F.F.A.° 1 1 ")3 1 T"; 0.21 2 1 83 2 13 0.33 3 2.02 2.18 0.16 Effect of Length of Milk Line Without Risers on F.F.A.° of Milk as Observed on Installation "D" Three milkings were observed, and samples were obtained for F.F.A.° from the "D" pipeline installation. The milk was produced by a large Holstein herd. The milk was ob- served to flow smoothly and quietly, without foaming. The results are shown in Tahle XII. The data show that the F.F.A.O of milk fat in the first trial increased from 1,84 to 1.96 or 0.12, in the second trial from 1,73 to 1.94 or 0.21, and in the third trial from 2.22 to 2.75 or 0.53 respectively. All three trials show a very slight increase in F.F.A.o as the milk passed through the 210 feet stanchion pipeline installation "D". Further investigations on installation "D" were discontinued, since the three trials did not show any considerable increase in F.F.A.°. ‘1' _‘A-I-n 5 ‘ 40, a TABLE XII F.F.A.° OF MILK FROM STANCHION BARN PIPELINE "D" N0 RISERS Trial <_n__ A_§.F.A.O 93 Milk N0. 20'sI 70's 130's 180's 210'H Total Inc. 1 1.84 1.80 1.84 1.96 0.12 2 1.73 1.90 2.00 2.20 1.94 0.21 3 2.22 1.74 2.07 1.96 2.75 0.53 1 Horizontal section of line Seasonal Differences in F.F.A.° of Milk Tests were made of milk for F.F.A.° from samples (a: 4" M.” V” - --““-- -r—‘T . a. .. ' v drawn by milking systesm "A", "C", and "E". Samples for tests were taken from well mixed bulk tanks at fairly even intervals over an eleven month period. The data secured are shown in Table XIII. It can be shown that the lowest F.F A.O values occurred in late spring and summer, from May until July, while the highest values occurred in the fall, from September until December. Table XIII also indicates that the F.F.A.° values during the winter, from January until March, were only slightly lower than in the fall. The average F.F.A.O values, in installations ”A", "C", and "E", in May through July were 2.77, 1.60, and 1.07 respectively. In September to December, the same installa- tions showed values of 3.62, 2.28, and 1.43 respectively. Pipeline installations "C" and "E" in January through march TABLE XIII SEASONAL DIFFERENCES IN F.F.A.O FROM VARIOUS MILKING INSTALLATIONS Pipe No. F.F.A.° use Seass Samslee ““ i3;}.:.22’“*13:;;?“ A May-July 12 2.77 5.95 1.58 Sept-Dec 28 3.62 5.41 2.55 A1 Jan-March 18 1.85 2.22 1.52 C May-July 8 1.60 2.16 1.21 Sept-Dec 5 2.28 2.80 1.87 Jan-March 4 2.22 2.57 2.00 E May-July 8 1.07 1.20 0.84 Sept—Dec 5 1.43 1.57 1.26 Jan-March 4 1.40 1.54 1.31 yielded F.F.A.O values of 2.22 and 1.40 respectively. Relationship between Rancid Flavor and F.F.A.O of Milk F.F.A.° determination was made on 502 milk samples in this study. The same samples also were tested organo- leptically for presence of rancid flavor. The data are shown in Table XIV. All milk of F.F.A.O of 2.0 and less, 200 samples, were satisfactory to taste and free of rancid flavor. When F.F.A.° values ranged from 2.0 to 3.0, as dfd 144 samples, 75.6 percent were found free of rancid flavor TABLE XIV RELATIONSHIP BETWEEN RANCID FLAVOR AND F.F.A.O OF MILK A; N3, F.F.A.° of Milk Flavor Samples 75-20 2.0—3.0 3.0-4.0 4.o—< No. 75 No. % No. 3% No. 5% satis- factory 310 200 100 109 75 S 5 7.1 - - Rancid 192 - - 35 84 4 65 99.9 ' 88 100 Total 50? 300 100 144 100. TO 100. 88 100 and 24.4 percent were found rancid. 0f 70 samples with F.F.A.° ranging between 3.0 and 4.0, 92.9 percent were found rancid by taste. All 88 samples, with F.?.A.O of 4.0 and over, were detected as having rancid taste. Relationship Between Rancid Flavor and Surface Tension Measurement Surface tension measurement and organoleptic exami- nation were made of 324 milk samples, using the procedure described in "EXperimental Procedure," page 27. The results are shown in Table XV. Two hundred twelve samples were determined satisfactory by organoleptic test. Of those, the majority, or 150, gave ea surface tension measurement of more than 48 dynes per centimeter. Fifty-eight samples measured between 46 and 48 dynes per centimeter, and only 4 tested less than 45