History of Soybean Crushing: Soy Oil and Soybean Meal - Part 4
A Special Report on The History of Soy Oil, Soybean Meal, & Modern Soy Protein Products
A Chapter from the Unpublished Manuscript, History of Soybeans and Soyfoods: 1100 B.C. to the 1980s
by William Shurtleff and Akiko AoyagiCopyright 2007 Soyinfo Center, Lafayette, California
The Interwar Period (1920-1939) . During the period of reconstruction after World War I, and especially starting in the early 1920s, Europe began to import record amounts of soy oil from Manchuria and Japan. By 1922 European oil imports had reached 72,576 tonnes, more than twice the oil equivalent of the net soybean imports. Oil imports reached a peak of 102,500 tonnes in 1927, then declined as Europe began to process more of its own oil from imported soybeans. From 1929-1933, net European imports of soy oil averaged less than 20% of the oil equivalent of net soybean imports; this figure dropped to 15% during 1934-1938. Since European soybean production had been almost negligible, European consumption of soy oil was approximately equal to net soy oil imports plus net soybean imports converted to oil equivalent (14% oil). On this basis, apparent consumption of soy oil in Europe (excluding the USSR) was about 68,000 tonnes annually from 1909-1913, almost zero during World War I, 109,000 tonnes annually from 1922-1924, reaching a peak of 259,000 tonnes from 1929-1933, then declining to 209,000 tonnes from 1934-1938, and reaching a low of about 4,500 tonnes during World War II (Burtis 1950).
The fact that Europe was dependent on imports for almost all of her vegetable oils shaped the basic nature of the oil processing industry in at least three ways: (1) All of the processing centers were concentrated in seaports or on navigable waterways, the largest being the Hamburg-Harburg area followed by Hull, Liverpool, Amsterdam, Copenhagen, and Trieste (Italy); (2) Because each country had only one or two processing centers, each of these was very large, and most also contained oil refining facilities; and (3) Because it was expensive to import oilseeds, processors went to great lengths to extract as much oil from them as possible. The combination of factors (2) and (3) led to the early popularity of solvent extraction, which requires large capacity to be economical and is more efficient.
Just before World War I, in 1913, Germany had briefly surpassed England in imports of soybeans, most of which were crushed to make 20,000 tonnes of oil. By the late 1920s Germany once again passed England to become Europe's leading soy oil producing and refining country. From 1929-1933 Germany imported twice as many soybeans for crushing as all other European nations combined. German oil mills, which were processing only 13,000 tonnes of soybeans in 1922, were crushing a remarkable 859,000 tonnes by 1928, then 1,197,504 tonnes in 1932 (the peak year; equivalent to 44 million bushels). By 1936 the figure had fallen to 569,000 tonnes. During this same period, soy oil production rose from 53,000 tonnes in 1925 to a peak of 188,000 tonnes in 1932, then plummeted to 80,000 tonnes in 1936, and to almost nothing after World War II was underway. Between 1928 and 1934 Germany was exporting large amounts of soy oil (peak of 45,000 tonnes in 1929) and soybean meal (peak of 188,000 tonnes in 1930). (Fig. ??.?).
Part of the secret of Germany's success lay in the development of high quality, cost effective solvent extraction systems for soybeans. Germany, however, did not invent solvent extraction. The process was first patented in England as early as 1856 and became established in Europe on a large scale in 1870. These early extractors were small, unagitated, single-unit batch extractors. From them, countercurrent extractors, in which the meal is moved counter to the flow of solvent, were developed (Markley 1951, p. 541). As mentioned above, England was doing solvent extraction of soybeans prior to 1909. Just prior to World War II the Germans set up a naphtha solvent plant in East Asia, but the oil recovered was considered inferior to that from the native mills (Howe 1918). Where in E. Asia?? What company?? By 1916 there were three benzene solvent mills in operation in Manchuria and Japan and several in England (Piper and Morse 1916). By 1920 solvent extraction was the favored method in England, which was then Europe's leader in solvent extraction. Thereafter, however, Germany began to take the lead. Germany's first successful solvent extraction systems were the Bollmann type (also called the percolator, Hansa-Muehle, or paternoster type) and the Hildebrandt total immersion type. Both were continuous systems. The first, patented in Germany in 1919 and 1920 by Hermann Bollmann (German patents 303,846 and 322,466 Refs??), utilized baskets mounted on an endless vertical chain, which conveyed the oilseeds through a spray of solvent; the second consisted of a gigantic vertical U-tube through which the solids were propelled counter-current to the solvent by means of perforated screws (Illust??).
Up until the early 1920s, solvent extraction was plagued by several major problems, which limited its popularity. The oil could not be used for edible purposes since the solvent (benzene or gasoline) imparted an odor, which could not be completely removed. Hence solvent extracted soy oil, limited to industrial uses, sold for 1/4 to 1/2 cent a pound less than mechanically pressed oil. The meal, too, carried the disagreeable odor of the solvent. And the solvents were highly flammable and explosive (US Tariff Commission 1920). By the mid-1920s, however, Germany had developed improved desolventizing processes using heat and a vacuum to get rid of the odor in the oil and meal. Careful engineering kept the problem of explosions under control. These advances led to a dramatic increase in German solvent extraction of soybeans and a corresponding decrease in soy oil imports from Manchuria. The solvent-extracted oil was increasingly used by the European margarine industry and the meal as a cattle feed, with some of it being ground to flour for use in breads. By 1930 nearly all European oil extraction was done using benzene (then spelled benzine) as a solvent. One factory, Hansa Mills at Hamburg, used both benzene and alcohol. Experiments using only alcohol were also conducted (Tsao 1930a; Horvath 1927). (But which company in Germany first did solvent extraction of soybeans?? Starting when?? Hansa Muehle??)
In 1934 German solvent systems started to be exported to major soybean crushers in the US. Thereafter and throughout the 1940s, German state-of-the-art technology was used in most solvent systems in the US, England, Netherlands, France, and Argentina. The German soybean crushing industry was the world leader and other countries borrowed heavily. In 1936 Durkee reported that the world's largest soybean crushing plant, located in Europe (probably the Hansa Mills in Hamburg, Germany) handled over 40,000 bushels (1,089 tonnes) of soybeans a day in a continuous process. In the early 1940s the Hansa Mills had four solvent extractors with a combined capacity of about 1,000 tonnes of soybeans a day, while the Harburger Oelwerke at Harburg operated nine Hildebrandt solvent extractors also having a combined capacity of 1,000 tonnes of soybeans a day (Goss 1944, 1946a,b).
To prevent the critical shortage of oils and fats that had developed during World War I, Germany began in 1934 to reduce her net soybean imports in a drive for great self-sufficiency in oils and fats. Shortly thereafter she began to stockpile more than 600,000 tonnes of various types of oils. Manchurian soybeans were imported from the Soviet Union and production of contracted soybeans was expanded dramatically in Romania (see Chapter 3).
Little is known about the development of the soybean crushing industries in countries other than Germany during the period between the two world wars. Yet some important research was done. In 1922 Moureu and Dufraisse (Ref??) in France demonstrated the importance and presence of antioxidants (mainly tocopherols) in vegetable oils. Hornemann (1925) showed that if the oil is removed from soybeans by pressure, all the fat-soluble vitamin remains in the cake, however if it is solvent extracted, all the fat-soluble vitamin remains in the oil.
One of the greatest economic and industrial triumphs of the soybean was in Denmark, oddly enough. Until about 1900 this country had been more than self-supporting in cereal grains, especially wheat. But US imports undercut the market, so the Danes took to raising livestock. They imported some soybeans, which they crushed for oil and meal, but also imported large quantities of meal, which they used for fodder. By 1930 Denmark had switched from a grain-based to an animal-product based economy, with 70% of the country's export trade consisting of dairy products, eggs, and meat. For their country's regained economic viability, the Danes gave thanks to the soybean ( Fortune 1930).
The War and Postwar Period (1940-1959) . During World War II, with imports of soybeans to Europe from East Asia largely cut off, soybean crushing, and hence soy oil and meal production, fell to almost zero. In Germany, although the USSR made all of its contracted shipments, no soybeans were available for crushing after 1941, when Germany attacked the USSR. The soybeans Germany had grown in Romania were used to make soy flour. Thus oil supplies in Germany again reached very low levels, as they had during World War I.
It was apparently during the war that Germany made a breakthrough discovery in improving soy oil stability and preventing off flavor development through the use of citric acid, added to prevent trace amounts of metals in the oil from causing oxidative rancidity, and through the removal of lecithin, also thought to cause off flavors (Goss 1944, 1946a,b).
Most of Germany's major oil milling centers were severely damaged during the war. In the Hamburg-Harburg area, the world's greatest oilseed center, four large mills, each processing over 1,000 tonnes of soybeans a day, were 15-80% destroyed by bombing. The Hansa Mill (Hansa Muehle), probably the most famous soybean crushing plant in the world at that time, was 80% destroyed. The mill of Noblee and Thoerl, the largest crushing and refining in the world, founded in 1883 and then handling over 1,000 tonnes of seeds a day, was 95% destroyed, having received between 2,000 and 3,000 bomb hits in the closing months of the war (Goss 1946a,b). During and after the war, American oil experts (such as Goss of the USDA's Northern Regional Research Center) studied the German oilseed industry with every means available, prompting German remarks that the Americans knew more about the industry than the Germans. In the postwar period, Germany and the rest of Europe rapidly rebuilt its soy oil processing capacity, while importing soy oil in the interim (30,000 tonnes in 1947). Germany started crushing soybeans again in 1949 (about 180,000 tonnes) and by 1959 was crushing almost 800,000 tonnes. By this time the effects of American Soybean Association promotional efforts were beginning to be felt (see Chapter 38).
The Modern Era (1960-1982) . Soybean crushing capacity and oil and meal production in Europe grew steadily throughout the 1960s, accelerating during the 1970s in response to the US agricultural exports and the rise of affluent diets in Europe. Low cost soybean meal fueled the growth of Europe's livestock and dairy industries, as consumption of meat, poultry, and dairy products reached record highs. European firms joined to form the European Community Oilseed Crushers Association (a good source of information). Soybean and other oilseed prices came to be quoted at Rotterdam (a major port). Most of Europe's soybeans (over 80%) were imported from the US, with some coming from Brazil and Argentina. In 1980 the US exported to Europe 12.6 million tonnes of soybeans, 43,000 tons?? of soy oil, and 5.5 million tons?? of soybean meal and cake.
Figure ??.? gives basic statistics on the European soybean crushing and soy oil industry in 1980. Note that although West Germany is the largest crusher and largest consumer of soy oil, the Netherlands and Denmark have the two highest per capita soy oil consumption rates. France, the second largest crusher, has the lowest per capita consumption rate; much of the oil is exported.
Total Per Caput Crushing
Population Soy Oil Soy Oil Capacity
Country (Million) Consump MT Cons (kg) Million MT Annual Per capita
W. Germany 61.4 550,000 8.96 4.30 6.8 8.8 13.1 28.5
Italy 56.8 300,000 5.28 1.68 2.1 1.1 22.7 26.4
UK 55.9 255,000 4.56 1.40 7.4 6.4 10.9 24.8
Netherlands 13.9 241,000 17.3 3.00 3.1 16.4 11.5 31.0
France 53.5 110,000 2.06 3.49 9.4 3.4 15.6 28.4
Belg-Lux 10.2 91,000 8.92 1.11 9.5 11.6 14.1 35.2
Denmark 5.1 62,000 12.1 .60 8.4 17.4 9.1 34.9
Ireland 3.2 12,000 3.75 ?? 12.1 3.9 8.3 24.3
TOTAL EEC 260.0 1,261,000 6.23 15.58 6.4 6.1 15.2 27.7
Soviet Union 2.50
Sources: Columns 1-3, Fangauf 1981; Column 4, Dunn et al. 1981. Columns 5-8, Leysen 1979, data for 1977.
Leysen (1979) in "Soy Oil Consumption Trends in the EEC," has analyzed the trends and forces behind these statistics. During the 1970s, soy oil consumption in the nine (10??) European Economic Community (EEC) countries rose rapidly , from 922,000 tonnes in 1971 to 1,546,000 tonnes in 1978, for a 68% increase in just 8 years. Per capita soy oil consumption in these countries averaged about 6 kg per year in 1978, as compared with about 18 kg in the US. Countries having high per capita margarine and moderate per capita butter consumption (such as the Netherlands and Denmark) tend to show high per capita soy oil consumption, whereas countries with high butter and low margarine consumption (such as France and Ireland) use relatively little soy oil. Italy, traditionally a liquid oil consumer, uses a fair amount of soy oil. In 1978 soy oil was the leading vegetable oil in the EEC as a whole, accounting for 35.9% of the total, followed by palm oil (14.4%), coconut oil (11.6%), and sunflower oil (10.2%). Soy oil was the leading oil in every EEC country except France (where it was peanut oil and Ireland (coconut oil).
A close look at the soy situation in West Germany yields additional information about the general situation in a number of other European countries. In 1980 Germany imported 3.9 million tonnes of soybeans, 84.7% of these from the US. These soybeans comprised 71.2% of Germany's oilseed imports, followed by sunflowerseeds and rapeseeds. Germany produced 683,000 tonnes of soy oil, which was 46.9% of the country's total vegetable oils, followed by rapeseed (24.8%) and sunflowerseed (20.3%) oils. Germany also imported 146,000 tonnes of soy oil, which was the number three oil import after palm and coconut oils. Germany consumed 550,000 tonnes of soy oil (38.9% of the total), followed by rapeseed (16.1%) and sunflowerseed (11.9%). The country also consumed 3.8 million tonnes of soybean cake and meal, which was 52% of all cakes and meals consumed (Fangauf 1981).
In Europe foreign tariffs have been designed to encourage the purchase of soybeans for crushing rather than the purchase of soy oil and meal, because the importing countries want the benefit of additional employment, value added to raw material, new plant investment, and reduced outflow of foreign exchange. In response to this trend, Americans largest soybean crushers and oil refiners (eg??) have greatly expanded their operations in Europe through construction of plants and investments.
One important development has been the work of the American Soybean Association (ASA) to increase awareness of soy oil in Germany by promoting identified soy oil. In 1960 soy oil had 20% of the German share; there were big market barriers to its expansion and no products identified on the label as containing soy oil. In 1976 (1977??) ASA started a generic soy oil campaign, showing distributors the advantage of identifying their product as containing a 100% soy oil. Consumers in Europe are more conscious than those in America of the value of replacing saturated fats by polyunsaturates. In 1976 Unilever's Blauband became the first major identified 100% soy oil brand. By 1981 there were ten identified soy oil products available in Germany, five for retail and five for wholesale. In large part because of this campaign, soy oil consumption in West Germany increased a remarkable 50% between 1977 and 1980, when it accounted for 47% of all Germany's vegetable oils. This represented a $1,000 million market for US soybean growers.
Soy oil's main future competitors in Europe appear to be rapeseed oil (especially that grown in Europe), and sunflowerseed oil. Europe's margarine industry remains the largest customer for the soy oil producer. In April 1982, in response to long and steady industry growth, a soy oil futures market opened in London.
HISTORY OF SOYBEAN CRUSHING IN THE UNITED STATES: SOY OIL AND SOYBEAN MEAL
The earliest commercial crushing of oilseeds in the US dates from 1802, when cottonseeds began to be crushed for oil. This was more than 100 years before the first commercial crushing of soybeans in the US began. The first patent on cottonseed crushing had been issued in 1796, and it was not until 1885?? that the first oilseed meal (cottonseed meal) was used in livestock feeds. During this early period the majority of America's edible oils and fats came from animals in the form of butter, lard, and edible tallow. Soy oil was a relative latecomer on the scene.
Early References and Research (1894-1909) . The earliest known reference to soy oil in the United States (and in English) appeared in 1894 in J.A. Roelofsen's "The Iodine-Absorption of Some Rarer Fatty Oils." In a long chart he listed "Chinese bean oil," the scientific name of whose seed was unknown, as being a yellow liquid with an iodine absorption (similar to today's iodine value) of 138.8. Given that soybeans and soyfoods had been written about quite extensively in the US since 1804, and that soy oil had been known in Europe since 1855, the rather late date of this first reference in America seems quite surprising. The first substantial US reference appeared in 1896 in Henry Trimble's "Recent Literature on the Soja Bean." Citing work done in Europe, he mentioned that soy oil may be extracted by use of pressure or solvents, is said to possess some laxative properties, is yellowish brown in color, has a slightly aromatic odor, and is intermediate between drying and nondrying oils. He cited a detailed list of properties given by Stingl and Morawski (1886). Except for a brief reference to imports of soybean cakes by Carson in 1909 and to early tests on the feeding value of soybean cake by Lindsey and co-workers at Massachusetts in 1909 (to be discussed in detail later), no other mention of soy oil, cake, or meal appeared in the literature until 1918 (Hen 16??), when Morse and Holmes published articles.
Introduction of Soy Oil and Cake to America (1910-1919) . Soy oil was introduced commercially to the US during the decade from 1910-1919. Although US researchers at the USDA and at state agricultural colleges and experiment stations had long been actively interested in soybeans and in promoting their cultivation by American farmers, they were concerned chiefly with the use of the soybean plant for fodder and soil improvement. There was little interest by 1910 in growing soybeans as a source of oil and meal, and almost no inkling that a paradigm shift was about to occur, in which these two products would make the soybean famous in the United States and throughout the Western world.
Starting in about 1910, the dramatic success of soy oil in Europe, including the processing of imported Manchurian soybeans to make oil and meal, inspired great interest and the beginning of soybean crushing experiments in the US. Despite the early attempts at crushing, the great majority of soy oil (an estimated 98-99%; Page et al. 1920) used during this decade was imported in the form of crude soy oil. Unlike Europe, which imported primarily whole soybeans and crushed them for oil and meal, the US chose to import crude oil, primarily because it could be imported less expensively than it could be produced from imported or domestic soybeans and because the US already had a surplus of cottonseed meal (whereas Europe had a shortage). There were not enough soybeans harvested as beans in the US during this decade to serve as a reliable source of oil, and those that were available were sold mostly at relatively high prices for seed or, during World War I, for food.
The demand for soy oil (as for all food grade and industrial oils) increased dramatically following the outbreak of the war, for various reasons: (1) America's main oilseeds, cottonseed and linseed, became even more scarce and expensive than they were before the war. The boll weevil was an increasingly important factor in declining cottonseed supplies; (2) Large quantities of oils and fats were being exported to American allies in Europe, whose supplies had been cut off; (3) Large additional supplies of oil were needed to make glycerine for nitroglycerine, used in explosives. Ten tons of oil are required to yield one ton of glycerine; in England, the price of glycerine increased fivefold shortly after the war started. One additional factor, not directly related to the war, which led to a rapid increase in use of vegetable oils, was the development of the hydrogenation process and the increased consumption of hydrogenated vegetable oil products in the form of shortenings and margarine; these served as substitutes for expensive lard and butter. Moreover soy oil was competitively priced; its average price from 1909-1912 was 6.75 cents per pound as compared with 6.5-6.75 cents for cottonseed oil and 9-13 cents for linseed oil (Thompson and Morgan 1912). Linseed oil would be even higher priced by 1916. All these factors combined to greatly stimulate imports of soy oil to the US from East Asia, using Pacific Ocean shipping routes, which were still open.
Several excellent reports published during this decade chronicled the rise of soy oil and the related crushing and refining industries in America. In 1912 Thompson and Morgan at the Delaware Agricultural Experiment Station published "Soy Bean Oil," a 13-page report, which was the first entire article about soy oil in the US. They compared the soybean with other major oilseeds (cottonseed and flax seed), and discussed imports, 13 soy oil constants, utilization, prices, and the qualities of the cake or meal. They concluded by stating prophetically " . . . it is believed that with the development of the soy bean oil industry it will be found that the beans can be profitably produced in this country for oil purposes."
In 1916 C.B. Williams, chief agronomist at the North Carolina Experiment Station, published two excellent reports, one focusing on the first crushing of American-grown soybeans in that state, and the other containing letters from various manufacturers describing how and why they were using soy oil. In 1919 Bailey and Reuter, oil chemists for the USDA, wrote "The Production and Conservation of Fats and Oils in the United States," in which they gave a good treatment of soy oil and concluded that "The soybean and the peanut constitute the two most promising possibilities for a large increase in our fat resources." But by far the most detailed and comprehensive early report was the 22-page "Survey of the American Soya-Bean Oil Industry," written in 1920 by Page and co-workers of the US Tariff Commission. The Commission's report led to the enactment of a tariff on soy oil, as described in the next section. A summary of the statistics drawn from these three reports on soy oil imports, domestic refining, exports, and utilization, is given in Figure ??.?
Soy oil was probably being imported to the US by 1909, since regular market price quotations were available from that year. In 1910 an estimated 16.4 million pounds (7,439 tonnes, worth $1,019,842) were imported, increasing to 41.1 million pounds (18,642 tonnes) in 1911 (Thompson and Morgan 1912). Imports dropped until 1914, then skyrocketed during the war, reaching the amazing peak of 336.8 million pounds (152,771 tonnes, worth $32.8 million) in 1918, but falling off rapidly after the war. Prior to 1914 Japan was the main supplier of soy oil to the US; this was actually Manchurian oil, transshipped at Kobe in Japan. England and Belgium also supplied?? substantial amounts to the US. After 1915 Manchuria became a major direct supplier, and from 1918 virtually all Manchurian oil was shipped direct from Dairen to Seattle, usually in cases of two 5-gallon tin cans, which typically lost 5% to leakage. Other major ports of entry were San Francisco and New York. Until 1919 East Asia supplied about 95% of the soy oil used in the US. Imports of defatted soybean cake during the period of 1912-1919 inclusive were very small, averaging only 2,943 tonnes annually (Page et al. 1920). Imports of soybeans from 1914 (the time of the first record) to 1920 were insignificantly small, only 3.25 million pounds (1,474 tonnes) a year.
The crude soy oil imported from East Asia was either used as is for industrial purposes or refined and deodorized in the US (to remove its distinctly beany flavor) for use in various edible products. In 1914 only 17% of the imported crude oil was being refined, but this figure jumped to 58% in 1919, showing the increasing use of soy oil for edible purposes. A small portion of the refined oil was exported; up until 1918 most of the exports were to Canada, but from 1919 most went to the Netherlands, Belgium, and England, to aid in their postwar recovery.
While we know that most crude soy oil was used for industrial purposes and most refined soy oil was used for edible products, we do not know what proportion was used for each, since statistics prior to 1931 are incomplete and there are huge discrepancies between early statistics for imports and utilization. Although industrial uses apparently greatly exceeded food uses for most of this period, the USDA (1966 Ref??) has estimated that during the war, 52% was used for nonfood products, and 48% for edible products. By far the greatest industrial use for soy oil was in soft soaps; usage increased from 1.2 million pounds (544 tonnes) in 1912 to 124 million pounds (56,246 tonnes) in 1917, at which time its use was almost equal to that of the leading oil, cottonseed oil, and it had largely replaced linseed oil. Soy oil, which was considered a semi-drying oil, was also widely used as an extender for linseed oil in paints, linoleum, and varnishes. In 1910, when linseed oil prices climbed from $0.40-$1.01 a gallon, the New York Oil, Paint, and Drug Reporter (Ref??) published the first US article on the use of soy oil in paints. Tests showed that up to 25% soy oil could be used to extend linseed oil with good results. That same year the Paint Manufacturers Association sponsored experimental plantings of soybeans (Ware 1936), and in 1917 The New York Times published an article on this work. In 1912 Toch published a long and detailed article on "Soya Bean Oil for Paint Purposes," suggesting the addition of 5-7% of a tungate drier to increase drying properties and make it "equal in every respect to linseed oil" for interior applications. As noted earlier, glycerine from soy oil was used in explosives. C.B. Williams (1916a) published excerpts of letters from 78 companies that were using soy oil in the manufacture of various products, both industrial and edible. Most praised the product and its potential. Numerous letters indicated that the largest uses were in paints, varnishes, and oilcloth. The remarkable versatility of soy oil was clearly demonstrated.
The earliest documented use of soy oil in foods in the US was in 1912, when 708,000 pounds (321 tonnes) were used in margarine. In 1914 soy oil was first used in shortenings; 1,585,000 pounds (717 tonnes) were utilized. Shortenings (also then called "lard substitutes or "lard compounds;" see Chapter 27) continued to be the main edible use for soy oil. In 1918, for example, 71% of the soy oil used in food products was used in shortenings, while 22% was used in blended salad and cooking oils (both apparently introduced that year; Piper and Morse 1923), and 7% in margarines. Although as edible or table oils, peanut and, to a lesser extent, cottonseed oils were considered superior to soy oil, improved methods of deodorizing and bleaching soy oil, and the use of quality-improving hydrogenation, gradually removed the former prejudice against its use as a table oil and gave it an improved image (Page et al. 1920). During the war, Holmes (1918) of the USDA Office of Home Economics ran tests on various oils, including soy oil, and found it to have an excellent digestibility (97.5%) by humans. Thus during World War I soy oil, as such, entered American kitchens for the first time. William Morse (1918a) wrote in the 1917 USDA Yearbook of Agriculture that soy oil "has a good color, has but a faint odor, and is rather palatable."
As mentioned earlier, the soy oil made in the US by crushing imported or domestic soybeans never amounted to more than 1-2% of total US soy oil consumption during the decade from 1910-1919. Yet it was during this period that America's soy oil crushing industry was born, and thus these developments, though small at first, came to assume great importance. Most of the soybeans crushed in the US during this period were crushed by existing cottonseed mills, and the oil was expressed using hydraulic presses or screw presses.
It is not certain when the first soybeans were crushed in the US. Weller (1970) thinks that they were probably crushed as early as 1910 by Chinese in California. He states (without citation, but probably correctly) that Oriental immigrants were then importing soybeans from China and Manchuria and crudely crushing them for cooking oil. Research is needed to verify this point.
The first documented crushing of soybeans in the US to obtain oil and meal took place in 1911 (probably not in 1910 as some Ref?? accounts say) in Seattle, Washington. The soybeans were imported from Manchuria by the Albers Brothers Milling Co. and sold to Herman Meyer, who operated a small hydraulic press mill in Seattle. His establishment was later called Pacific Oil Mills. The crude oil was sold locally for use in making soap and paint, and the meal, brand-named Proteina, was sold to farmers as a high-protein livestock fodder. It was found, however, that the oil and meal could be imported less expensively than they could be produced domestically from imported soybeans. Meyer's crushing operations were, therefore, discontinued after the initial shipment of beans had been processed (Markley and Goss 1944; Dies 1942). Yet the pattern based on the success of the European (and to a lesser extent the Manchurian) patterns of soybean processing, was established from the very outset; it has dominated soybean processing and utilization to this day.
The earliest recorded crushing of American-grown soybeans for oil and meal took place on 13 December 1915 in North Carolina at the Elizabeth City Oil and Fertilizer Company in Elizabeth City, Pasquotank County, in the far northeast corner of the state. This cottonseed oil mill, incorporated in 1912, was located along Knobbs Creek and Lamb's Ferry Road. The soybean crushing was done under the direction of William Thomas Culpepper Sr. (1884-1945), manager of the firm, as part of his efforts to promote local soybean production. At that time there was a surplus of soybeans in North Carolina, which was America's leading soybean producing state. Many farmers had planted soybeans instead of cotton, since the latter's price was often below production costs. There was also a growing importation of and interest in soy oil nationwide, and the local cottonseed mills were searching for a way to prolong their operating season.
During 13-20 December 1915, the Elizabeth City Oil and Fertilizer Company did a test run in which 272 tonnes (10,000 bushels) of soybeans were crushed and the oil expelled in the mill's six Anderson expellers. From each 1,000 lb of soybeans, the mill was able to obtain 124-135 lb (16-17.5 gallons, weighing 7.72 lb each) of crude soy oil and about 825 lb of soybean meal; the balance of 40-51 lb was processing loss. Before the tests the mill had contracted to sell all of the oil to a leading manufacturer (?? of what) at reasonable prices. Most of the resulting meal, reported to be of excellent quality and containing 5.0 to 5.5% oil, was sold to a fertilizer manufacturer. The experiment was so successful that the mill continued to crush local soybeans. Other North Carolina cottonseed oil mills followed suit within several weeks and by the spring of 1916 mills in at least nine North Carolina cities and towns (the largest were at Winterville, New Bern, and Washington) had crushed about 80,000 to 100,000 bushels (2,177 to 2,722 tonnes) of soybeans. By 1917 some 150,000 bushels (4,050 tonnes) of local soybeans had been crushed (C.B. Williams 1916b; Morse 1918a; Holmes 1918; Jordan 1918). The USDA played an important role in coordinating and studying the crushing operations. Many more soybeans would have been crushed, but for the extremely high price of the seed, which was in demand for planting and food. In 1916, for example, German interests are reported to have purchased and exported the entire local supply at prices as high as $4.50 per bushel (Markley and Goss 1944; Dies 1942).
Soon the idea of crushing locally grown soybeans spread to the southern states. By 1916 the boll weevil, which entered the US in Texas in 1892 and rapidly spread eastward, made cotton growing unprofitable in various parts of the South. Thus both soybeans and peanuts were welcomed by farmers and millers as alternative oilseed crops. In August 1916 The New York Times reported that the Louisiana Cottonseed Crushers Association had voted unanimously in favor of development of the soybean in that area for use as an oilseed, as soy oil was rapidly cutting into cottonseed oil sales. During the following months many cottonseed oil mills throughout the Cotton Belt, realizing the potential of the soybean as an oilseed, contracted with farmers for the seed of their 1917 crop; this led to a marked increased in southern soybean acreage. Soybeans imported from Manchuria were also processed in southern mills to meet the rapid growth in demand for oils (Morse 1918a, Piper and Morse 1923).
Soybeans grown in the Corn Belt were first crushed for oil and meal in late 1917 or early 1918 by the Chicago Heights Oil Manufacturing Company (located just south of Chicago, Illinois), operated by George Brett and I. Clark Bradley. Using screw presses (expellers), which were generally used for crushing corn germs (the oil-rich corn embryos), they experimentally crushed a small amount of soybeans. In late 1920, since soybeans were in short supply and most of the crop was sold for planting, Brett and Bradley bought and crushed 10 carloads of soybeans from North Carolina and Virginia. Hydraulic presses were used for soy oil extraction in 1922 and 1923. The company sold the oil with some difficulty and had great difficulty selling the meal. Bradley noted that
In the three years from 1920 we coaxed and forced feeders to try the meal. We hauled meal all over the state, gave it to them free. We sent it to experiment stations. We exhibited it at state and county fairs; we made soybean flour and sent samples to bakers, had it blended at a flour mill with wheat flour, and gave five-pound bags to hundreds of grocery stores who would consent to accept it.
Bradley and Brett continued their pioneering work toward the established of a soy oil processing industry in the Corn Belt until August 1923, when the company went out of business for lack of enough soybeans to keep the mill supplied (Dies 1942; Markley and Goss 1944).
These four early experiments with soybean crushing in Seattle, North Carolina, the Cotton Belt, and Illinois, laid the foundation for America's soybean crushing industry that would emerge during the 1920s and 1930s, and also served as a key stimulus to US soybean production.
The earliest protein concentrates used in the US livestock feeds were cottonseed meal (introduced in 1885), corn gluten feed and corn gluten meal (1888), meat scraps (1890), tankage (dried animal residues, 1900), and linseed meal (1900). The earliest known reference to defatted soybean cake in the US was in 1909, when Carson reported that 2,213 pounds (1.0 tonnes) had been imported in 1907. Also in 1909 the first results of feeding tests using soybean cake were published, as described below. Defatted soybean cake and meal were unimportant products in the US between 1907 and the early 1920s and relatively little research was done on them, whereas a great many feeding tests were done in Europe. There were various reasons for this. First, since Europe imported and crushed large amounts of soybeans, uses had to be found for the meal; America imported mostly oil and did very little crushing. Second, imports of soybean cake and meal to the US were very small. As noted earlier, during the period of 1912-19, they averaged only 2,943 tonnes a year. Third, most of the early cake imports from East Asia entered the US in Seattle and were used as livestock feed mostly in the Pacific Coast states, since feed shipped over the rockies from the Midwest was expensive there and since shipping the soybean cake to the Midwest was likewise not cost effective. However, little livestock research of any kind was done at that time by agricultural experiment stations on the West Coast. In 1911 Sawer reported that the first large shipments of soybean cake had just arrived in Seattle and predicted that it would be widely used there by dairymen and livestock feeders for the reasons given above. In 1916 Henry confirmed Sawyer's prediction, adding that the cake, though high priced, was widely fed to Pacific Coast poultry and dairy cows, and especially esteemed by dairymen, who often fed it in large amounts to cows on official tests. In the US, as in Europe, the cake, ground to a meal, was used mostly for feeding purposes, but occasionally as a fertilizer or in commercial fertilizers. Some of the meal was also used by kennel owners who found it to work very well in dog foods (Piper and Morse 1923).
The earliest known scientific report on the feeding value of soybean cake or meal in the US was published in 1909 by Lindsey and co-workers at the Massachusetts (Hatch) Agricultural Experiment Station. Their cake had had its oil expressed, presumably using an expeller or screw press, by the V.D. Anderson Co. of Cleveland, Ohio. The researchers found, as their counterparts in Europe were finding at the same time, that soybean cake fed to milch cows did not modify the composition, flavor, or yield of the milk or butter produced, although it did make the butter slightly softer than usual (how? Change composition??). There were not many tests in the early years, but they increased gradually, typically starting at agricultural experiment stations in states where soybean crushing operations were established. Thus in North Carolina, where crushing of US soybeans began in 1915, Gray (1916 Ref??) and Kaupp (1917 Ref??) did early feeding tests of soybean cake to hogs and chickens. In most tests soybean cake was compared with defatted cottonseed or linseed cake, tankage, or meat scraps, and it was often fed mixed with about 3 parts of corn.
Considerable confusion arose in the early US literature relating to defatted soybean cake and meal because of the shifting and overlapping meanings of the term "soybean meal," as explained at the beginning of this chapter under Etymology. Prior to about 1908 the term referred to a full-fat meal made by grinding whole dry (unheated) soybeans. Early feeding tests on this meal were done by Brooks (1893, 1894, 1895) on dairy cattle, Georgeson et al. (year??) on hogs, and many others thereafter. Piper and Morse (1923) contributed greatly to this confusion by using the term "soybean meal" to refer to both full-fat and defatted products and by occasionally referring to solvent-extracted meal as "cake." Nevertheless Piper and Morse gave the best early summary of feeding tests in the US and Europe using soybean cake, meal, and ground whole soybeans. They compiled the first statistics on the composition of the cake and meal in Manchuria, China, and the US. The cake contained an average of 6.80% fat (range, 5.13-8.77%), 42.6% protein (40.6-44.6%), and 11.95% moisture (7.59-17.4%). The solvent extracted meal contained an average of 1.96% fat (range, 1.33-2.20%), 43.6% protein (41.1-46.7%), and 10.95% moisture (8.98-12.87%). The corresponding figures for cottonseed cake were 7.41, 40.3, and 6.62%; for solvent extracted linseed meal they were 2.49, 37.5, and 9.63%. A history of the early work using the soybean plant (as forage, ensilage, etc.) and whole or ground soybeans as feed, is given in Chapter 2.