K.T. Achaya
Dr Achaya participated in the 1993 FAG/WHO Expert Consultation on Fats and Oils. His areas of expertise include dairy science, food and nutrition and the history of Indian foods.
Traditional ghani technology
Crushing oilseeds
Oil yield
Recent evolution of oil processing
Advantages and disadvantages of ghani crushing
Bibliography
For thousands of years, fats and oils have been important in food preparation in India. Metal frying pans that are remarkably similar in design to those used today have been found in archaeological excavations of the Harappan civilization of circa 2000 BC. A number of oleaginous materials such as sesame, rape and mustard seeds and coconut were known sources of oil (Achaya, 1990). In addition, a variety of animal fats were used. However, the exact way that oil was obtained from oilseeds is uncertain.
In Sanskrit literature of about 500 BC there is a specific reference to an oil-press, although it was never described (Monier-Williams, 1899). Juices were extracted from vegetable materials as early as 1500 BC using either a mortar and pestle or a grinding stone working on a flat stone. Linguistic evidence suggests that it is from these two crushing systems that presses for both oilseeds and sugar cane developed in the form of a mortar-and-pestle arrangement powered by animals. This system Is commonly called the ghani, or the kolhu or chekku (Achaya, 1993).
Ghani operation has been noted in Afghanistan, Sri Lanka and Myanmar, which had cultural ties with India. The device is widely used in the Sudan to crush sesame seeds, though its antiquity there has not been documented. Despite subsidies by the Indian Government to support village industries, in recent times ghani operation has declined precipitously in the land of its origin.
The oilseeds and subsequently the expressed oil are held in a scooped circular pit in the exact centre of a circular mortar made of stone or wood. In it works a stout, upright pestle which descends from a top curved or angled piece, in which the pestle rests in a scooped-out hollow that permits the pestle to rotate, eased by some soapy or oily lubricant. Today the single angled piece takes the form of two shorter pieces pinioned or chained together. The bottom of the lower angled piece is attached to a load-beam; one end of the load-beam rides around the outside of the barrel, while the other is yoked to the animal. The load-beam is weighted down with either heavy stones or even the seated operator. As the animal moves in a circular ambit, the pestle rotates, exerting lateral pressure on the upper chest of the pit, first pulverizing the oilseed and then crushing out its oil.
Within India there are regional variations in ghani design (Patel, 1943;Chaudhuri and Selvaraj, 1985), which probably arose from the nature of the oilseeds that were regionally available for crushing. The large granite ghanis of southern India have a capacity of 35 to 40 kg, requiring two animals yoked side by side and two operators, one for the animals and the other near the mortar. The load-beam is very long and curved and rides on a strong outer groove on the mortar. These ghanis have a life of four to five years, after which the pit is too worn to be useful. The wooden ghani of western India has a capacity of 8 to 15 kg, has an oil outlet at the base of the pit (which is kept plugged during crushing) and frequently has the operator seated on the load-beam. The Bengal ghani has a small capacity of only 5 to 10 kg per charge and is usually used to crush a mixture of rape and mustard seeds to yield a bouquet of flavours. The pit is small and the pestle is tall and has a stout base. The operation is prolonged so as to permit slow enzymatic liberation of several pungent alkyl isothiocyanates from the precursor glucosinolates in the prevailing warm, moist conditions. Punjab ghanis are of similarly small capacity but generally carry a short pestle.
For the mortar, the trunks of hard woods such as the tamarind tree (Tamarindus indica), neem (Azadirachta indica), jack (Artocarpus heterophyllus), baheda (Terminalia bellirica), shirish (Albizia lebbeck) and sal (Shorea robusta) have been utilized regionally, all these being very large trees (Patel, 1943). Pit designs also vary with region, and could even take the form of a wooden sleeve that sits snugly in the cavity and is less expensive to replace (Patel, 1958). Even wooden strips laid radially in the pit cavity are in use. The pestle is generally made of baheda, shirish or babul (Acacia nilotica) wood, with a bulbous tip sometimes clad with lengthwise metal strips. The shape and design of the pestle end must match that of the pit to avoid excessive dead space (Nag, 1982). The curved or angled piece was once fashioned out of a single large piece of wood; a shortage of these has led to the use of two pieces, the top one angled or curved and the bottom one straight; these are tied, chained or pinioned together for easy detachment when the pestle has to be removed. The strong load-beam has to be designed so that lateral pressure on the animal does not force it to lean from a comfortable upright stance during ambulation. Trained male animals, cattle or buffalo, are generally used, usually blindfolded to avoid dizziness and distraction; however, on small ghanis in certain areas even human labour, both of men and women, is employed (Achaya, 1993).
In the crushing of 10 kg of sesame seed in a ghani, about three-fourths of the material is placed in the pit and the rest is evenly laid out all around the flat rim (Patel, 1943; Nag, 1982). The animal is prodded and allowed to perambulate for a few minutes until pulverized seed is found to climb the walls of the pit. The animal is halted, and 180 ml of water is sprinkled around the chest and 120 ml poured into the pit. A further 5 minutes of pestle rotation will cause about three-fourths of the seed to be pulverized, after which another 300 ml of water is poured evenly around the pithead. The material built up in the chest is raked using a crowbar, and the pieces are broken up by hand and cast into the pit. After the animal has resumed movement, the rest of the seed is evenly pushed in all around. The operator now tests the solid material by balling it in his or her palm; if it crumbles too easily, more water is needed. The layer of built-up material is again broken up, and brisk ambulation is resumed. After about 45 minutes, a sudden release of frothy oil floods the surface. Another 100 ml of water is sprinkled over the oil, the animal is stopped and the oil is allowed to settle. A final quantity of about 20 ml of water is now brushed over the compacted cake surface using the edge of the palm, after which the animal makes a few more rounds. The operation is stopped, the two curved pieces are detached and the pestle is lifted out and laid aside. If the ghani has a drainpipe, it is unplugged and the oil is drawn into a vessel. Otherwise the oil released into the pit is mopped up with a piece of cloth and wrung out by hand into vessels. While the cake is still hot and before it has set really hard, it is prised out as thick slabs from the chest using a crowbar.
Rape and mustard seeds need more water during crushing than sesame, and copra rather less. The oilcake is not raked during linseed crushing, but only at the very end. Safflower seeds are always very carefully decorticated by passage between grinding stones, sieving and winnowing: only practically pure meats are subjected to ghani crushing. During crushing of groundnuts at least part of the shells are retained in the ghani so as to ensure formation of a granular and compact cake.
At the point of maximum contact, the pressure in a ghani is about 10 kg/cm2 (Gujarathi, 1982), about one-third of that in a small screw-press and about one-tenth to one-hundredth of that in a large modern expeller. The pressure in the ghani is largely determined by the weight placed on the load-beam, usually 115 to 160 kg, which is transmitted by way of the curved piece to the top of the pestle.
The fit of the pestle within the pit is important. Experiments have shown that an inclination that exceeds 21° from the perpendicular causes so much lateral pressure that the mass will not climb the walls of the pit. Too much dead space in the pit will have the same effect.
The phased additions of about 7,5 percent water during ghani operation have a major role. The first addition provides the pestle with a grip on the dry oilseed, and the friction produces heat. The second portion, with the heat present, cooks the ground seed. This is analogous with what happens in the stack precooker in modern screw-press operation. Protein is denatured and coagulated, and as the moisture level reaches a critical point, oil is rather suddenly displaced from the cells. (In the Russian Skipin process, in which oil is extracted by displacement with hot water, this critical moisture level has been ascertained to lie within the limits of 14 to 18 percent (Alderks, 1948)). The cake at this stage turns granular and cohesive, and will not reabsorb the expelled oil. After the oil has appeared, the third addition of water serves to hydrate and coagulate gums and phospholipids. This phase is analogous to modern oil degumming. The last brushing with water serves to clear surface oil on the cake and give it a sheen.
An oil-rich seed such as sesame seed or groundnut yields about 5 percent less oil in a ghani than in a modern expeller, mainly because of insufficient pressure. Ghani oilcake carries about 15 percent residual fat, about twice that of screw-press oilcake (Achaya, 1993). In fact, in modern commercial Indian practice oilcake produced by crushing rape and mustard seeds in the ghani is put through screw-presses to obtain about 2 percent oil; this is added to the pungent ghani oil already obtained, to raise total yield.
At the beginning of the twentieth century, there were an estimated 500 000 ghanis in India which crushed 97 percent of all oilseeds (Achaya, 1990). The remaining 3 percent, mostly rape and mustard seeds, coconut and groundnuts, were processed in rotary units which were no more than mechanized ghanis installed in factories. Virtually all the oil used in the country, about 800 000 tonnes, was processed by ghanis. These numbers dropped sharply as power-driven screw-presses, hydraulic presses and solvent-extraction units came into operation. A further switch to rotaries for rape and mustard seeds and coconut, and to expellers for groundnuts and castor beans, rapidly brought down ghani usage for all oilseeds to 40 percent in the 1930s and 28 percent in the 1940s (Achaya, 1990).
In the 1930s, Mahatma Gandhi inspired the formation of an organization devoted to the resuscitation of village industries, including oilseed crushing, which is now called the Khadi and Village Industries Commission (KVIC). This organization has steadily evolved a succession of improved oilseed crushing devices. The first of these, the Wardha ghani, was still worked by animals, but it was followed by several units that employed electric power for traction: the rotating barrel power ghani, the overhead power ghani, the portable power ghani and the portable overhead power ghani. Assistance by way of loans for both capital and working expenditure is granted by the KVIC both to individuals and to cooperative organizations to modernize traditional ghanis with power-driven pestles or to install the new power ghanis.
By the mid-1950s the number of ghanis had fallen to 300 000, and in 1983, despite the efforts of the KVIC, the figure was placed at 100 000 to 150 000 (Achaya, 1990). About 60 000 units were beneficiaries of KVIC assistance, of which 37 percent were traditional ghanis, 50 percent earlier forms of the power ghani and 13 percent later designs.
At present, just 4 percent of all oilseeds are pressed in ghanis. The proportions of individual oilseeds crushed in ghanis have been estimated as follows: safflower seeds, 40 percent; sesame seeds, 24 percent; rape and mustard seeds, 6 percent; groundnuts, 2.5 percent; and copra, 1 percent (Achaya, 1993). The major drawback is the low production capacity of ghanis. Even modern ghani units can press 100 kg of oilseeds per working day at best. As long as the present subsidies, such as exemption of sales tax on ghani oil, are retained as a measure of commitment to village industries, ghani operations can be expected to survive.
When ghani crushing was widespread, fresh oil was in greater demand than it is today. Flavour, which was traditionally an important attribute of all oils, and particularly of rape and mustard, coconut, groundnut and safflower oils, was best in oils produced from mild ghani crushing. Both storage quality and nutritive value were perceived as being high, although this is not borne out by modern studies. Today homemakers, especially in urban areas, demand bland and refined packaged oils.
Since vegetable oils are naturally sterile, problems of hygiene in ghani oil are unlikely. Turnover of oil in the home is so rapid, and usage of oil in India so low, that oxidative and lipolytic deterioration resulting from storage is also insignificant. Ghani cake is known to be exceptionally hard and dry and is not prone to mould infestation unless inadvertently wetted.
However, the ghani has disadvantages which are mainly economic in nature. Traditional ghanis have a maximum capacity of about 50 kg per day, and modern powered units only about twice that much. As a result, running costs are disproportionately high. If animals are used, they need to be trained, and they are expensive to feed. Artisan training is also essential. Ghani oilcake as prised out of the unit after crushing is extremely hard and is not accepted by the trade for further solvent extraction, as are expeller oilcakes.
In ancient times, ghani crushers in India were recorded as being a separate caste, and this distinction still persists (Bose, 1975). Since the start of the twentieth century, as the demand for ghani oil has dropped sharply, younger people have shifted to more remunerative occupations, turning away from ghani crushing as they have from many other artisanal activities in a rapidly changing social, technological and economic environment.
Use of ghani crushing in India has probably stabilized at the current level of subsidized operations. In the future, power-driven devices are certain to displace traditional ghanis worked by animal traction. There may still be room for powered ghanis in India and perhaps even in other developing countries with limited local supplies of raw materials for oilseed extraction, and there may be a place for batteries of power ghanis to multiply oil output from a common shaft in factory operations.
Achaya, K.T. 1990. Oilseeds and oilmilling in India: a cultural and historical survey. New Delhi, India, Oxford and IBH.
Achaya, K.T. 1993. Ghani: The traditional oilmill of India. Kemblesville, Pennsylvania, USA, Olearius Editions.
Alderks, O.K. 1948. Cooking of meats and recovery of the oil. In A.E. Bailey, ed. Cottonseed and cottonseed products: their chemistry and chemical technology, p. 615-650. New York, Interscience Publishers.
Bose, N.K. 1975. The structure of Hindu society, trans. André Beteille. Hyderabad, India, Orient Longman.
Chaudhuri, J.L. & Selvaraj, K. 1985. Technological developments in ghani oil industry. Khadigramodyog, Oct./Nov.: 75-84.
Gujarathi, P.V. 1982. Design and efficiency of power ghani system. In V.H. Potty, S. Hariprasad & M.R. Mukunda, eds. Village oil industry, p. 91-100. Mysore, India. Central Food Technological Research Institute.
Monier-Williams, M. 1899. A Sanskrit-English dictionary, Delhi, India, Motilal Banarsidass. Reprinted 1963.
Nag, T.K. 1982. Village oil industry. Power-ghani installation: a handbook. Bombay, India, Khadi and Village Industries Commission.
Patel, J.P. 1943. Oil extraction. Maganwadi, Wardha, India, All-India Village Industries Association. 3rd ed.
Patel, J.P. 1958. Village oil industry. Bombay, India, Khadi and Village Industries Commission.
