Solvent Extraction

The objective is complete recovery, or as near to it as the process can practically get. Where a press leaves oil behind in the cake, extraction is able to reach it, because dissolution does not depend on the oil having a mechanical path out. The solvent penetrates the prepared solid, oil dissolves into it, and the resulting solution is drained away from the depleted solids. Two separations then follow: the oil is separated from the solvent, and the solvent is separated from the solids.

That last point is what makes the process an engineering undertaking rather than a simple washing step. A solvent that dissolves oil well and separates from it cleanly is, by the same physical properties, volatile and flammable. Everything difficult about the process follows from the need to keep that fluid contained, recover essentially all of it from both product streams, and hold it away from any ignition source throughout. The oil recovery is the easy part; the solvent management is the process.

The two methods are not competitors offering the same service at different prices. They do different physical jobs, and the industry uses them in the combinations that suit the seed. Pressing removes the oil that will flow. Extraction removes the oil that will not. Where a seed carries a lot of oil, running both makes sense: a prepress stage takes out the bulk mechanically and extraction finishes the recovery on the cake. Where a seed carries less oil, as soybeans do, there is little for a press to express and direct extraction of the prepared seed is the normal route.

The consequence is that the same commodity can be crushed by more than one route, and the routes leave different amounts of oil in the meal. That difference is real and it is why extraction was developed, but its magnitude is a plant- and seed-specific matter and no figures are given here. What can be said qualitatively is the direction: extraction recovers more oil from the same seed than pressing alone, and correspondingly leaves a meal lower in residual oil.

Soybean crushing produces oil and meal, and both are co-products in the full sense. This is not a case of a valuable product and a residue. The meal is the larger stream by mass and carries much of the crush’s value; the oil is the more valuable per unit. A soybean crushing plant is built to sell both, and the relative price of the two is what drives crush margins. Neither is the primary product and neither is a by-product of making the other.

Soybean oil

The extracted oil, recovered from the solvent solution as crude oil. A co-product, going on to degumming and further refining before food use.

Soybean meal

The depleted solids, once residual solvent has been removed and protein quality adjusted. A co-product and one of the world’s principal protein feed ingredients, sold on protein.

Neither stream leaves the extractor finished. The oil requires refining; the solids require desolventising before they are meal at all. Those are separate methods, described separately, and they are not optional steps a plant might skip.

The process is thermally demanding, because separating solvent from both product streams and recovering it for reuse requires energy. Its emissions profile is dominated by solvent losses to air, which is why solvent recovery is both an economic priority and a regulated one — the same engineering serves both ends. Its material streams are fully utilised: oil and meal both have established markets.

Residual solvent in food and feed products is separately regulated, with limits and analytical requirements set by the applicable food and feed authorities. Those limits are jurisdictional and are not reproduced here.