Scientists at China's Dalian Institute of Chemical Physics have tested a way to refine oil using engineered membranes instead of heat, which, under laboratory conditions, used an estimated 91% less energy than conventional distillation.

Distillation normally separates crude by boiling it repeatedly so its components vaporise and condense at different temperatures, one of the most energy-hungry steps in industry. But this new approach, which the researchers call molecular refining, pushes the oil through membranes that sort molecules by size and chemical properties instead.

The work was published in the journal National Science Review.

The detail matters for how seriously to take it. The team worked with light naphtha, one fraction of crude, not a full barrel. They ran a simulated 15-component mixture through two membranes, one sorting molecules by size and one by chemical type, directing them into three useful streams: feedstock for ethylene, ingredients for plastics, and components for petrol.

Recovery of the target materials ran at 85% to 90%. Because it works at molecular level rather than by boiling point, it is also more precise than distillation, which means less valuable product wasted into the wrong stream.

The 91% energy figure is an estimate from a lab test, not a plant, and this is one of several competing efforts. Separate teams at KAIST in South Korea and Georgia Tech published their own membrane-refining results in Nature and Science within the past few months, so the field is moving fast but nobody is running this at industrial scale yet.

The South China Morning Post reported the Dalian result and the molecular-refining method. Chemistry World notes that separating crude oil into fractions consumes roughly 1% of all the energy used on Earth, which is the scale of the prize if any of these methods reach production. Phys.org, covering the parallel work, reported a pilot membrane running stably for more than 250 hours, an early sign the approach can hold up outside a single lab run.

Refining is about 1% of global energy use, so cutting it by even a fraction is enormous, and that is the real story here rather than any single lab figure.

For South Africa the angle is specific. We reported this week that the country has lost half its refining capacity in a decade because the old plants were too small, too old and too expensive to upgrade against giant integrated refineries abroad. Membrane refining, if it scales, changes what a refinery has to be.

A process that separates oil at close to room temperature with a fraction of the energy is cheaper to run and can be built smaller, which is exactly the economics that could make local refining viable again rather than a stranded cost. But that is a long way off, with lab results not plants, competing designs and no industrial deployment yet.

However, it is the first development in a while that points at a future where a mid-sized economy could refine its own fuel without the scale penalty that closed our plants. Worth watching, not banking on.

You might also like our piece on how SA lost half its refining capacity, the Wits topological light breakthrough in communications, and how the revised electricity pricing policy is reshaping energy economics.

Get more SA tech and business news and subscribe to The Open Letter.