MDI's active chamber: how the compressed-air engine was meant to work

For a few years in the mid-2000s, a car that ran on air looked like one of the most exciting ideas in transport. The French company Motor Development International (MDI), led by the engineer Guy Nègre, showed small city cars powered by an engine that used compressed air instead of petrol. Fill the tanks at a compressor, drive off, and the only thing coming out of the exhaust is cold air.

This page used to carry an enthusiast's explanation of that engine, and encyclopedia articles still point to it. The original is long gone. What follows is a new, independent explainer written in 2026, with the benefit of knowing how the story turned out: MDI's air cars never reached volume production. That is not a judgement on anyone's sincerity. It is simply where two decades of prototypes, announcements and partnerships ended up, and the physics goes a long way to explaining why.

The basic idea: a piston engine without combustion

A conventional engine burns fuel inside a cylinder. The hot gas expands and pushes a piston down, and the crankshaft turns that push into rotation.

An air engine skips the burning. Air is squeezed into a strong tank at very high pressure, around 300 bar in MDI's published designs, roughly 300 times atmospheric pressure. To drive, the engine lets small doses of that air into a cylinder, where it expands and pushes the piston. No fuel, no flame, no exhaust gases beyond the air itself.

Stated like that, it sounds like a near-perfect city car. The difficulty is in the details of expansion.

How the active chamber was meant to work

When compressed air expands, it cools sharply. Anyone who has felt the nozzle of an emptying gas canister go icy has seen the effect. In an engine, that cooling is a problem twice over. Cold air pushes less hard than warm air at the same pressure, so each dose does less work. And the chill can freeze moisture in and around the engine.

MDI's answer was the active chamber. In the company's descriptions, incoming high-pressure air did not go straight into a cylinder and blast the piston. It was admitted first into a small chamber where it could begin expanding in a controlled way, with the engine's mechanism arranged so the piston lingered near the top of its stroke while that happened. The aim was to get the air to do its work more gradually, giving it more time to pick up heat from the engine's surroundings and less chance to waste its pressure in one sudden burst.

The company also described "dual-energy" versions. In those, a small burner could heat the air on its way into the engine. Warmer air expands with more force, so heating it lets the same tank of air drive the car further, and outside town the car could run with fuel as well. The Wikipedia article on compressed-air cars notes that heating stored air with fuel to boost output is an old trick, and that it was used in torpedoes in the early twentieth century.

Seen this way, the active chamber was an attempt to fight the central weakness of compressed air as a fuel: it carries very little energy, and some of what it carries is lost to cold every time it expands.

Why 300 bar is not as much as it sounds

The pressure in an air car's tank sounds enormous, and in mechanical terms it is. But pressure is not the same as stored energy.

The most energy you can theoretically get back from compressed air comes from letting it expand slowly at a steady temperature (engineers call this isothermal expansion). Working through the standard formula for air stored at 300 bar and released down to atmospheric pressure gives an upper limit of about 140 megajoules per cubic metre of tank, or roughly 40 watt-hours per litre. That is the perfect case, with no losses at all. A real engine, losing heat and pressure along the way, gets considerably less.

For comparison:

Even granting that a battery-electric drive is far more efficient than a petrol engine, compressed air starts a long way behind. A tank big enough to give useful range has to be large, strong and expensive, and the car built around it has to be extremely light. That is why the MDI designs were small, light city cars, and why their claimed ranges were hard to reach in independent testing.

There is also the energy needed to fill the tank. Compressing air to 300 bar heats it, and unless that heat is captured and reused, it is thrown away. Studies of grid-scale compressed-air energy storage, which faces the same physics on a bigger scale, show how much this matters: the oldest plant, at Huntorf in Germany, stores and returns energy with an efficiency well under half, and newer designs that store the heat of compression aim for around 70 per cent.

A sudden breakthrough that never quite arrived

Part of the appeal of the air car was the sense that a small team had found a shortcut that big carmakers had missed, and that the change could come suddenly. The record shows a slower, more ordinary story.

• 2002 to 2004: MDI's car made public appearances, and production was predicted to be months away.
• February 2007: Tata Motors, the Indian carmaker, signed an agreement with MDI covering further development of the technology and the right to use it in India. This was the high-water mark of public interest.
• 2009: a Tata engineering executive was reported as saying low engine temperatures were causing problems, the cooling issue described above.
• 2012: Tata said the engines had completed a first phase proving the technical concept and had moved to a second phase.
• 2017: Tata said the first vehicles could be available by 2020. They were not.
• 2020s: with no air cars on sale, mentions of the project faded. Wikipedia's article on the company reports that two French companies linked to it were placed in compulsory liquidation in 2023, and that the Luxembourg holding company was declared bankrupt in January 2026.

Several other businesses around the world took rights to build or sell the cars, and in at least some cases they waited for technology or vehicles that did not arrive. Guy Nègre died in 2016.

None of this means the engineering was a fraud or that the people involved were not serious. It means the compressed-air car faced a hard limit on stored energy, a real problem with cold, and then a fast-moving competitor: lithium-ion batteries, which kept getting cheaper and better through exactly the years the air car was stalled.

What compressed air is good for

Compressed air did not disappear from the energy conversation. It is useful wherever its weaknesses matter less:

• Grid-scale storage, where air is pumped into underground salt caverns and the heat of compression can be stored and reused. A small number of plants operate around the world, and new ones have opened in China in recent years.
• Tools and industry, where air power is valued for safety and simplicity rather than efficiency.
• Hybrid assistance, briefly. In 2013 Peugeot and Citroën announced a "Hybrid Air" system using compressed air to help a petrol engine in town, but the development was indefinitely delayed in 2015.

The lesson for the next big idea

The air car is a useful test case for any emerging technology. The questions worth asking are the dull ones: how much energy does it really store, what does it cost to make, what happens in cold weather, and who has actually built one in numbers? The technologies that changed New Zealand driving over the past twenty years, battery electric cars charged at home, did so by passing those tests rather than by sounding revolutionary.

For the rest of the story, see our look back at the air-car years from 2007 to 2016, and our piece on energy storage of every kind, including compressed air.