Helium depletion

Why does a gas essential to hospital scanners keep escaping into party balloons?

Helium is the second most abundant element in the universe. It is also the only substance that stays liquid down to absolute zero at ordinary pressure, which is why every superconducting MRI magnet in every hospital sits in a bath of it at about 4 kelvin. There is no substitute at that temperature.

And you can buy a tank of it at a supermarket to fill balloons for an afternoon. Two things that both look true — an irreplaceable medical input, and a novelty item — and the obvious explanation, that it must not really be scarce, turns out to be wrong. So what is actually going on?

Begin with where terrestrial helium comes from, because that is where the strangeness starts. It is not abundant on Earth in the way it is in stars. Almost all of it is radiogenic: alpha particles thrown off by decaying uranium and thorium in the crust are helium nuclei, and over hundreds of millions of years they seep upward and collect in the same impermeable traps that hold natural gas. Helium is therefore a mineral deposit in the strict sense — accumulated over geological time, at a rate no human process can accelerate.

Now ask about the other end of the journey. What happens to a helium atom after it has done its job? Every other industrial gas we release stays in the atmosphere, where it is at least in principle recoverable. Helium does not. It is light enough that, at the temperatures of the upper atmosphere, a fraction of atoms exceed escape velocity and leave for space, with charged-particle loss over the poles doing the rest. Its atmospheric abundance sits at around five parts per million and is not going up. So the loss is not dilution, which could be reversed at a cost. It is genuine removal from the planet.

Hold those two facts side by side — accumulation over 10^8 years, removal in something like 10^6 — and the physical picture is already complete. What remains to explain is the price, and here the reasoning has to leave chemistry.

Ask who decides how much helium is produced. Not a helium company. Helium is separated cryogenically from natural gas that is being extracted for fuel, and only where the gas happens to carry a fraction of a per cent or more of it. So the quantity available responds to the demand for methane, not to the demand for helium. If the helium price triples, essentially no new supply appears; if a single large separation plant goes offline, a global shortage follows. That is a supply curve that cannot do the job we normally rely on prices to do.

Then ask who bears the cost of a released atom. Nobody identifiable. The buyer paid for it, used it, and vented it. The loss falls on every future user of low-temperature physics, none of whom is party to the transaction and none of whom can charge for the damage — because the sink is the sky, and no one owns it. This is the shape of a collective-action problem rather than a chemistry problem: each individual decision is entirely reasonable, the aggregate is plainly not, and no participant has either the incentive or the standing to change it alone.

Policy made this sharper rather than softer. The United States built a strategic Federal Helium Reserve near Amarillo from the 1920s, and then in 1996 Congress mandated that it be sold off on a schedule to repay its debt. Selling a stockpile on a legislated timetable pushes the price toward whatever clears the schedule, not toward what the resource is worth, and for years the world's helium was cheap for that reason. Later legislation moved sales to auction, and the reserve was finally sold to a private industrial-gas firm in 2024.

Honesty requires a caveat, though, because "helium is running out" is often overstated. The shortages of the last two decades were mostly plant outages and logistics, not geology, and new capacity in Qatar, Russia and a promising find in Tanzania has repeatedly loosened the market. The problem is not that the last atom is in sight. It is that a non-renewable, non-recoverable resource has been priced as though it were neither.

The most useful correction is that this is not a story about waste in the moral sense. A balloon is a small share of demand, well behind cryogenics, semiconductor manufacture, fibre-optic drawing and leak detection. If every balloon vanished tomorrow the structural problem would remain, because the same venting happens invisibly in laboratories and factories. The balloon is a symptom that the price contains no information about irreversibility, not the cause of the shortage.

The second refinement is that "recycling" cuts both ways. Modern MRI installations increasingly use closed-cycle recondensing systems, and some designs need only a few litres sealed in for the life of the magnet rather than a periodic top-up. Recapture is technically straightforward. It is just capital-intensive, and it only gets built when the gas is expensive enough to justify it — which is precisely what the byproduct supply structure and the decades of subsidised release prevented.

Finally, note what makes helium unusual among scarce materials. Copper is finite too, but a copper atom used today is still on Earth tomorrow, sitting in a landfill at some recoverable concentration. Helium is the rare case where consumption and destruction are the same event.

• How helium-3, a genuinely different and far rarer isotope, is sourced and why fusion and quantum-computing research compete for it.
• What a severance-style royalty or a mandated capture rule on vented helium would actually change, and who would resist it.
• Why no other coolant works below about 4 kelvin, and how far dry dilution refrigerators can substitute.

Every term the collection defines is gathered in the glossary.