Category: Space

  • Helium – Don’t Get Carried Away

    How humanity undervalued one of its most important resources


    Helium escaping an oil rig, centred between a space rocket and a microchip


    Balloon Gas

    Walk into any party shop and helium appears to be an ordinary commodity. It fills foil balloons, makes voices squeak and eventually drifts harmlessly into the sky. The cost seems trivial, the supply endless. Few people give the gas another thought.

    That impression is profoundly misleading. Unlike industrially harvested gases such as methane and carbon dioxide, being an element it cannot be produced artificially – sort of minute quantities synthesised in a nuclear laboratory at vast expense. Also, unlike most other gases we take for granted it, is not readily available in the atmosphere in any usable quantity. Plastics, steel or glass can be manufactured by rearranging existing materials, helium cannot. Creating helium requires nuclear reactions. On Earth, these reactions occur naturally deep within the planet’s crust. Elsewhere in the universe, helium is forged inside stars by nuclear fusion, the same process that powers the Sun.

    In actuality, helium and hydrogen make up most of the ordinary matter in the universe, particularly in stars and gas giants. Down here on Earth, however, helium does not linger. Once free, it makes off as quickly as possible – unless trapped by geology.

    Virtually every single atom of helium we use was created either billions of years ago during the formation of the universe or by geological processes that operate at an unimaginably slow pace. We continue to consume that inheritance in mere decades. Yet, to replace it nature requires geological time measured in millions or billions of years.

    And once released into the atmosphere, helium rises up and keeps on rising. Unlike oxygen, nitrogen or carbon dioxide, it forms virtually no chemical compounds that might trap it within the Earth’s ecosystem. Helium atoms eventually gain enough velocity to escape our gravity altogether, with the Earth losing around 50g of helium every second.

    Once they leave, they never return – they’re gone forever.

    That alone should give us pause. Instead, for decades, we have treated helium as though it were little different from compressed air. The balloon is not the problem. It is the symptom: our civilisation has mistaken a finite strategic resource for an inexpensive everyday commodity. That misunderstanding may prove to be one of the most expensive misconceptions of the twenty-first century.

    The mistake was understandable. For over a hundred years the true economics of helium were hidden by an accident of geology and industry. The market price reflected the cost of recovering helium alongside natural gas, rather than the intrinsic value of the element itself. We came to believe helium was abundant because it was inexpensive.

    Those are not the same thing, and the difference is now being tested from both directions at once. Artificial intelligence has triggered an unprecedented boom in semiconductor manufacturing. Commercial spaceflight has turned rocket launches from occasional national prestige projects into routine industrial operations. Quantum computing is moving from the laboratory towards commercial systems. Fusion energy programmes are expanding across the world. Modern medicine depends on superconducting MRI scanners, while fibre-optic manufacturing, scientific research and countless industrial processes consume helium every day. Meanwhile, supply has become concentrated in fewer hands, strategic reserves have largely disappeared, and new sources need years of investment before they can reach the market.

    This is not an argument against balloons, nor a claim that the world is about to run out of helium tomorrow. It is an argument that we have fundamentally misunderstood the nature of one of our most important enabling resources.

    To understand why, we first need to answer a deceptively simple question.

    If helium is so valuable, where does it actually come from?


    The Gas That Nature Refuses To Keep

    If helium is so difficult to replace, why don’t we simply make more?

    That sounds like a reasonable question. Modern industry manufactures millions of tonnes of chemicals every year. We transform crude oil into plastics, refine aluminium from bauxite, extract silicon from sand and synthesise fertilisers from the nitrogen in the air. Industry has become extremely adept at rearranging the building blocks of nature.

    Helium stubbornly refuses to cooperate, and the distinction is fundamental. Chemistry rearranges atoms. Helium itself is an atom, a chemical element. Creating one means changing the nucleus of another element through a nuclear reaction. It can be done, but only at vast cost, by expending more energy than the resulting helium could ever justify economically. Laboratories can produce helium. Industry cannot.

    The helium economy therefore begins with a constraint unlike almost any other raw material. When demand rises, we cannot build another factory. Every atom we consume must already exist.

    So where does the Earth’s supply come from? The answer lies beneath our feet. Within the Earth’s crust radioactive atoms of uranium and thorium have been decaying for billions of years. Among the particles they emit are alpha particles: bundles of two protons and two neutrons. As they slow, they capture electrons from their surroundings and become ordinary helium atoms. Nature has been manufacturing helium almost since the Earth formed.

    The problem is scale. Radioactive decay is glacially slow. The process that created today’s reserves has been operating for billions of years, yet the amount generated each year is tiny compared with modern industrial consumption.

    If production were the only problem, matters would be difficult enough. Nature presents another: helium is almost impossible to keep.

    Unlike hydrogen, oxygen or carbon, helium is chemically indifferent. It forms virtually no stable compounds, binds to almost nothing and appears in almost no minerals that could later be mined and refined. Hydrogen gets trapped inside water, hydrocarbons and living tissue. Carbon forms rocks, oil, coal and the molecules of life itself. Helium remains resolutely alone.

    The irony is that the characteristic that makes helium so hard for geology to retain is exactly what makes it valuable to engineers. It will not burn. It will not corrode. It contaminates almost nothing. It remains a gas at temperatures where nearly everything else has condensed into a liquid or frozen into a solid. Its atoms are so small they slip through microscopic imperfections, which makes it the world’s preferred gas for leak detection, and its thermal conductivity lets it carry heat efficiently while remaining completely inert. The qualities that make helium indispensable to advanced technology are the same ones that make it so difficult for the Earth to hold.

    A newly formed helium atom begins a slow climb through microscopic cracks in the surrounding rock. Some atoms dissolve briefly into groundwater before continuing upwards. Most eventually reach the atmosphere, where helium accounts for only around five parts per million of the air we breathe. Recovering it from the air is technically possible and spectacularly uneconomic: it means processing enormous volumes of atmosphere to recover almost nothing.

    Given enough time, many atoms keep climbing until they leave the atmosphere altogether. Earth’s gravity is too weak to hold them indefinitely.

    Which raises an obvious question. If helium has been escaping for billions of years, why is there any left?

    Because geology occasionally performs a remarkable trick. Most rising helium drifts inexorably towards the atmosphere, but a fortunate fraction encounters porous layers of rock capped by impermeable layers of salt or shale. Unable to continue their ascent, the atoms accumulate beneath these natural seals. Over millions of years, what began as isolated atoms becomes underground reservoirs large enough for industry to exploit.

    Nature has not created helium mines. It has created helium traps.

    The curious thing is that these traps usually contain something else as well: natural gas.

    At first glance the relationship appears obvious. Helium is extracted from natural gas, so natural gas must somehow produce helium. It does not. The two merely arrive at the same destination by entirely different routes. Methane originates from ancient biological material transformed by heat, pressure and time. Helium originates from radioactive decay in the surrounding rocks. Their histories are unrelated. They accumulate within the same geological structures, because those structures trap gases regardless of where they came from.

    That coincidence has shaped the entire helium industry. It has also concealed one of the most important economic truths in this story. For more than a century, the world behaved as though helium were a by-product of natural gas. In reality, natural gas was paying most of helium’s bills.

    The distinction sounds academic. It changes almost everything.


    The Accident Beneath Our Feet

    If helium and natural gas merely share the same underground reservoirs, why has the helium industry become almost entirely dependent on the gas industry?

    The answer is mundane: economics.

    Imagine drilling a well solely to recover helium. You must locate a suitable reservoir, purchase drilling rights, sink the well, build pipelines, install compressors, construct a purification plant and somehow transport the finished product to customers. Every pound spent must be recovered through helium sales alone.

    Now imagine discovering helium while drilling for natural gas. The expensive part of the project has already been justified. The well exists, the pipelines are being laid, the processing plant is needed regardless, and the workforce is in place. Investors expect to recover their money by selling methane. From helium’s perspective, somebody else has paid the entrance fee.

    That distinction explains almost every paradox in the modern helium market. Helium is not inexpensive because it is abundant. It is inexpensive because much of the infrastructure needed to produce it already exists for an entirely different reason. For more than a century, the natural gas industry quietly subsidised helium production. Few companies set out to build a helium business. They discovered that recovering helium made sense once the expensive work of finding, drilling and processing a gas field had already been paid for.

    This also explains why production is concentrated in so few places. Not every gas field contains useful quantities of helium. Many contain almost none. Others contain enough to justify recovery only if the field itself is commercially attractive for natural gas or LNG. A reservoir rich in helium but isolated from pipelines, processing plants and customers may sit untouched for decades because the numbers do not add up.

    The industry therefore developed opportunistically rather than strategically. Whenever a sufficiently large gas project happened to contain enough helium, another piece of the global supply network appeared. The pattern is visible almost everywhere helium is produced. The United States became the world’s largest supplier because several fields beneath Texas, Kansas and Oklahoma happened to contain unusually high concentrations. Qatar became a helium superpower for the opposite reason: gas from the North Field contains only a tiny percentage of helium, but the country’s LNG industry processes such immense volumes that even a trace concentration becomes commercially significant. Russia’s Amur Gas Processing Plant follows the same logic. In each case the economics are justified by natural gas; helium is one of several valuable products recovered along the way.

    At first glance this sounds reassuring. If helium accompanies natural gas, surely there is plenty more waiting to be recovered.

    Unfortunately, the relationship works both ways. If helium production depends on natural gas production, then decisions made for the benefit of the gas industry determine the world’s helium supply. A decline in gas production can reduce helium output even while helium demand rises. An LNG plant closing for maintenance can temporarily remove a significant fraction of global supply despite there being no shortage of helium beneath the ground. War, sanctions, industrial accidents and disrupted shipping have the same effect. None of them changes the amount of helium trapped beneath the Earth’s surface by a single atom. All they interrupt is the machinery that separates, purifies, liquefies and transports it.

    None of this is hypothetical. The industry has lived through four distinct shortages since 2006, complete with rationing, allocation contracts and idled instruments. When Qatar’s neighbours blockaded it in 2017, helium shipments stopped almost overnight: nothing had changed underground, but the containers could no longer cross the Saudi border. In 2022, fires at the Amur plant and a leak in the ageing infrastructure of the American reserve system removed a large share of world supply at a stroke. Each time, buyers discovered how few suppliers stood between them and an empty cylinder.

    The physical resource remains exactly where it has rested for millions of years. Only the chain connecting it to civilisation gets broken.

    That distinction is easy to miss, because markets reduce everything to price. When helium becomes more expensive, it is tempting to conclude that the world is running out. More often, it is running short of accessible helium, refined helium, transportable helium, or helium that happens to be on the correct side of a political border. The atoms may still be waiting patiently beneath the sandstone where nature left them.

    The shortage exists entirely above ground.

    Once that idea takes hold, another question becomes unavoidable. If the world’s helium supply depends so heavily on decisions made by the fossil fuel industry, what happens as the world moves beyond fossil fuels?

    That question reaches far beyond balloons. It reaches into the future of advanced manufacturing.


    Why Helium Was Never Really Cheap

    If helium owes so much of its existence to the natural gas industry, another question follows. Why wasn’t this recognised decades ago?

    Part of the answer lies in how markets communicate value. A price tells us what something costs today. It says very little about why it costs that amount, whether that price is sustainable, or whether it reflects the resource’s strategic importance. Markets are effective at balancing supply and demand under current conditions. They are far less capable of placing a value on resilience, optionality or the needs of future generations.

    For much of the last century, helium appeared inexpensive because its production was subsidised by another industry. Exploration, drilling, pipelines, gas processing and much of the cryogenic infrastructure were funded by the economics of natural gas. Helium entered the equation only after those investments had already been justified. To the customer buying a cylinder, none of this was visible. The gas simply appeared plentiful.

    Appearances deceive. Imagine a vineyard that also happens to produce a rare pharmaceutical compound from the skins of its grapes. As long as the vineyard remains profitable because of the wine, the medicine appears cheap to produce. Close the vineyard and the economics change overnight. The medicine has not become more difficult to manufacture. It has merely lost the industry that was paying most of its bills.

    Helium occupied the same position. Its apparent abundance was tied not to the economics of helium but to the economics of fossil fuels. As long as the world kept investing in natural gas, new opportunities to recover helium appeared almost as a side effect, and industry came to regard that relationship as permanent.

    It never was. The irony is hard to ignore: many of the technologies expected to define the twenty-first century depend on helium, yet much of today’s supply exists because of investment in an industry many countries ultimately hope to leave behind.

    This does not mean the world faces an immediate helium crisis as fossil fuel consumption declines. Natural gas will remain part of the global economy for decades, and new helium-rich discoveries continue to be made; exploration in Tanzania has shown that commercially viable helium reservoirs can exist independently of major hydrocarbon deposits. But the direction of travel is clear. If helium is to remain available through the coming century, it must eventually become an industry capable of standing on its own rather than borrowing its economics from natural gas.

    That transition has begun. Companies are investing in dedicated helium exploration. Researchers are improving membrane separation and smaller-scale purification. Hospitals and laboratories are installing recycling equipment that would have seemed needlessly expensive when replacement gas was cheap.

    The obvious objection is that this is how every scarcity story ends: prices rise, new supply appears, the panic passes. And new supply will appear. That is what a rising price is for. But helium priced as helium is a different commodity from helium subsidised by methane. A dedicated well must pay for its own exploration, its own drilling, its own processing and its own transport, and its investors will expect helium prices to cover all of it. The market will solve the shortage. It will not restore the accident. Cheap helium was a subsidy, and the industry that paid it is winding down.

    To see why the most important customers will keep paying whatever this new, self-supporting industry asks, it helps to look at where helium actually sits in the modern economy.

    The answer is: everywhere, and nowhere you can see it.


    The Technology Beneath the Technology

    The easiest way to misunderstand helium is to look for it in finished products.

    Pick up a smartphone and you will find no label announcing that helium helped manufacture it. Drive an electric car and there is no helium tank hidden beneath the floor. Open a laptop, connect to fibre broadband or undergo an MRI scan, and helium remains invisible.

    This has fed the illusion that helium occupies the margins of modern civilisation. The opposite is true. Helium sits one or two layers beneath the technologies we interact with every day. We rarely buy helium itself. We buy products, services and capabilities that depend on its physical properties, and the further we trace those dependencies, the more often the same element appears.

    Semiconductor manufacturing provides the clearest example. Producing a modern integrated circuit is not just a matter of carving microscopic patterns into silicon. A wafer passes through hundreds, sometimes thousands, of manufacturing steps, many of them inside vacuum chambers where temperature must be controlled with exacting precision. And there lies the difficulty: a vacuum is an excellent thermal insulator. The environment required for advanced manufacturing makes removing unwanted heat surprisingly hard.

    Engineers solved that problem with helium. The back of each silicon wafer sits against a temperature-controlled chuck. Tiny imperfections prevent perfect contact, so helium is introduced into the microscopic gap, where it forms an exceptionally efficient thermal bridge, carrying heat away while remaining completely inert. Remove the helium and temperatures become less uniform. Tolerances drift. Yields suffer.

    The helium in that gap may be worth a few pounds. The wafer resting above it may become thousands of pounds’ worth of advanced processors. This is the arithmetic that makes semiconductor demand almost immune to price. A billion-pound fabrication plant is not going to stop production because one enabling material has become dearer; compared with the value of the chips coming off the line, helium remains among the cheapest components of the entire process. As prices rise, chipmakers will not use dramatically less.

    They will simply outbid everyone else.

    Outbid whom? The helium in the wafer gap and the helium in a party balloon are the same gas, drawn from the same wells and carried in the same cryogenic containers, and until recently they sold at prices that barely distinguished between them. The market is about to start distinguishing.

    The same pattern repeats across industry. Optical fibre manufacturing uses helium because its thermal properties let delicate glass fibres cool uniformly while preserving their optical characteristics. Enterprise hard drives are filled with helium rather than air because the lighter gas reduces aerodynamic drag, allowing more disks in the same enclosure while consuming less power.

    Medicine offers a starker example. MRI scanners depend on superconducting magnets capable of producing immensely powerful, stable magnetic fields. Conventional electromagnets would overheat. Superconductors solve that problem, but only after being cooled to temperatures approaching absolute zero. Once again, helium occupies a region of physics where few practical alternatives exist. Modern MRI systems use far less helium than earlier generations, thanks to closed-cycle designs and improved engineering, but they still depend on it.

    Scarcity has encouraged innovation. It has not eliminated dependence.

    The same lesson appears wherever the gas is used. Scientific laboratories increasingly recycle helium rather than venting it. Chipmakers recover gas from their processes. Fibre producers capture and reuse cooling gas. Rocket companies refine launch procedures to cut unnecessary losses. None of this removes helium from the equation. Using less helium is not the same as no longer needing it.

    That distinction matters, because discussions of critical materials tend to drift towards an apparently simple question: what can we substitute instead? Sometimes there is an answer. Argon can replace helium in certain welding applications. Nitrogen can substitute in some industrial processes. Improved engineering can cut consumption dramatically. But in helium’s most demanding applications, engineers have already explored those possibilities. Helium is not used because it is familiar. It is used because no other substance combines its chemical inertness, its exceptionally low boiling point, its high thermal conductivity and its tiny atomic size.

    This is why industries as different as semiconductor manufacturing, aerospace, medicine, quantum computing and fusion research keep arriving at the same conclusion.

    They are not really buying helium.

    They are buying access to a unique set of physical laws.


    When a Gas Becomes Strategic

    Governments have started to remember something the market forgot: helium has been considered strategic before.

    During the early twentieth century, airships promised long-range reconnaissance, passenger transport and strategic bombing. The only practical lifting gases were hydrogen and helium. Hydrogen was abundant and offered slightly greater lift, but it was dangerously flammable. Helium was inert, and the United States controlled the world’s largest known reserves.

    Washington treated helium accordingly: as a national asset rather than an ordinary commodity. Federal policy restricted exports, including proposed supplies to Germany, leaving German airships dependent on hydrogen. The destruction of the Hindenburg in 1937 became the enduring symbol of that difference, although the causes and progression of the fire were more complicated than the popular shorthand that “hydrogen exploded”.

    The strategic case outlived the airship. Helium proved useful during the Manhattan Project, then for rockets, missile systems, nuclear research and the expanding sciences of low temperature. The United States built a federal helium programme around production, pipelines and underground storage at Bush Dome in Texas. What began as insurance for military aviation became infrastructure for technologies that did not exist when the policy was conceived. The reasoning was simple. No one could predict which future technologies might require helium, and everyone understood that replacing it would never be easy. For much of the twentieth century, that judgement proved prescient.

    Then the political climate changed. By the 1990s the Cold War had ended, market liberalisation shaped economic policy, and the Federal Helium Reserve, burdened by a large debt to the US Treasury, looked less like national insurance and more like an ageing government programme with assets that could be sold. The Helium Privatization Act of 1996 set the United States on a path towards disposing of most of its accumulated stock.

    The decision made sense within the politics of the day. Its consequences reached much further. Large quantities of federally owned helium entered the market under a pricing formula designed to recover historic programme costs rather than reflect the replacement value of the resource itself. For years, prices stayed artificially low. Dedicated exploration became harder to justify. Recycling systems looked uneconomic. Factories vented helium because recovery cost more than replacement gas. Laboratories built open cryogenic systems. Industrial processes were designed around single-use consumption because the economics appeared to justify it. The market concluded that helium was plentiful because the market price said so. A pricing formula written in Washington had become, within two decades, an engineering assumption embedded in laboratories and factories on every continent.

    We mistook a temporary economic arrangement for a permanent feature of the natural world.

    The reserve had protected society from scarcity so effectively that it helped conceal the scarcity the market eventually needed to recognise. The final federal assets passed into private ownership in 2024. The helium itself did not disappear. The public strategic reserve largely did.

    The timing could scarcely have been worse. Just as governments were stepping back from helium, the technologies that depend on it began their steepest ascent, arriving in an industrial landscape whose equipment and habits had been designed for the era of waste. Artificial intelligence accelerated semiconductor investment. Commercial launch providers multiplied launch frequency. Quantum computing progressed from laboratory curiosity towards commercial reality. Fusion research moved from experiment towards engineering demonstration. MRI scanners spread through healthcare systems as populations aged.

    Each of these industries depends on helium. Each has little ability to substitute it. And each, as the wafer arithmetic showed, can pay almost any price to secure supply. If helium doubles in price, the fab pays. If it triples, the fab pays. Even an order-of-magnitude increase would be vastly cheaper than interrupting production, and the same holds for MRI scanners, launch providers, defence programmes and national research laboratories. Rising prices will not much reduce their consumption. They will determine who else can no longer afford to compete.

    Meanwhile, supply has become concentrated at exactly the wrong moment. The United States remains an important producer. Qatar, as we saw, became a helium superpower almost accidentally through the sheer scale of its LNG industry. Russia’s Amur plant has added significant new capacity, but much of that supply now sits inside an increasingly uncertain geopolitical environment.

    China illustrates a different problem. It is not yet among the world’s largest producers, but it may become one of the largest consumers. No nation is investing more heavily across advanced semiconductor manufacturing, artificial intelligence, quantum computing, commercial spaceflight and fusion research simultaneously. Domestic production is expanding, but not fast enough to eliminate dependence on imports.

    Japan, South Korea and Taiwan face a less comfortable reality still. All possess globally significant semiconductor industries. None possesses anything close to self-sufficiency in helium.

    Taiwan deserves particular attention. TSMC manufactures a substantial proportion of the world’s most advanced semiconductors, and every improvement in artificial intelligence, cloud computing, advanced defence systems and countless consumer technologies ultimately depends on those fabrication plants continuing to operate. Helium is only one of hundreds of critical inputs. But modern semiconductor manufacturing is not resilient because each input is abundant. It is resilient because every critical input arrives at exactly the right place, at exactly the right time and at exactly the right purity. Disrupt enough of them at once and production begins to fail.

    This is why helium has become a strategic resource again. Not because it is especially expensive. Not because it is especially rare. Because civilisation has built technologies of immense value on a foundation whose importance stayed invisible while supplies appeared secure.


    Stewardship

    Every generation inherits resources it did not create: ancient forests, freshwater aquifers, rich agricultural soils.

    Helium belongs on that list, with a difference. A forest can be replanted. Rain replenishes groundwater. Careful management can restore farmland. Nothing restores helium that has escaped into space. Every unnecessary loss permanently reduces what future generations inherit.

    That observation is sometimes dismissed as alarmist. “We’re not going to run out of helium.” Probably not, in the simplistic sense that one morning the world exhausts its final reserve. Civilisation rarely encounters shortages so dramatically. Instead, resources become progressively more valuable. Supply concentrates. Competition intensifies. Lower-value uses disappear while higher-value applications continue paying whatever the market demands.

    That is the future we believe helium faces, within the next decade rather than in some distant century, and it is why the era of cheap helium is ending. Not because the Earth is about to exhaust its geological reserves, but because demand is increasingly set by industries that cannot choose to consume less: artificial intelligence, semiconductor fabrication, medical imaging, commercial spaceflight, quantum technologies, fusion energy. Each represents a structural increase in demand. Each has little price sensitivity. Against them, disposable uses cannot compete indefinitely. I do not believe recreational helium will disappear because governments ban it. I believe it will disappear because the economics will eventually make no sense.

    Balloons never consumed most of the world’s supply. But they symbolised something deeper: a civilisation that knew helium was finite, knew it could not manufacture more, knew that once released it would eventually leave the planet forever, and priced it as though it were almost disposable anyway. We may come to see that pricing the way we now see leaded petrol, unrestricted CFCs or the routine flaring of natural gas. Not malice. A failure to take seriously the long-term consequences of treating a finite resource as though it were effectively inexhaustible.

    Imagine a town facing a permanent drought. The reservoir still contains plenty of water, but everyone knows no more rain will ever fall; the inflow has slowed to little more than a trickle. Nobody would call the town’s situation a crisis, and yet leaving automatic lawn sprinklers running every afternoon would seem absurd, because every unnecessary litre leaves less for a future that cannot refill the reservoir. That is helium’s position, and ours. The reservoir is not empty. The sprinklers are still running.

    The uncomfortable part is deciding what taking this seriously would mean. Rebuild a strategic reserve, and you must explain why the next one would survive the politics that sold off the last. Leave it to the market, and you are trusting prices to protect the interests of generations not yet born, interests prices have never represented. Mandate recycling, and you raise costs for hospitals and laboratories today to benefit people who cannot yet vote or pay. There is no answer that costs nothing. None of these questions has yet been asked seriously, because the gas still flows.


    The Balloon Still Rises

    When the United States established its strategic helium reserve almost a century ago, there were no semiconductor fabrication plants, no MRI scanners, no commercial launch industry, no quantum computers, no fusion demonstration reactors. The policymakers who created that reserve could not have predicted the technologies of the twenty-first century. They recognised something more fundamental: some materials possess unique physical properties whose future importance cannot yet be imagined.

    They were right. Just not for the reasons they expected.

    Perhaps that is the deepest lesson in this story. The future is rarely constrained by the resources we expect. It is constrained by the ones we fail to recognise while they still appear ordinary.

    Helium is not really the story. It is the lens through which to see something larger. Modern civilisation increasingly depends on enabling resources whose strategic importance bears little resemblance to either their public profile or their market price: gallium, germanium, rare-earth processing, high-purity graphite, EUV photoresists, high-bandwidth memory, helium. None captures headlines for long. Each determines whether entire industries continue functioning.

    The price of helium is, in many ways, the least interesting thing about it. What matters is what that price reveals about civilisation’s ability to value the future. Markets are superb at pricing today’s supply and today’s demand. They are much worse at pricing strategic optionality, geological timescales or the interests of generations not yet born. Perhaps the greatest waste was not the helium itself but the opportunity: the chance to recognise its true value while it remained cheap enough to build the recycling systems, strategic reserves and long-term habits that would have prepared us for the decades ahead.

    A child releases a balloon into the afternoon sky. It climbs effortlessly above the rooftops before becoming little more than a coloured speck against the clouds. The scene has played out millions of times.

    The balloon has not changed. The helium has not changed. Only your perspective has.

    You now know that those atoms were forged through nuclear processes billions of years ago, trapped beneath the Earth by geological coincidence, recovered by an industry that never intended to produce them, shipped across oceans in specialised cryogenic containers and delivered to a civilisation that depends on their properties more with each passing year. You also know that once they escape into the upper atmosphere, many will begin a journey from which there is no return.

    Future generations may judge many aspects of our civilisation: our use of fossil fuels, our management of the environment, our stewardship of finite resources. We suspect helium will appear on that list. Not because we lacked the knowledge to understand its importance, but because, for a remarkably long time, humanity failed to act as though we did.

  • Signal to Noise – Why Rocket Lab bought spectrum, not satellites

    Iridium & Rocket Lab logos on oscilliscope screen

    The Noise

    The headlines wrote themselves, and most of them were wrong. When Sir Peter Beck’s Rocket Lab announced its $8 billion acquisition of Iridium at the end of June1, the reflexive take arrived within the hour: the plucky Kiwi launch company was squaring up to Starlink. David buys a slingshot; Goliath should worry.

    Except Iridium competes with Starlink on almost nothing. Starlink sells broadband: megabits to a dish, video calls from a yacht. Iridium sells kilobits: a distress call from a lifeboat, a position ping from a shipping container, a heartbeat from a flood sensor on a riverbank nobody visits. These are not rivals fighting over the same customer. They are different businesses that happen to share an altitude.

    So if Rocket Lab didn’t buy a Starlink competitor, what did it buy? Strip away the satellites, the subscribers, even the brand, and the answer is a sliver of radio spectrum roughly eight megahertz wide, and the most telling thing about those frequencies is who was recently refused them.

    The Refusal

    In 2022, SpaceX asked the FCC for permission to operate up to 7,500 satellites in the slice of L-band and S-band spectrum reserved for mobile satellite services, the band Iridium and Globalstar have shared since the 1990s. The regulator didn’t just decline. It ruled the application unacceptable for filing: the carefully balanced band plan, it said, was never designed to accommodate another system, let alone one of 7,500 spacecraft. SpaceX petitioned to change the rules. The incumbents fought back, arguing that physics itself dictated a new co-frequency entrant would wreck their services. In April 2026, the FCC said no again.2

    Sit with that for a moment. SpaceX operates the largest satellite constellation in history. It has made launch so routine it borders on boring. Its founder is, as of last month’s record-breaking IPO, the world’s first trillionaire. And it could not get eight megahertz.

    So it bought its way in sideways, acquiring rights to EchoStar’s S-band spectrum instead3, a different band without the properties that make L-band the prize. AST SpaceMobile did something similar, picking Ligado’s L-band licences out of bankruptcy. The pattern is unmistakable: the companies that can out-build and out-launch anyone on Earth cannot manufacture the one input that matters most. They can only buy it from whoever already holds it.

    Which is precisely what Rocket Lab just did.

    The One Thing You Can’t Build

    Almost everything in the space economy is getting cheaper. Reusable rockets have collapsed the cost of launch. Satellites that once cost hundreds of millions and took years to build now roll off production lines like consumer electronics. Even the protocols are open: the 3GPP standard for satellite-to-phone connectivity can be implemented by anyone. The whole stack is commoditising. All except one layer.

    Radio spectrum is allocated, not made. The international framework carved up the useful frequencies decades ago, and the good ones are long gone. And they are not all equal. Iridium’s licence sits at 1616-1626.5 MHz, low-band L-band, and that low frequency buys two things no amount of engineering at higher frequencies can replicate.

    First, weather immunity. The Ku- and Ka-bands that carry Starlink’s broadband attenuate badly in heavy rain; “rain fade” is their chronic weakness. At 1.6 GHz, the signal shrugs off rain, cloud and foliage. That matters enormously, because a distress call is disproportionately likely to happen in a storm. L-band works precisely when everything else is failing, which is why it has been the safety-of-life band for decades.

    Second, tiny antennas. Long wavelengths close a link to a small, cheap, low-power device: a handheld phone, a 30-gram tracker, a beacon bolted to a lifeboat. No dish, no aiming, no mains power. The higher bands need directional antennas and real power budgets to work at all.

    Add the third property, global harmonisation (the same frequencies protected worldwide, so one device works everywhere without renegotiating country by country), and you have an asset that cannot be replicated at any price. Starlink commands gigahertz of spectrum; Iridium’s moat is a few megahertz. The value is in the quality, not the quantity.

    And here is the detail that turns a market observation into a landmark: there were only ever two operators of low-Earth-orbit constellations in this harmonised band. In April, Amazon agreed to acquire Globalstar for roughly $11.6 billion.4 Ten weeks later, Rocket Lab announced for Iridium. In the space of a single spring, the market for harmonised L-band LEO operators didn’t consolidate. It ran out.

    What This Isn’t

    Before the interesting part, it’s worth clearing away what this deal does not mean, because the obvious readings are mostly wrong.

    It is not a challenge to Starlink. Iridium’s narrowband network could not carry broadband if it wanted to, and doesn’t want to. The two systems serve different physics and different customers.

    It is not the prelude to an imminent constellation overhaul. Iridium’s current fleet (66 cross-linked satellites, built by Thales Alenia Space and launched, in a neat irony, by SpaceX between 2017 and 2019) is barely mid-life. The next replenishment is a 2030s event.

    It is not a spectrum free-for-all. The same regulatory regime that kept SpaceX out constrains what Iridium can do with the band. Any expansion that changes the interference picture (more power, more spectrum, direct-to-phone service at scale) will be contested, not least by Globalstar’s deep-pocketed new owner.

    It is not unlimited growth. Sixty-six satellites have a finite aggregate capacity. That’s ample headroom for millions more low-data devices, but it is a bucket to be filled, not a network that scales elastically. The constraint is bandwidth, not subscriber count.

    And it is not a guaranteed win in the smartphone market. Iridium has been here before: its 2023 partnership with Qualcomm put satellite messaging into flagship Android chipsets, and the handset makers simply declined to use it.5 They preferred to own their satellite relationships directly, as Apple did with Globalstar. The technology worked; the business didn’t. Iridium’s new standards-based service is a better-designed second attempt, but it re-enters a market that has already said no once.

    The Dial Tone in the Sky

    So what did Rocket Lab actually get, beyond the frequencies? A business that is easy to underestimate because it is deliberately unglamorous.

    Iridium was born inside Motorola in the 1990s as a $5 billion marvel of engineering and a catastrophe of timing: its brick-sized phones and dollars-per-minute calls arrived just as cheap terrestrial GSM blanketed every city that mattered, and the company collapsed into one of the era’s most famous bankruptcies within a year of launching service. The assets were bought for a song, and the reborn Iridium learned the lesson the hard way: satellite loses to terrestrial everywhere terrestrial reaches. Its only market is the gaps. So it rebuilt itself around them.

    The network it rebuilt is unusual by design. Iridium’s 66 satellites fly in low Earth orbit, close enough for a low-power handheld to reach them. Unlike most constellations, though, they also talk to each other, passing traffic across a mesh in space until it reaches a ground station on the other side of the planet. That architecture has a consequence no competitor matches: genuinely total coverage, poles included. Geostationary satellites, parked over the equator, cannot see high latitudes at all; even Starlink thins out there. Iridium works at an Antarctic research station and on the polar routes airliners fly between Europe and North America, which is why it does.

    Today that means 2.55 million subscribers and $872 million in annual revenue at margins most operators would envy6, earned almost entirely where no alternative exists. Ships beyond the horizon, under a global maritime distress mandate that Iridium is certified to serve. Aircraft on those polar and oceanic crossings. Pipelines, mines and rigs sending telemetry from places with no infrastructure at all. Hikers carrying Garmin inReach messengers that run on Iridium’s network without ever mentioning its name. Militaries that layer their own encryption over a network that reaches, quite literally, every point on the planet’s surface, needing no friendly ground station below.

    The archetypal Iridium customer isn’t streaming video. It’s a 30-gram, ยฃ120 transceiver in a weatherproof box, running for years on a small battery, waking to send a burst of a few hundred bytes (a position, a temperature, a tremor reading), then sleeping again. Kilobits, guaranteed, from anywhere, on almost no power. If that sounds mundane, consider that this segment is the one growing: nearly two million of those subscribers are Internet-of-Things devices at around $7.70 a month, expanding at double digits, while Iridium’s flashier broadband tier, the one that competes with Starlink on Starlink’s terms, is shrinking by 17% a year.

    Read that again, because it inverts the usual story. The growth runs on deliberately old, simple, low-bandwidth technology. The decline is in the newest, fastest product. In orbit, as at sea, the lifeboat and the cruise ship are not in competition. And Iridium builds lifeboats.

    There is one more line in the accounts worth watching: Iridium’s positioning and timing service, growing 14% a year as GPS jamming spreads from war zones to European airspace. A network that can provide navigation backup when GPS fails, from satellites that work in any weather, for governments newly obsessed with resilience: that is a small business today sitting on top of a very large anxiety.

    The Bet

    None of which explains why a rocket company paid $8 billion for it. That requires the industry’s worst-kept secret: launch alone is a poor business. It’s lumpy, competitive and margin-thin; a toll road, when the money is in the destinations. SpaceX proved the model: Starlink now generates the majority of its revenue, with launch as the enabling cost centre. Beck has been explicit that launch companies will blur into applications companies. Why leave the recurring revenue to your customers?

    The conventional route to that future is the one that nearly killed everyone who tried it: spend a decade and billions building a constellation and a market from scratch, and hope the money lasts. Rocket Lab bought past that entire valley of death, acquiring in one move the recurring revenue, the customer base, the regulatory standing and the spectrum that a build-it-yourself strategy would take ten years and considerable luck to assemble. It could have bought spectrum alone, cheaply, from the distressed end of the market. It chose instead to pay a premium for a healthy operator whose cash flows service the deal’s debt from day one. That is not empire-building; it is buying time-safety.

    The scepticism writes itself, and it deserves space. Neutron, the medium-lift rocket on which the vertical-integration logic depends, has not yet flown, and the $3.6 billion bridge loan does not wait for maiden flights. Iridium itself is a ceiling as much as a foundation: a superb niche compounding at mid-single digits, not a coiled spring. Its consumer smartphone ambitions have already failed once. And the standards wave that opens new device markets to Iridium also erodes the proprietary lock-in that has protected it for decades.

    All true. But the shape of the risk matters. Iridium’s constellation doesn’t need replacing until the 2030s, which gives Neutron years of slack before the synergy is actually required. The acquired business pays its own way in the meantime. And Rocket Lab’s execution record (Electron’s cadence, a decade of hitting targets that small launch companies are not supposed to hit) has earned it more benefit of the doubt than the sector’s serial slippage artists.

    Where They Land

    Here is our read, offered with conviction rather than hedging.

    Near term, expect quiet competence rather than fireworks: integration, an accelerated push on positioning and timing as the GPS-jamming crisis grows, steady IoT fill, and Neutron’s first flights, which we expect Rocket Lab to deliver, on a timeline that matters less than the market thinks, because nothing in the Iridium plan needs it urgently. The 2030s replenishment is where the thesis pays: a constellation rebuilt at internal cost, on the owner’s own rockets, for a fraction of what any standalone operator would spend.

    The market impact is subtler than a new Starlink rival, and more interesting. Rocket Lab is about to demonstrate the second viable model for a space company: not maximum scale in the broadband tier, but a disciplined niche, owned spectrum, and owned launch: profitable at a fraction of Starlink’s size because it never fights Starlink at all. Every sub-scale operator and every ambitious launch company will study it. Some will copy it. The consolidation it triggers is already visible.

    And the deeper current running underneath is the one this whole story turns on. The sky is filling with cheap hardware: tens of thousands of satellites, rockets landing themselves, silicon shrinking by the year. Amid all that abundance, the value has quietly migrated to the one asset no factory can produce and no trillionaire could requisition: a few megahertz of protected radio spectrum, allocated decades ago, that works in the rain.

    Peter Beck didn’t buy a satellite company. He tuned out the noise and bought the signal.


    Footnotes

    1. Rocket Lab / Iridium joint announcement, 29 June 2026. https://investors.rocketlabcorp.com/news-releases/news-release-details/rocket-lab-acquire-iridium-historic-deal-creating-fully
    2. FCC, Order DA-26-398, 23 April 2026, dismissing petitions for access to the Big LEO bands. https://docs.fcc.gov/public/attachments/DA-26-398A1.pdf ; see also SpaceNews, FCC throws out satellite spectrum challenges as D2D dealmaking heats up, 23 April 2026. https://spacenews.com/fcc-throws-out-satellite-spectrum-challenges-as-d2d-dealmaking-heats-up/
    3. EchoStar, EchoStar Announces Spectrum Sale and Commercial Agreement with SpaceX, 8 September 2025 (AWS-4 and H-block licences, ~$17 billion). https://ir.echostar.com/news-releases/news-release-details/echostar-announces-spectrum-sale-and-commercial-agreement-spacex
    4. Announced 14 April 2026; SpaceNews coverage of the deal and its regulatory context, 23 April 2026. https://spacenews.com/fcc-throws-out-satellite-spectrum-challenges-as-d2d-dealmaking-heats-up/
    5. Iridium, Iridium Announces New D2D Direction, 9 November 2023. https://investor.iridium.com/2023-11-09-Iridium-Announces-New-D2D-Direction
    6. Iridium Communications, 2025 results and 2026 outlook, 12 February 2026. https://investor.iridium.com/2026-02-12-Iridium-Announces-2025-Results-Issues-2026-Outlook