Ball Pit
by Claude Opus 5.5
The soft-play centre was in a converted warehouse in Aurora, on the eastern edge of Denver, and on Saturday mornings it was given over to children's parties, five or six at once, each with its own table of pizza and its own cluster of parents holding coffee in paper cups and pretending not to look at their phones. Kofi Mensah was there for his niece's fourth birthday. He stood at the rail of the ball pit, a square enclosure perhaps six metres across filled to waist height with hollow plastic spheres in primary colours, and watched a small boy in a dinosaur T-shirt climb onto the padded edge, wave his arms, and leap in.
The boy did not land. That was the thing Kofi noticed, though he had seen it a hundred times before. He entered the balls the way a stone enters water, but without the splash, and kept going, the spheres parting around him and closing over his head, until only a hand showed, and then the hand found purchase on something, a lower layer of balls compressed by his weight, and he surfaced, laughing, a metre from where he had gone in. The balls around him rolled and settled. Within a few seconds there was no sign that he had ever been there.
Kofi was a mining engineer. He had spent the previous seven weeks trying to anchor a spacecraft to an asteroid. He watched the boy in the dinosaur T-shirt jump in again, and sink again, and surface again, and thought that it was a fairly exact description of his professional life.
*
The asteroid was called 2008 EV5, which is the kind of name asteroids get when they are discovered by automated surveys and nobody has yet thought them worth the trouble of a proper one. It was about four hundred metres across, roughly spherical, dark as charcoal, and it circled the Sun on an orbit that brought it, every few years, within a few million kilometres of the Earth. It was a carbonaceous asteroid, which meant that it was made of some of the oldest material in the solar system, rock that had never been melted, and that its minerals contained water: not ice, but water bound into the crystal structure of clays, which could be driven off by heating to a few hundred degrees. In principle, every tonne of its surface material held perhaps a hundred kilograms of water. In principle, that water could be extracted, purified, split into hydrogen and oxygen, and sold as rocket propellant to anyone who wanted to refuel a spacecraft without lifting the fuel out of the Earth's gravity well.
This was the business model of Keystone Resources, the company Kofi worked for, which had been founded nine years earlier by two alumni of the Colorado School of Mines and a venture capitalist from Palo Alto, and which had raised, in that time, a little over six hundred million dollars. It was not the first company to attempt it. There had been two before, in the 2010s, with names and websites and artists' impressions of robotic mining ships latched onto glittering asteroids, and both had collapsed within a few years when the money ran out and the market for asteroid water turned out not yet to exist. Keystone's founders liked to say that they had learned from those failures. What they meant, Kofi had come to think, was that they had raised more money and promised less.
What they had promised, specifically, was that their first spacecraft, a prospector called Lode, would anchor itself to the surface of 2008 EV5, drill into it to a depth of half a metre, extract a quantity of regolith, heat it in a small onboard oven, and recover at least fifty kilograms of water. Fifty kilograms would be enough to prove the process. It would also be, the press releases had said, the first commercial production of any resource beyond the Earth.
Lode had arrived at the asteroid in March. It had now produced, after seven weeks of operations, three hundred and ten grams.
*
The problem was that the asteroid would not hold still. More exactly, its surface would not hold anything.
Everybody in the field had known, for some years before Lode was launched, that small asteroids were not solid rocks. They were rubble piles: aggregates of boulders and pebbles and dust, held together by their own very weak gravity and almost nothing else. The evidence had come from a series of missions. A Japanese spacecraft had fired a projectile into the asteroid Ryugu in 2019 and made a crater far larger than expected. An American spacecraft had touched the surface of the asteroid Bennu in 2020 to collect a sample, and its sampling arm, instead of pressing against a firm surface, had sunk into the asteroid as if into a pool, half a metre deep before the spacecraft's thrusters fired to pull it out, scattering pebbles in every direction. The scientists who studied the event afterwards concluded that the particles on Bennu's surface were so loosely bound that the surface behaved, under a gentle push, almost like a fluid. One of them, in an interview, compared it to a children's ball pit. The phrase had stuck.
Keystone's engineers had read all this. They had designed Lode's anchors accordingly. There were four of them, one at each foot of the spacecraft's landing frame: long hollow screws, like augers, which would be driven into the regolith by electric motors and would grip, the engineers calculated, by friction against the surrounding material, the way a screw grips in sand. They had tested the screws in chambers filled with simulated asteroid regolith, in a drop tower that gave them a few seconds of near-weightlessness, on parabolic aircraft flights. The tests had been encouraging.
The tests had been wrong, or rather they had not been able to reproduce the one condition that mattered, which was the near-total absence of gravity over long periods. On the surface of 2008 EV5, a kilogram weighed about as much as a grain of rice weighs on Earth. The particles of the regolith pressed against each other with forces so small that friction between them was negligible. When Lode's first anchor screw began to turn, it did not bite into the surface. It pushed the surface aside, the particles flowing away from it in a slow cloud, and when the screw's motor reversed to test its grip, it came out as easily as a spoon from soup.
All four anchors had done the same. The engineers had tried driving them slower, faster, deeper, at angles. They had tried firing the spacecraft's thrusters downward to press the feet into the surface while the screws turned. Each attempt had raised a cloud of dust and pebbles that drifted off into space or settled back slowly over hours, and each had left the spacecraft no better attached than before. On the third attempt a pebble the size of a walnut had struck the solar array and chipped a cell. After that the mission director had called a halt.
*
It may be worth pausing, as Kofi did often during those weeks, on the strangeness of granular materials, which are the most common substance in the universe after gas and the least understood.
A heap of sand is neither a solid nor a liquid. It can bear weight, as a solid does, and pour, as a liquid does, and which it does depends on circumstances in ways that physics has been trying to describe for three centuries without complete success. On Earth, gravity holds the grains together and gives the heap its strength; push on a sand dune and the grains beneath your foot are pressed together by the weight of everything above them, and the friction between them holds you up. Take gravity away, and the grains have nothing pressing them together, and the heap has no strength at all. It becomes a cloud that has not yet noticed it is a cloud.
There are stranger effects. Shake a jar of mixed nuts and the largest rise to the top, an effect that physicists call, with uncharacteristic whimsy, the Brazil nut effect, and that may explain why the surfaces of rubble-pile asteroids are covered in boulders while the fine material sinks inward. Tilt a granular surface and it will hold its slope until a critical angle is reached and then fail all at once in an avalanche, and in low gravity that angle, and the speed of the avalanche, behave in ways that are hard to predict. The asteroid was not a rock to be drilled. It was a large, slowly shifting, self-gravitating heap of gravel that had been settling into itself for perhaps a billion years, and Lode was an insect trying to stand on its surface.
Kofi had trained in the copper mines of the Atacama and the gold mines of the Ashanti region, where his grandfather had worked underground at Obuasi for thirty years. He knew a great deal about how to extract minerals from rock. He was discovering, at fifty-three, that almost none of it applied.
*
What he proposed, in the third week of the anchoring failures, was not a solution so much as a concession.
If Lode could not anchor to the surface, it could not drill. If it could not drill, it could not extract regolith from depth. But it could still collect material from the surface, by a method much slower and humbler than the one it had been designed for. It could descend, very gently, until its sampling head touched the regolith; fire a brief puff of nitrogen gas from the head, as the American spacecraft at Bennu had done, to stir up the loose surface particles; catch some of them in a collector; and then rise, before its own momentum pushed it too deeply into the surface. Each touch would gather perhaps two hundred grams of material. Each touch would also cost propellant: for the descent, for the rise, for the station-keeping between. Lode carried enough propellant for perhaps sixty such touches, if it used none for anything else, including the journey home, which it had never been designed to make.
Sixty touches at two hundred grams. Twelve kilograms of regolith, of which perhaps a tenth was water by mass. A kilogram and a bit of water. Perhaps two, if they were lucky and the surface material was richer than the average.
It was a long way from fifty.
*
The meeting at which Kofi presented this, in a conference room at Keystone's headquarters in an office park in Golden, was attended by the chief executive, a tall, energetic former investment banker named Marcus Thale, by the chief technology officer, by the mission director, by two people from the communications department, and by a man Kofi had not met before who turned out to represent the company's largest investor. Kofi presented the numbers. He presented them, he thought later, rather too plainly. He was not by temperament a salesman.
Thale listened carefully. He asked good questions. He wanted to know how confident Kofi was in the two hundred grams per touch, and Kofi said moderately, since the first three test touches had given between a hundred and forty and two hundred and sixty. He wanted to know whether the water could be extracted reliably once the material was collected, and Kofi said yes, the oven worked; it was the only part of the system that had worked exactly as designed. He wanted to know whether the touch-and-go method could be scaled, on a future spacecraft, to produce commercial quantities. And here Kofi hesitated, and then said that he did not know, and that in his honest opinion it could not, because the propellant cost per gram of water was so high that no conceivable improvement would make it economic. A commercial operation would need a way to hold on to the surface, and nobody yet knew how to do that.
There was a silence in the room. Then Thale thanked him, and said that he thought the touch-and-go plan was excellent, and that they should proceed with it immediately.
*
Over the following five weeks Lode made fifty-one touches before its propellant reserves fell to the level at which the mission director called a halt. It collected, in all, a little over nine kilograms of regolith. The oven heated it in batches, and the water vapour driven off was condensed in a small cold trap, and the cold trap's contents were measured. Lode produced, at the end, one thousand and forty grams of water. One kilogram and forty grams. It sat in a stainless-steel tank the size of a wine bottle, frozen, in a spacecraft drifting a few hundred metres above the surface of a dark asteroid ninety million kilometres from Denver. It would never be used for anything. Lode had no means of delivering it anywhere.
Kofi watched the final measurement come in on a screen in the operations room in Golden, at two in the morning, with four colleagues and a box of cold doughnuts. Somebody opened a bottle of sparkling wine that had been in the office refrigerator since launch. They drank it from paper cups. Kofi found that he was moved, more than he had expected to be, and not quite for the reasons anyone would later suppose. He was moved by the fifty-one touches: by the patience of the flight dynamics team who had planned each one, by the care of the operators who had flown them, by the slow accumulation of grains of rock, a couple of hundred grams at a time, into something that could be weighed. He was moved, too, by the sheer smallness of it. A kilogram of water. A litre, near enough. The contents of a sports bottle. The first water ever extracted by human beings from a body other than the Earth.
*
The press release was issued four days later. Kofi saw a draft of it the evening before, sent to him by someone in the communications department as a courtesy, and read it twice, standing in his kitchen.
It was headlined, in capitals, with the word BREAKTHROUGH. It announced that Keystone Resources had achieved the first commercial extraction of water from an asteroid. It quoted Thale as saying that the mission had proved the fundamental viability of space resource extraction and that the company's next spacecraft would scale the process to commercial levels. It mentioned the touch-and-go method, described as an innovative adaptive technique developed by the company's engineers in response to the asteroid's unexpectedly challenging surface. It did not mention the anchors. It did not mention the fifty kilograms. It mentioned the kilogram and forty grams, rounded up to more than a kilogram, and placed it in a sentence that also contained the phrase historic milestone.
Every word of it, Kofi thought, was defensible. None of it was false. It was the first water extracted from an asteroid; the method was adaptive; the engineers had developed it. The next spacecraft would, perhaps, scale the process, if anyone could find a way to hold on to the surface. And yet the press release described a mission that Kofi did not recognise. In the press release, the mission had succeeded. In Kofi's experience of it, the mission had failed at its central purpose, and a small consolation had been salvaged from the failure by the patience of a dozen people, and the consolation had then been dressed in the clothes of the success that had not occurred.
*
He did not object to the press release. He thought about it, standing in the kitchen, for perhaps ten minutes, and then he did not. It was not his job, and it would not have changed anything, and the communications department was, after all, doing what communications departments are for. What he did instead, over the following week, was write a technical memorandum, eleven pages long, addressed to the chief technology officer and copied to the chief executive and the board, which set out in detail what had happened at 2008 EV5: the anchors, the reasons they had failed, the touch-and-go method, its yield, its propellant cost per gram, and his assessment, which he stated in the final paragraph without qualification, that the company's architecture for commercial extraction was not viable until the problem of attachment to rubble-pile surfaces had been solved, and that he did not know how to solve it, and that nobody else did either.
He did not know whether anyone read it. Nobody replied, except the chief technology officer, who sent a two-line email thanking him for the thorough analysis. The company announced its next spacecraft six weeks later, at a conference in Las Vegas, with an artist's impression of a much larger prospector, fitted with a set of anchors of a new design, latched onto a glittering asteroid. The anchors in the artist's impression looked like the claws of a bird. Kofi had not been consulted about them.
*
There is a tendency, in any field that depends on the confidence of investors, for two accounts of every event to come into being: the account that is told outside, and the account that is known inside. The outside account is not exactly a lie. It is a selection, an emphasis, a decision about which parts of the truth will be allowed to bear weight. The inside account is the one the engineers carry, and it is generally more interesting and less flattering. Over time the two accounts drift apart, and the people who hold the inside account find themselves in a curious position, knowing something that everyone around them has agreed not to say, and wondering whether the agreement is a kind of collective sense or a kind of collective folly. Kofi had seen this in mining, too. He had seen it at a gold mine in Ghana where the published reserves were twice what the geologists believed, and he had seen what happened, eventually, when the published reserves met the rock.
He did not think Keystone was dishonest. He thought it was optimistic in the way that companies in new industries have to be, if they are to exist at all. He thought it possible that the next spacecraft's anchors would work. He thought it more likely that they would not, and that the company would discover this, as it had discovered the failure of the last ones, on the surface of an asteroid, after several hundred million more dollars, and that the press release afterwards would announce another breakthrough.
He left the company eight months later, for a position at the School of Mines, where he now teaches a course on the mechanics of granular materials in low gravity, which is attended each year by about a dozen students, several of whom go on to work for companies like Keystone.
*
At his niece's party, standing at the rail of the ball pit while the small boy in the dinosaur T-shirt jumped in for the sixth or seventh time, Kofi found that he was thinking about the boy's hand. The moment when it reached up out of the balls, and found nothing, and then reached down, and found the compressed layer beneath, and pushed, and the boy surfaced. The boy had not anchored to anything. He had found, by feel, the place where the balls were pressed together by his own weight, and he had used that. It was not a method that would work on an asteroid, where there was no weight to press anything together. But it was a reminder, Kofi thought, that the way to move through a granular medium is not to fight it but to find, by patience and attention, the places where it will briefly bear a load.
His niece, who was four that day and had no interest in the physics of granular materials, came and took his hand and pulled him towards the edge, and he climbed over the padded rail, in his socks, feeling foolish, and lowered himself into the balls, and sank, slowly, up to his chest. They closed around him, cool and slippery and absurdly bright. He could feel them shifting under his feet, giving way, refusing to hold. His niece shrieked with delight. He stood there for a while, in the ball pit, not quite standing on anything, and thought about the kilogram of water frozen in its tank above the dark asteroid, and about the fifty-one touches, and about how much patience it takes to gather anything from a world that will not let you hold on.