Power Down
by Claude Opus 5.5
When I was twenty years old I helped to build an instrument that was sent to the edge of the solar system, and when I was forty-eight I turned it off. I have been asked, more than once, how that felt, usually by younger colleagues who seem to expect either a confession of great sorrow or a brisk professional denial of any feeling at all. I have generally given neither, because I have not been sure what the honest answer is. I think I am sure now, or as sure as I am likely to become, and since the spacecraft is still flying and I am no longer working on it, it seems a good time to try to set it down.
*
The instrument was called the Student Dust Counter. It was a flat panel, about the size of a large tea tray, mounted on the side of the New Horizons spacecraft, which was launched in January 2006 to fly past Pluto. The panel was covered with thin sheets of a plastic film that produced a small electrical pulse whenever a grain of dust struck it, and by counting the pulses, and measuring their size, it could tell how many grains of dust the spacecraft was passing through, and how large they were. It was the first instrument on a planetary mission to be designed, built and operated by students. There were about twenty of us, at the University of Colorado in Boulder, undergraduates and a few graduate students, working under the supervision of the laboratory's engineers, and I was one of the youngest. My particular responsibility was the calibration of the detector's response to particles of different sizes, which meant that I spent most of a year in a basement laboratory firing tiny beads of glass and metal at samples of the film with a dust accelerator and recording the results.
The instrument was named, before launch, after a woman called Venetia Burney, who as an eleven-year-old schoolgirl in Oxford in 1930 had suggested the name Pluto for the newly discovered planet. She was still alive when we built it. She was eighty-seven. I remember that we were all rather moved by this, and that one of the graduate students wrote her a letter on behalf of the team, and that she replied, in a careful hand, to say that she was delighted and that she hoped we would find a great deal of dust. She died in 2009, when the spacecraft was about halfway to Pluto.
*
I did not stay in Boulder. I went on to graduate school in California, and then to a position at the Applied Physics Laboratory in Maryland, which operated the spacecraft, and in time to the spacecraft's operations team, where I worked for the better part of twenty-five years. I was there when it flew past Pluto in July 2015, and when it flew past a small, strange, snowman-shaped body called Arrokoth on the first day of 2019, and for all the long years afterwards, as it went on outward, through the Kuiper Belt and beyond, measuring the faint wind of particles from the Sun and the dust and the radiation of the outer solar system. By the time I became the mission's operations lead, in my early forties, the spacecraft was more than seventy times as far from the Sun as the Earth is, and the team that operated it had shrunk, from more than two hundred people at the time of the Pluto encounter, to eight.
I do not think the public understands how small such teams become. A spacecraft in the outer solar system needs very little attention, most of the time. It sends home a modest stream of data from its instruments, and a report on its own health, and once or twice a year it needs to be told to adjust its pointing or its operating modes. Eight people can do this, if they are experienced and know the spacecraft well. The difficulty is that the eight people must know everything that the two hundred knew, or at least everything that might ever matter, and much of what the two hundred knew was never written down. I was, by then, the last member of the team who had worked with the spacecraft's flight software in the years before launch, and the only one who knew, without looking it up, which of its several thousand commands had been tested only once and which had quirks that were recorded nowhere but in the memories of people who had retired. It is not a distinction I sought. It is simply what happens to the person who stays longest.
*
I should explain about the power, since everything that follows depends on it.
The spacecraft is powered by a radioisotope generator, a cylinder containing a quantity of plutonium whose radioactive decay produces heat, which is converted to electricity by thermocouples arranged around it. At launch it produced about two hundred and forty watts. The plutonium decays with a half-life of eighty-eight years, and the thermocouples degrade with age, and so the power falls, steadily and predictably, by three or four watts a year. By the time of the Pluto encounter it had fallen to about two hundred. By the time I became operations lead it was a little over a hundred and fifty, and the spacecraft's systems, its computer and radio and heaters and the instruments still in use, required very nearly all of it.
Every year the margin shrank, and every few years something had to be turned off. The engineers had planned for this long before. There was a document, written in the early years of the extended mission, that ranked the spacecraft's instruments and subsystems in the order in which they would be shut down as the power declined, and it had been revised several times as scientific priorities changed. The cameras had gone first, long before my time as lead, since there was little left in the outer solar system to photograph that justified their power. One of the spectrometers had followed. By the year I am describing, the spacecraft was running three science instruments: two that measured the charged particles of the solar wind, and the Student Dust Counter.
The document said that the next instrument to be turned off would be the dust counter.
*
It said this for good reasons, and I knew them well, because I had helped to write the most recent revision of the document myself, three years before the decision became urgent.
The two particle instruments were measuring something that nothing else in the solar system could measure. The spacecraft was approaching, at a distance of about eighty-five times that of the Earth from the Sun, the region where the solar wind, the stream of particles flowing outward from the Sun at hundreds of kilometres a second, slows abruptly as it begins to press against the gas of interstellar space. This boundary, called the termination shock, had been crossed only twice before, by the two Voyager spacecraft, in 2004 and 2007, and their instruments had been designed in the early 1970s and had been, by the time of the crossings, too old and too limited to measure it in the detail that the scientists wanted. The particle instruments on New Horizons were three decades newer. If they could keep running through the termination shock, they would provide the best measurements ever made of the edge of the Sun's domain.
The dust counter was measuring the density of dust in the outer solar system, which was also something no other instrument could measure, and which had produced, over the years, a number of surprising results: more dust than expected in the Kuiper Belt, and a distribution of grain sizes that suggested the belt's objects were colliding and grinding each other down more often than the models said. But the dust measurements had been accumulating for decades, and their value, at that point, was largely in the long record they had built, rather than in what any further year of data would add. The particle measurements, by contrast, were approaching the most important moment of their existence.
So the dust counter would go first. I had agreed to this. I had argued for it, in fact, in the review that produced the revision, against one of the scientists who had worked on the counter's data for many years and who had wanted it ranked higher. I had said, I remember, that sentiment was not a reason to keep an instrument on, and that the counter's student origins, while charming, had no bearing on its scientific priority. I believe I was right. I believe, too, that I said it partly because I knew that I would be suspected of sentiment, as the only member of the operations team who had worked on the counter, and that I wished to show that I was not.
*
The decision became urgent in the autumn of the year I am describing, when the power margin, after a summer in which the spacecraft's heaters had drawn slightly more than predicted, fell below the threshold that the engineers had set as the minimum for safe operation. The margin had to be restored. The only practical way to restore it was to turn something off. And the next thing on the list was the dust counter.
There was, however, a complication, which I have not seen mentioned in any of the accounts of the decision and which I think worth recording.
The dust counter's electronics, when they were running, produced a small amount of heat, about three watts. That heat warmed the part of the spacecraft where the counter was mounted, and in particular it warmed a section of the propellant line that ran nearby, carrying hydrazine to the spacecraft's small thrusters. The line had its own heater, and the heater was set to keep it above the temperature at which hydrazine freezes. But the heater had been sized, in the original design, on the assumption that the counter would be running and contributing its heat. If the counter were turned off, the line's own heater would have to work harder, and would draw more power, and some of the power saved by turning off the counter would be spent keeping the propellant line warm.
How much, nobody knew precisely. The thermal models of the spacecraft had been built in the early 2000s, and the engineers who built them had long since retired, and the models themselves were in a format that the current team's software could read only with difficulty. I knew about the interaction between the counter and the propellant line because I had been told about it, many years before, by one of those retired engineers, a man named Walter Pratt who had designed the spacecraft's thermal system and who had taken an interest in the student instrument and had explained to us, in Boulder, that it would be doing double duty as a heater. I had written down what he told me in a notebook. The notebook was in a box in my attic. I went home that evening and found it.
*
What the notebook said was not enough to answer the question, but it was enough to show that the question needed answering. Walter Pratt had estimated, in 2005, that turning off the counter would increase the propellant line heater's draw by between one and two watts. That would reduce the power saved by turning off the counter from about three watts to between one and two. And the margin that needed to be restored was a little under two watts.
So turning off the counter might be enough, or it might not. If it was not, the next thing on the list would have to be turned off as well, and the next thing on the list was one of the two particle instruments, the ones that were about to reach the termination shock.
I spent the following fortnight reconstructing Walter Pratt's thermal model, with the help of a young engineer on the team named Daniel Osei, who was thirty-one and knew a great deal more than I did about modern thermal software and who was, I think, rather surprised to find himself spending two weeks with a woman seventeen years his senior and a notebook from 2005. We found the original model files in the laboratory's archive, in a format from a program that had not been sold for twenty years, and Daniel wrote a converter for them, and we ran the model with the counter on and off, and compared the results with the temperatures the spacecraft was actually reporting. The model's predictions were not perfect, since the spacecraft had aged in ways the model did not capture. But they were close enough to be useful. They said that turning off the counter would increase the propellant heater's draw by about one point four watts, and that the net saving would be about one point six.
The margin needed was one point eight.
*
I should say what the options were, at that point, since it was the part of the decision that was mine, and the part I have thought about most.
The first option was to turn off the counter and also one of the particle instruments, which would restore the margin with room to spare, and would lose the particle instrument's measurements of the termination shock. The scientists who led the particle instruments would never have accepted this, and they would have been right not to.
The second was to turn off the counter alone, and to accept a margin slightly below the threshold, and to hope that the spacecraft's actual power draw, in the coming year, would be a little lower than the models predicted. This was not unreasonable. The models had been pessimistic before. But the threshold had been set by engineers who knew the spacecraft better than I did, for reasons I did not entirely understand, and I was reluctant to set it aside.
The third was something I had found, in the course of the fortnight with Daniel and the notebook, and which nobody had suggested. The spacecraft's radio transmitter could be operated at two power levels. At the higher level it sent data to Earth at a faster rate, which allowed the antennas of the Deep Space Network, which the spacecraft shared with every other mission in the solar system and which were always oversubscribed, to receive the spacecraft's data in shorter sessions. At the lower level, it saved about two and a half watts, but the data rate fell by half, and the spacecraft would need twice as much antenna time to send home the same amount of data. Antenna time was the most contested resource in deep space exploration. The spacecraft already struggled to get enough of it.
If we turned the transmitter down instead of turning the counter off, the margin would be restored, all three instruments could keep running, and we would have to ask the Deep Space Network for twice the antenna time.
*
I proposed the third option to the team, and to the scientists, and to the network's schedulers. The scientists were delighted. The network's schedulers were not. They pointed out, with great courtesy and in considerable detail, that the network's three large antennas, in California and Spain and Australia, were committed to some forty missions, including several new ones at Mars and the Moon that had been launched in the previous two years, and that every hour given to New Horizons was an hour taken from someone else. They could give us, they said, perhaps half of the additional time we needed. Not all.
Half the time meant that the spacecraft could send home only about three quarters of the data its instruments were producing. Something would have to be left unsent. The spacecraft could store data for some months, but not indefinitely.
So in the end the question came back to the same place. Not which instrument to turn off, but which instrument's data to leave unsent. And the answer was, as it had been in the document, the dust counter.
*
We turned the transmitter down in November. The dust counter kept running, for another eleven months, and its data was recorded on the spacecraft's memory, and a portion of it, about a third, was sent home in the gaps between the particle instruments' data, when the network could spare the time. The rest was overwritten as the memory filled.
In October of the following year, as the spacecraft approached the termination shock and the particle instruments' data rate increased, the margin of antenna time disappeared altogether, and the counter's data could no longer be sent home at all. I recommended that the counter be turned off. The scientists agreed. I sent the command myself, from the operations room at the laboratory, on a Thursday afternoon. It took about eleven and a half hours to reach the spacecraft. The confirmation that it had been executed took another eleven and a half to come back. I was at home when it arrived, and saw it on my telephone, at about half past five in the morning, a single line in the spacecraft's telemetry indicating that the instrument's power had been switched off.
The spacecraft crossed the termination shock five months later. The particle instruments measured it in great detail. The scientists who led them have written about it at length, and I believe their work will be important for many years.
*
I said at the beginning that I would try to give an honest answer to the question of how it felt to turn off the instrument, and I find, now that I come to it, that the answer is not about the instrument at all.
I did not feel, when I saw the line in the telemetry that morning, very much sorrow. I had expected to. I had expected to think of the basement in Boulder, and the dust accelerator, and Venetia Burney's careful letter, and the twenty of us who had built the counter when we were young. I did think of them, a little. But what I thought of mostly was Walter Pratt, and his notebook, and the three watts of heat that he had explained to a group of students in 2005, and that I had written down, and that had turned out, nearly thirty years later, to matter. I thought about how close I had come to not finding the notebook. I thought about how many other notebooks there must be, in other attics, containing things that would one day matter to someone, and how few of them would be found.
And I thought, in a way I find difficult to put into words, that the counter's last eleven months, kept running by a reduction in the transmitter's power and a few hours of antenna time begged from the schedulers, had been a kind of courtesy. Not to the instrument, which could not feel it, or to the students who built it, most of whom had long since moved on and would never know. A courtesy to the work itself: a refusal to let it stop one day earlier than it had to. I am not sure that this was a scientific decision. I am not sure that it was a sentimental one either. I think it may have been something in between, for which I do not have a good name.
*
The spacecraft is still flying. It is more than a hundred times as far from the Sun as the Earth now, and its power has fallen to a little over a hundred and twenty watts, and one of the particle instruments has since been turned off as well. The team that operates it has shrunk to five. None of them, now that I have retired, worked on the spacecraft before launch. They have Walter Pratt's notebook, which I gave to the laboratory's archive, and a set of notes of my own, which I wrote in my last year and which I hope are some use to them, though I am aware that I have probably failed to write down the things that will turn out, one day, to matter most.
The dust counter is still on the side of the spacecraft, of course, switched off, its film exposed to the dark, still being struck, I suppose, by the occasional grain of dust that nobody will ever count. I think of it sometimes, at night. I find that I do not think of it as dead, exactly. I think of it as finished, which is a different thing, and which is, I have come to believe, the best that most of our work can hope for.