NASA just bought Voyager 2 another year of life without launching a thing. The whole budget is measured in watts, and it shrinks by four every year.

In early August, a small team at NASA's Jet Propulsion Laboratory sent a command toward a spacecraft 13 billion miles away, and then did the hardest part of the job, which was wait. Voyager 2 is far enough out now that a radio signal takes about 20 hours to reach it, and another 20 to come back — close to two days for a round trip. The engineers could not see the spacecraft, could not touch it, could not send a technician or a spare part. They had reasoned their way to a maneuver they were confident would work, nicknamed it the Big Bang, fired it off into the dark, and then sat for the better part of two days to learn whether their reasoning was right. It was. And what they won was not a discovery or a new destination. It was time — at least another year of it. To understand why that is one of the best deals in the entire space program, you have to stop looking at the spacecraft and start looking at its budget.

Because Voyager does not run on money, day to day. It runs on watts. And the whole story of the mission's final act is the story of a budget that shrinks by about four watts every single year and cannot be topped up.

The only line item that matters is power

Voyager 2 launched in August 1977. It is 49 years old, which means it is powered the way you powered things in 1977 if you needed them to last: a radioisotope thermoelectric generator, a box of plutonium that makes electricity from its own decay heat. That was a superb engineering choice and it has a merciless feature. The plutonium decays on a fixed schedule that no ground command can alter, and as it decays it makes less heat, and less heat makes less power. Every year the spacecraft has roughly four fewer watts to work with than the year before. There is no refueling a decaying isotope. The power budget only goes one direction, and everyone on the mission has known for decades exactly which direction that is.

So run the mission the way its operators actually have to: as an accountant running down a fixed, shrinking endowment, deciding each year which line items still earn their keep. Every instrument on Voyager 2 is a draw on that account. Every heater that keeps the electronics from freezing is a draw. The radio that phones home is a draw. When the incoming power falls below what all the draws demand, something has to be switched off, and switched off for good, because there is no turning the sun back up. The engineers at JPL are not really explorers at this point. They are portfolio managers for a declining trust, and the Big Bang was a rebalancing.

The engineers at JPL are not really explorers at this point. They are portfolio managers for a declining trust, and the Big Bang was a rebalancing.

What they actually cut, and why it was clever

Here is the part that makes this a budget story and not a physics one. Without intervention, the mission would have had to turn off another science instrument on Voyager 2 before the end of this year — a real loss, because the spacecraft is already down to just three working instruments. It has been shedding them for a while: the plasma science instrument went dark in late 2024, the low-energy charged particle detector in early 2025. What remains are three sensors that still do genuinely unique work — a magnetometer reading the magnetic field of interstellar space, a plasma-wave instrument listening to the density of the gas between the stars, and a cosmic-ray detector counting the energetic particles out there. Those three are the science. Those three are the reason anyone still pays the bill.

So instead of cutting a fourth instrument, the team went after overhead. The Big Bang shut down a pair of heaters and an onboard digital tape recorder that was no longer recording anything useful — it was being kept powered essentially for the waste heat it threw off — and substituted lower-power ways of keeping the critical electronics warm enough to function in deep space, where the ambient temperature is a problem you have to spend energy to solve. In plain terms: they cut the utility bill instead of the research staff. They found their savings in the parts of the spacecraft that cost watts without producing data, and they spent those recovered watts buying another year for the three instruments that do. That is exactly the move a good operator makes when the top line is fixed and falling: protect the revenue-producing assets, strip the overhead.

Now price the alternative

The instinct with a 49-year-old machine is to ask whether it is worth the trouble. Follow the money and the answer is almost embarrassingly lopsided. The recurring cost of keeping Voyager 2 alive is a modest operations team and a share of time on the Deep Space Network — the handful of giant radio antennas that talk to everything NASA flies beyond the Moon. That antenna time is the real invoice, and it is not nothing: the Deep Space Network is oversubscribed, and every hour a 70-meter dish spends listening to a whisper from interstellar space is an hour it is not spending on a mission at Mars or Jupiter. But it is measured in antenna hours and a small annual budget, not in billions.

Against that, price the replacement. There isn't one. There is no probe funded, built, or on a launch pad that will reach interstellar space to do this work if Voyager stops. The two Voyagers are the only human-made objects out there in the interstellar medium taking measurements, and the earliest any successor could plausibly get there is decades away, after a mission that does not currently exist is conceived, funded, built, launched, and then spends a quarter-century in transit. The capital cost of Voyager was sunk in 1977. Everything it returns now is marginal science on a fully depreciated asset, which is the cheapest kind of science there is. The cost-per-instrument-year of squeezing another year out of the existing spacecraft rounds, against the cost of building a new one, to zero.

That is the whole business case, and it is why the Big Bang was worth a two-day wait and a one-shot command with no undo. You do not spend engineering brilliance extending a dead-end asset. You spend it here because the asset is irreplaceable and the marginal cost of another year is a rounding error. The math does not just close. It closes more decisively than almost anything else in the budget.

The mission ends when the watts do

JPL plans to run the same maneuver on Voyager 1 in the coming months. Voyager 1 is even farther out — the most distant thing we have ever built — and it faces the identical arithmetic, the same four-watts-a-year decline, the same shrinking list of things it can afford to keep switched on. If it works there too, the two spacecraft could keep returning data into the 2030s, one careful watt-saving decision at a time.

It is worth being clear-eyed about how this ends, because the ending is written into the power curve. These missions will not stop because the science runs out or because the instruments fail. They will stop because the plutonium finally cannot make enough electricity to keep even a single sensor and a radio alive at the same time, and no maneuver, however clever, can recover a watt that the physics no longer provides. The Big Bang did not change that destination. It bought distance to it. Every year JPL wrings out of these spacecraft is a year of data from a place nothing else can reach, purchased for the price of an operations team and some antenna time, against a replacement cost that is effectively infinite. When the accountants and the engineers agree that completely, you take the deal. NASA took it, waited two days in the dark to make sure it had worked, and went looking for the next four watts.