The spacecraft that is running out of power but not out of fuel
Voyager 1 still carries 67.8 per cent of the plutonium it launched with in 1977. It has about 45.3 per cent of the electricity. Radioactive decay accounts for barely half of what went missing, and the rest of the story is the part nobody teaches.
On 18 November 2026, Voyager 1 passes a mark no machine built by anyone has passed before: one light-day from Earth. That is 25.902 billion kilometres, or 173.1 astronomical units, and it is defined by the clock rather than the ruler. A command leaves the Deep Space Network and arrives a day later. The spacecraft answers, and the answer lands a day after that. Engineers describe the working rhythm plainly.
you send your message and then you go home
That is Suzanne Dodd, the Voyager project manager at NASA's Jet Propulsion Laboratory, speaking in October 2026. It took the spacecraft 17,971 days to cover the distance light covers in one. Light is about 17,600 times quicker, and at 16.99 kilometres per second the second light-day will take another 48.3 years, arriving in about 2075.
The distance is the headline. The power is the problem.
Two-thirds of the fuel, under half of the electricity
Voyager carries no solar panels. At 172.4 astronomical units from the Sun, sunlight is roughly thirty thousand times fainter than it is at Earth, so a panel would be useless. Instead, three radioisotope thermoelectric generators turn the heat of decaying plutonium-238 into electricity. At launch they produced about 470 watts between them.
Plutonium-238 has a half-life of 87.7 years. That is the number every textbook reaches for, and it does exactly one job: it tells you how much fuel is left. After 49.2 years the answer is 0.5 raised to the power of 49.2 divided by 87.7, which is 0.678. Just over two-thirds of the original plutonium is still sitting in those generators, still warm, still decaying.
So the obvious prediction is 470 watts times 0.678, or 319 watts. The reported figure is nothing like it. NASA published 225 watts for 2023, and at the agency's own stated decline of about four watts a year that puts late 2026 at roughly 213 watts. That is 45.3 per cent of the launch power against 67.8 per cent of the launch fuel, and a gap of 106 watts between the prediction and the spacecraft.
Put the subtraction side by side and the reframing is hard to miss. Voyager has lost 257 watts since 1977. Radioactive decay explains 151 of them, which is 59 per cent. The remaining 41 per cent went somewhere else entirely.
Where the other 106 watts went
A radioisotope generator is two machines in one casing. The first is a heat source, and its behaviour is governed by the half-life. The second is a converter: banks of thermocouples that make electricity from the temperature difference between the hot fuel and the cold of space. Thermocouples degrade. They also deliver less for every watt of heat as the heat source cools, because the temperature difference across them narrows. Both effects compound, and neither appears anywhere in the half-life.
The result is a power supply that behaves as though its half-life were about 43.1 years, roughly half the plutonium's. Run the arithmetic the other way and the point lands harder still: if the electricity had followed the fuel alone, Voyager would not have dropped to today's 213 watts until 2077. Its power supply is running about 51 years ahead of its fuel.
The familiar rule of thumb fails in the opposite direction. Four watts a year is accurate for recent years, but run it back to launch and it predicts 273 watts today, 60 watts too high. The true average is 5.2 watts a year: about 6.6 before 2001, about 3.9 since. A linear rule is a good description of a curve over a short stretch and a poor one over a long one, which is the whole lesson of the graph.
Try it yourself
Set the years since launch and the reported power, and the calculator below does the comparison that the news coverage leaves out. The defaults are Voyager 1 in November 2026.
| In the year 2026 | Value |
|---|---|
| Fuel still in the generators | 67.8% |
| Electrical power still available | 45.3% |
| Fuel-only prediction | 319 W |
| Watts lost since launch | 257 W |
| Watts the decay explains | 151 W |
| Gap between prediction and spacecraft | 106 W |
| Share of the loss explained by the fuel | 59% |
| Share left to the converter | 41% |
| Effective half-life of the power | 43.1 years |
| Year the fuel alone would reach this power | 2077 |
| How far the power runs ahead of the fuel | 51 years |
The launch power, the 2023 reading and the half-life are reported figures. The present-day power is an estimate built from NASA's own four-watts-a-year decline. Push it to 205 or 225 watts and the share explained by decay moves only between 57 and 62 per cent, so the conclusion survives the uncertainty.
What a shrinking budget actually costs
Eight of Voyager 1's ten instruments are now switched off. The most recent shutdown, in April 2026, left two running. In September 2026 the team went further with a manoeuvre they nicknamed the Big Bang: three devices that also help keep the spacecraft warm were switched off and replaced by three lower-power alternatives, all in one go, because changing them one at a time would have let part of the spacecraft get too cold. Without it, NASA says, another instrument would have gone dark as early as the summer of 2027.
Six weeks later, on 18 and 19 October, the Deep Space Network could not find the signal. It returned on 20 October. JPL's preliminary finding is that one group of thrusters went offline and the backups took over; the root cause was still open. The timing is suggestive and proves nothing, which is a more useful thing for a student to sit with than a tidy answer.
Every one of those decisions is made through a two-day loop. A command sent today is confirmed the day after tomorrow. At 160 bits per second the spacecraft can return about 1.73 megabytes a day, so a single three-megabyte phone photograph would occupy the link for 41.7 hours. Nobody can supervise a fault at that distance, which is why the spacecraft has to be able to protect itself and then wait.
The classroom packet
Voyager Power & Half-Life Math
A 90-minute case study for grades 10–12, college and adult education. Students build the exponential decay model everyone expects to work, test it against three published NASA figures, and find the 106-watt gap for themselves. Then they budget the watts, design backwards from a deadline, and write a recommendation about a repair whose result takes two days to come back.
- Lead standard CCSS HSF-LE.A.1: a linear rule of thumb and an exponential law held against the same data.
- 22 pages — 11-page teacher guide, 10-page student packet, cover.
- Full answer key with every step worked, plus a rubric for the written recommendation.
- One 90-minute block, two 45-minute periods, or a documented 50-minute short path.
- Print and go. A scientific calculator is the only equipment.
Left: the classroom packet on Teachers Pay Teachers. Right: Jim Bell’s account of the forty-year Voyager mission, told through the NASA men and women who flew it — the people who make the shutdown decisions this article is about — an Amazon affiliate link. As an Amazon Associate this site earns from qualifying purchases, at no additional cost to you.
Sources
The light-day date and distance, the 38,000 mph speed, the September swap, the October signal loss and the quotation come from JPL news release 2026-078, NASA's Voyager 1 Conserves Power, Resolves Brief Communications Pause, 29 October 2026. The instrument shutdown, the 23-hour command time and the 0.5 watt scan motor come from the NASA Science Voyager blog, 17 April 2026; the Voyager 2 swap and the phrase “at least an extra year” from the same blog, 4 August 2026. The 25.902 billion kilometre figure is NASA's live tracker; 225 watts as of 2023, and the 16 and 160 bits per second data rates, are from NASA Science, Voyager: The Spacecraft; about 158 watts per generator at launch is from NASA's Voyager 1 mission page. The 315 watts for early 2001 is reported through a secondary compilation of JPL weekly status reports and agrees with the four-watts-a-year rule run back from 2023. News coverage: David Dickinson, Universe Today, 6 November 2026. Estimated, not measured: the present-day figure of about 213 watts, and therefore every percentage derived from it. Modelled, not measured: the heliocentric distance of 172.4 astronomical units, from a simple circular-orbit geometry rather than an ephemeris. Every calculation above and in the classroom packet was run in code before publication.