There's a number that stops engineers cold when they first encounter it: one kilowatt. That's roughly what the James Webb Space Telescope uses to run everything — four science instruments, communications, propulsion, thermal management, and onboard computers, all operating simultaneously from a halo orbit 1.5 million kilometers from Earth. Less than a household kettle. Less than a microwave.
That figure is usually presented as a fun fact. It's actually a design thesis — and understanding it reveals something important about how Webb's engineers thought about power redundancy in a mission where "send a repair crew" was never on the options list.
The Solar Array Margin Is the Redundancy
Webb's solar array is sized at close to two kilowatts — roughly double the observatory's operating load. That gap isn't waste. It's the power system's primary redundancy strategy.
Solar cells degrade. Radiation exposure, thermal cycling, micrometeorite impacts — all of it chips away at generation capacity over time. By building in a two-to-one margin between what the array can produce and what the spacecraft needs, engineers created a buffer that absorbs years of degradation without requiring any active intervention. No backup array. No battery-switching logic under normal operations. Just headroom, baked in at the design stage.
This is a fundamentally different philosophy than, say, a crewed spacecraft where you can add redundant power buses and have astronauts manually reconfigure systems if something trips. Webb's location at L2 made servicing impractical in a way that Hubble's low Earth orbit never was. The engineering response wasn't to add more redundant hardware — it was to make the primary system robust enough that redundancy rarely needs to activate.
The passive cooling architecture follows the same logic. The sunshield blocks heat from the Sun, Earth, and Moon, cooling the telescope's cold side without drawing significant power. An active refrigeration system would require power, moving parts, and failure modes. The sunshield requires none of those things once deployed — it just works, continuously, as long as it's pointed correctly. Passive systems don't fail the way active ones do.
The 344 Problem: When Redundancy Isn't Available
Here's where the power story gets complicated. That elegant margin-based approach only works if the solar array deploys in the first place.
Webb launched with 344 single-point failures on its books — components or steps where one mistake compromises the mission. Roughly 80 percent of those were tied to the post-launch deployment sequence. The solar array was first in that sequence — Webb needed it to deploy within minutes of separation from the Ariane 5 to have power at all.
You can't redundancy-engineer a deployment sequence the same way you can a power bus. Adding backup hinges and redundant release mechanisms adds mass and complexity, which creates new failure modes. The engineering answer here was different: exhaustive testing, incremental verification, and a design philosophy that treated each mechanism as something that had to work correctly rather than something that could fail gracefully.
Northrop Grumman's approach was to break the problem into unit-level and subsystem tests, building up an analytical model piece by piece because no facility existed large enough to test the full observatory under space conditions. The redundancy, in this case, lived in the knowledge base — the confidence built through years of testing that each mechanism would behave as modeled.
What the Extended Mission Actually Inherits
Webb is now operating well into its extended mission phase, and the power system's design choices are paying off in a specific way: the observatory is doing science that its designers only partially anticipated. NASA's Pandora mission launched specifically to complement Webb's exoplanet observations, addressing stellar variability problems that weren't fully understood until around 2017 or 2018. Webb's instruments are being pushed into observation modes and target lists that evolved after launch.
That flexibility exists partly because the power margin held. The two-kilowatt array degrading toward a one-kilowatt load over years means the spacecraft still has room to operate all four instruments, run the fine guidance sensor, and maintain thermal control without making hard trade-offs between subsystems. The margin that looked like over-engineering in 2021 is operational flexibility in 2026.
The deeper lesson isn't about solar arrays specifically. It's about where you put your redundancy budget when you can't service what you've built. Webb's engineers chose passive over active wherever possible, margin over backup systems, and exhaustive pre-launch verification over in-flight reconfiguration. Every one of those choices was a bet that the mission would outlast its original design life — and so far, that bet is paying off in rock-cloud discoveries 700 light-years away.
