Lunar Power

South-polar solar

Solar Power at the Lunar South Pole: Why Illumination Is Not the Same as Power Density

The rim is sunlit. The sun angle is 1.5–6 degrees above the horizon. Those are two different statements.

The south-polar rim advantage is real: selected ridgeline points on Shackleton crater and the de Gerlache ridge see sunlight more than 90% of the year, compared with roughly 50% at equatorial sites. The 14-day equatorial night — the constraint that dominates every power architecture at lower latitudes — largely disappears at the poles. Dark intervals are hours to days rather than two weeks. That is the single strongest site-selection argument for the poles, as covered on the Shackleton & the South Pole page.

But illumination time is not power density. The Moon’s 1.54° axial tilt means sunlight at the poles arrives at a grazing angle — typically 1.5° to 6° above the horizon, depending on the lunar month and the elevation of the array site. At a 3° elevation angle, a horizontal solar panel receives only about 5% of the irradiance (1,361 W/m²) it would get under a directly overhead sun. A panel optimized for equatorial noon is nearly useless at the poles.

The solution is to turn the panels sideways. A vertical array facing the sun at a 3° elevation angle intercepts the full cross-section of the beam. At the lunar south pole, vertical is what horizontal is everywhere else. This is not a clever engineering trick. It is the baseline architecture — every serious lunar solar design for the poles uses vertical arrays, and NASA has funded five companies to develop them. Building

Vertical solar arrays

Vertical Solar Array Technology: Lockheed Martin, Astrobotic, and Honeybee Robotics

Up to 65 feet tall, autonomously deployable, retractable for relocation, and self-leveling on uneven terrain.

NASA’s LVSAT (Lunar Vertical Solar Array Technology) program, under the Space Technology Mission Directorate, awarded contracts to five companies to design deployable vertical solar arrays for the south pole. The designs share common requirements: autonomous deployment up to roughly 10–20 meters (32–65 feet), retraction for relocation, self-leveling on slopes, and survivability under regolith dust exposure and thermal cycling.

Lockheed Martin’s prototype completed testing at its Littleton, Colorado facility. It demonstrated autonomous deployment and retraction on uneven surfaces, self-leveling on slopes, and survivability under regolith-dust exposure — the dust-on-optics problem that degrades performance over time. The vertical orientation has a secondary benefit: dust settles less on a vertical surface than on a horizontal one, reducing the accumulation rate that has historically degraded solar arrays on the Moon and Mars. Building

LunaGrid and the power network

Astrobotic LunaGrid: A Cabled Power Network for the South Pole

Not one array. A network of nodes connected by transmission cables and mobile charging rovers.

Astrobotic’s LunaGrid is designed as a network of solar-generating nodes connected by power-transmission cables up to 2 km long, with mobile CubeRovers carrying wireless chargers to extend power to assets beyond the cable reach. The first node lands on a Griffin lander and deploys a VSAT to begin generating power. Subsequent missions add nodes, each connected to the grid. By 2028, Astrobotic plans a full operational grid at the south pole that can serve the Artemis base or any other surface asset.

The architecture is significant because it separates power generation from power consumption. A habitat does not need its own array if it can draw from the grid. An ISRU plant 2 km away on an inner crater wall does not need a collocated reactor if a cable reaches it. The grid turns power from a per-installation problem into a shared utility — the same architectural transition that electrification brought to terrestrial industry in the early 20th century. Building

Gap-bridging storage

Surviving the Dark Hours: Batteries, Fuel Cells, and the Gap the Poles Shrank

The pole did not eliminate the night. It shrank it from 354 hours to hours or days.

Even at the best south-polar rim sites, illumination is not 100%. Short dark intervals of hours to a few days occur when terrain features shadow the array or when the solar geometry briefly dips below the horizon. The storage system must bridge these gaps without carrying the mass required for a full 14-day equatorial night.

Two technologies compete for gap-bridging:

Regenerative fuel cells (RFCs). During sunlit periods, excess solar power electrolyzes water into hydrogen and oxygen, which are stored as gases. During dark intervals, the fuel cell recombines them to generate electricity and produces water as a byproduct, which is recycled. RFCs have higher specific energy (Wh/kg) than batteries for long-duration storage because the energy is stored in the reactant tanks, which can be sized independently of the cell stack. NASA’s baseline surface architecture uses RFCs for multi-day gap bridging.

Batteries. Lithium-ion batteries are simpler, lighter for short gaps (hours), and more efficient in charge/discharge cycling. They degrade with thermal cycling and radiation exposure, but at the moderate south-polar thermal environment (~170–220 K) degradation rates are slower than at equatorial sites. Batteries and RFCs are complementary: batteries for short gaps (hours), RFCs for longer ones (days).

The mass savings from a polar site are dramatic. An equatorial base operating at 40 kW needs storage sized for 354 hours of continuous night — roughly 14,000 kWh at 40 kW average load. At a south-polar site with 48-hour maximum dark intervals, the same base needs roughly 1,920 kWh — a factor of seven less. The poles did not solve the power-storage problem. They shrank it by an order of magnitude.

Fission surface power

Nuclear Fission Reactors on the Moon: 40–100 kW by 2030

Power that does not care whether the sun is up, whether you are at the pole, or whether you are on the crater floor.

A White House Executive Order in early 2025 directed NASA and the Department of Energy to develop a nuclear reactor ready for lunar launch by 2030. The current program targets a 100 kWe fission surface power (FSP) system using a closed Brayton cycle power conversion system, replacing an earlier 40 kWe / 6-tonne design. Three teams — Lockheed Martin, Westinghouse, and a third contractor — received roughly $5M each for Phase 1 concept designs in 2022. The revised program halts the earlier timeline (which had hoped for a 2026 launch) and streamlines NASA’s process to support commercial builders at higher power levels.

Fission power has one overwhelming advantage over solar: it works everywhere. At the equator, it powers through the 14-day night. At the poles, it powers the permanently shadowed crater floor. On the far side, it operates without Earth line of sight for solar-array tracking telemetry. It does not degrade with dust accumulation. It does not need vertical deployment. It scales independently of site illumination geometry.

The trade-offs are mass, regulatory complexity (launching enriched uranium requires interagency and international review), thermal management (the reactor must reject waste heat to the vacuum via radiators), and the fact that it is not yet built. The KRUSTY experiment — NASA’s Kilopower Reactor Using Stirling Technology — demonstrated a 1 kWe fission system using a highly enriched uranium core, a sodium heat-pipe cooling system, and Stirling engines in 2018 at the Nevada National Security Site. Scaling from 1 kWe to 100 kWe is not just making it bigger — it requires different reactor geometries, different conversion systems, and significantly more shielding. Building

Power for the shadowed floor

Powering the Permanently Shadowed Crater Floor: Cable, Reactor, or Both

Every watt consumed below the rim must be generated above it — or generated locally by a reactor that changes the thermal environment.

The rim-to-floor corridor is where the power architecture meets the site geography. ISRU extraction and cold compute operations on the shadowed floor need power, but sunlight does not reach them. Two paths:

Cabled power from the rim. Solar arrays on the illuminated ridge generate power, and a high-voltage DC cable (5–10 kV) runs down the slope 10–20 km to the work zone on the crater floor or inner wall. The cable must survive the thermal gradient between the ~200–300 K rim and the ~40–88 K floor, and it must cross the terminator dust-levitation zone. Conductor mass scales with distance, voltage, and power level. At 250 kW delivered over 15 km at 10 kV DC, cable mass is on the order of several hundred kilograms — within the payload capacity of a cargo lander. The challenge is not the cable. It is installing, anchoring, and maintaining it across geologically active terrain in vacuum.

Local fission reactor on the floor. A reactor placed near the extraction zone eliminates the cable and the rim-to-floor transmission losses. But a 100 kW reactor operating on a ∼40 K crater floor radiates waste heat into an environment that has been thermally stable for two billion years. The thermal budget constraint applies: a reactor placed on the deepest floor could warm the surrounding cold trap and destroy the volatile inventory it was deployed to exploit.

The likely architecture uses both: cabled solar power from the rim for the primary work zone at 60–90 K on the inner wall, and a small reactor as backup and night-gap coverage, placed away from the coldest traps. Projected

The demand gap

10–40 kW vs. 490 kW: The Power Gap Between Outpost and Factory

NASA is designing for a research station. The industrial question asks what happens when the station becomes a factory.

NASA’s current Artemis surface-power baseline is roughly 10–40 kW — sized for a research outpost: lighting, life support, communications, science instruments, a small rover fleet, and initial ISRU demonstration hardware. The fission surface power program targets 40–100 kW in the same range.

The industrial architecture analyzed on this site needs on the order of 490 kW: roughly 250 kW for ISRU resource processing (excavation, thermal extraction, electrolysis, gas handling), 100 kW for habitat and life support at crew scale, 80 kW for surface mobility and logistics, and 60 kW for compute, communications, and base systems.

That is a factor of 12–50 above the current Artemis baseline. The gap is not an error. NASA is sizing a research outpost. The industrial question — the one that makes the Moon a supplier — begins where the outpost power budget ends. Bridging it requires either many vertical solar arrays operating as a grid (Astrobotic’s LunaGrid model), multiple fission reactors, or both. The mass driver alone — a 3.2-km electromagnetic launch track — is a multi-megawatt system. The industrial Moon is a power-infrastructure problem before it is anything else. Projected

Record

Provenance and Data Sources