The frame
A Flow With a Buffer
Energy activates every process on a site, human and industrial. What sets it apart among the three substrates is that it can be banked - so the discipline is buffering.
No process runs without power, so energy is produced and drawn down hour by hour like the other two flows. It differs in one respect that governs how it is run: a kilowatt-hour can be stored. Batteries hold it for minutes to hours; regenerative fuel cells, splitting and recombining water, hold it across the two-week lunar night. That buffer is what lets generation and demand run on different clocks, and managing it is the whole discipline - sizing the store against the longest dark, the deepest shadow, and the worst dust.
Energy moves through a four-part chain - generation, storage, distribution, and rejection - and the last part is the one most accounts drop. Every watt used becomes a watt of heat that has to be radiated away, and the radiator is sized in square meters and heat-pipe meters, not watts. The portable figure the whole model runs on is 0.37 t per kWe: each kilowatt of installed capacity costs roughly 0.37 tonnes of landed energy-chain mass once generation, storage, distribution, and rejection are counted. Building
Solar first
Solar Goes First, and Not as a Hedge
Cheap, fast, and dependent on nothing outside the program, solar stands a site up before a reactor can arrive, deploy, and be verified.
Solar is the first power everywhere in the inner system, and the reason is schedule as much as mass. On mass, solar with a regenerative fuel cell lands at about 17 kg/kWe at the lunar pole against roughly 60 for the 100 kWe reactor class; it covers 96.4% of a lunar outpost’s energy and 100% of its firm early block with no reactor at all. On schedule, every input a reactor needs sits on a clock the program does not control - one Western enrichment cascade turning out a few hundred kilograms a year, plutonium-238 produced at about half a kilogram a year against a target near 1.5, and launch approval on its own regulatory track, with the first flight article rated at 40 kWe rather than 100.
Solar is fast because nothing outside the program has to agree. That one sentence is the commissioning ruling: a site is stood up on arrays and fuel cells, and fission is brought in later for the jobs arrays cannot do. Observed on the mass and schedule figures; Building on the reactor timeline.
The crossover
The Crossover Is an Area, Not a Distance
A reactor becomes load-bearing at a size, not at a range. Mars is the exception, and the reason is weather.
The common picture puts a line somewhere past Mars where solar quits and fission takes over. The real crossover is set by area, not distance. A reactor becomes the enabling condition at roughly 2 to 3 MW, the point where the solar array outgrows the usable sunlit ridge and the regenerative-fuel-cell inventory needed to cross the dark passes about 100 tonnes. Below that, arrays win on mass and schedule; above it, the array stops fitting on the terrain and the stored reactant mass becomes the binding cost. Where a reactor is justified earlier, the reason is redundancy, shadow, plume tolerance, and room to grow - not energy per kilogram.
Mars breaks the pattern, and distance is not why. Planet-wide dust storms cut insolation for weeks and coat every panel, so a Mars site wants firm power earlier than its sunlight budget alone would suggest. The frost line past which solar collapses outright is real, but it is a working boundary that moves with reactor specific power and array packing, not a fixed wall. Building
The chain
Generation, Storage, Distribution, Rejection
The buffer and the water inventory are frequently the same tank, which concentrates both the leverage and the risk.
Storage is where energy and water meet. A regenerative fuel cell banks energy by splitting water into hydrogen and oxygen and recovers it by recombining them, so the energy buffer is literally a volume of water held back from every other claimant. At a plant producing 230 tonnes of water a year, the reactant charge is about 4.3 tonnes - seven days of production parked as inventory. The store is the product.
Holding the buffer as water concentrates five failure modes - freezing, leakage, electrolyzer loss, boil-off, and contamination - in one substance. The mitigation is to segregate the inventory into independent volumes, not to diversify the substance; there is no second working fluid a lunar site can make.
Closure
The Closure Ladder Is an Energy Ladder
Sort every manufacturing step a site could close by the energy its transformation costs, and the build order falls out on its own.
Local production closes in an order set by specific energy: bulk emplacement near 0.001 kWh/kg, salvage around 0.04, structure about 0.5, water 1 to 4 by grade, oxygen near 5 to 7, metals 20 to 40, and finished devices between 103 and 105. That is seven orders of magnitude from the bottom rung to the top, and each rung waits for the power plant to grow by roughly a factor of ten. The closure schedule is therefore an output of the energy buildout, not an input to it - the link between power and what a site can make runs the direction most plans assume it does not. The ladder itself lives in Materials; energy is what paces it. Building
Contested
Deliberately Not Claimed
The load-bearing uncertainties, tagged with the confidence they earn.
Solar first is settled for the early and uncrewed phases. The mass advantage over the reactor class and the no-external-agreement schedule argument both hold on measured hardware; arrays and fuel cells are flight-proven.
The 2–3 MW crossover is a modeled boundary. It moves with reactor specific power and with how densely an array can be packed on a given ridge, so the exact megawattage at which fission becomes load-bearing is site-specific.
The reactor supply timeline rests on external clocks. HALEU enrichment throughput, plutonium-238 output, and launch approval are each outside any space program’s control, and a slip in any one moves when fission can actually arrive.
Record
Provenance and Data Sources
- Confidence
- Observed for array and fuel-cell mass and the 0.37 t/kWe factor - Building for the crossover size and frost-line location - Projected for the reactor supply timeline
- Status date
- Q4 2026
- Discipline
- Buffering - a flow with a store, sized against the longest dark, the deepest shadow, and the worst dust
- Key figures
- Solar + RFC ~17 kg/kWe at the lunar pole vs ~60 for the 100 kWe reactor class; 96.4% of lunar outpost energy on solar; crossover at 2–3 MW; 0.37 t per kWe landed
- Related
- Water is the fuel-cell working fluid; Materials carries the closure ladder energy paces; Substrates for the three-flow frame
- Weakest link
- The reactor supply timeline and the Mars dust-storm firm-power case - both move when a reactor is actually needed