ENERA — Power for a Multi-Planetary Civilization

Energy is the precondition for everything else. The food architecture, the cognitive substrate, the coordination layer, the off-Earth presence, the climate transition — none of them are bounded by intelligence or by capital so much as by the kilowatt-hours available at the right places at the right cost. The energy questions split cleanly into three. The first is the technological question: which energy architectures are physically and economically feasible at planetary and multi-planetary scale. The second is the deployment question: how the architectures are built out, where, by whom, and against which alternatives. The third is the political-economy question: who owns the architecture, who pays, and how the surplus is distributed. ENERA treats the first two as the program scope and the third as a constraint on what the first two are allowed to do. The literature has frequently conflated these, and the conflation has produced confused arguments on all sides.

The three questions are different

The phrase “the energy transition” gets used to mean at least three distinct things, and the conflation has cost the field a decade of clarity. The first is the technology claim: that solar PV, wind, batteries, electrolyzers, fission, and (eventually) fusion have crossed cost-and-performance thresholds that make a fully-electrified, post-carbon energy system technically feasible at terrestrial scale. The second is the deployment claim: that the rate at which we are actually building the technology — the rate at which existing fossil capacity is retired, the rate at which transmission is built, the rate at which storage is sited — is the binding constraint on the transition, not the underlying technology. The third is the off-Earth claim: that durable human presence beyond Earth requires energy architectures that Earth has not yet had to build, because the Earth surface is a uniquely benign energy environment that hides constraints which Mars, the Moon, and deep space do not hide.

The three claims are independent. A society can succeed at the technology claim and fail at the deployment claim; this is the present state of the global energy transition, where solar PV and battery storage have undercut the cost of fossil-fueled new build but the deployment rate remains substantially behind what stabilization at 1.5–2 °C would require. A society can succeed at deployment and fail at the off-Earth claim; this is what we currently expect to happen by default, with a fully electrified Earth-side energy system that does not project to Mars or the lunar polar craters or deep space. And a program can be substantively about all three, but the engineering, the cost models, and the failure modes are different across each, and treating them as a single problem produces the wrong design choices.

The most-cited starting point for the off-Earth conversation is Peter Glaser’s 1968 Science paper “Power from the Sun: Its Future,” which proposed orbital solar arrays beaming microwave power to Earth from geostationary altitude, and which has remained the canonical architectural reference for space-based solar power for nearly six decades.¹ Glaser was right about the physics. The economics, in 1968, did not work — the launch cost of a gigawatt-scale solar power satellite at Saturn V or Shuttle launch prices exceeded the entire global energy market — and SBSP became, fairly, a perpetual-twenty-years-away technology for thirty years. What has changed in the last five years is not the physics. The physics has been settled since Glaser. What has changed is that the launch cost, the phased-array hardware, the in-space-assembly tooling, and the photovoltaic mass-per-watt have all crossed credibility thresholds simultaneously. SBSP is no longer a thought experiment; it is a deployed-flight-experiment program with results from Caltech, JAXA, AFRL, and shortly the Chinese space program.

So the three questions need to be held apart. The terrestrial transition is one program, the off-Earth power architecture is a second, and the integration question — how the two systems compose into a single planet-and-beyond energy system — is a third. ENERA’s research scope is the off-Earth architecture and the integration. The terrestrial transition is the underlying market and policy environment that determines what cost curves are available and what political-economic constraints the off-Earth architecture has to live within. The remainder of this page works through the architecture in this order.

The cost curves we are downstream of

Before the off-Earth architecture, the terrestrial cost environment. ENERA is downstream of the cost-curve trajectory of utility-scale solar PV, wind, batteries, electrolyzers, and reusable launch. The cost trajectory has been the most consequential phenomenon in the global energy system over the last fifteen years and the trajectory determines which off-Earth architectures are economically viable in the next decade.

Utility-scale solar PV reached an unsubsidized levelized cost of energy in the range of $36–60/MWh by the early 2020s, and despite a partial reversal in 2022–2024 driven by interest rates and supply-chain inflation, the long-run trend has continued downward.[^2] Onshore wind has been in a similar range. The standard reference framework — Lazard's annual *Levelized Cost of Energy+* analysis — places solar PV and onshore wind as the lowest-cost new-build power generation for the tenth consecutive year as of the 2024 analysis.[^2] The 2024 LCOE for utility-scale solar PV was reported at$61/MWh; for onshore wind, $50/MWh; for combined-cycle gas turbines (the next-cheapest source),$76/MWh. The implication for the off-Earth architecture is that any space-based power system has to compete, on terrestrial markets, against a moving target whose floor is approximately $30/MWh and whose primary disadvantage (intermittency) is being eroded by parallel cost reductions in lithium-iron-phosphate stationary storage. SBSP has to clear$50–80/MWh delivered, including the rectenna real-estate and the transmission interconnect, to be competitive against terrestrial alternatives. This is not trivial; it is the binding constraint on the SBSP business case.

The launch-cost cost curve is the second curve. Falcon 9 reusable, the dominant commercial heavy-lift launcher, operates at approximately $1,500–2,500/kg to LEO at marginal cost. Falcon Heavy reaches approximately$1,500/kg. Starship, currently in flight test and projected to reach full reusability over the next several years, has a published nominal target of $13–30/kg to LEO at full payload-and-cadence ramp, against a baseline of approximately 100–150 tonnes per launch at marginal launch cost on the order of$2 million. These are the figures SpaceX has stated; the figures industry-external observers will accept depend on Starship’s flight-rate ramp, which is the load-bearing variable. At $1,500/kg, SBSP gigawatt-scale stations are competitive with terrestrial nuclear new-build; at$200/kg, they undercut everything; at $30/kg, the Earth-side energy system becomes a launch-and-rectenna system within a generation. The ENERA program does not assume the lower end of the Starship curve. It assumes the upper end and reports the sensitivity.

The electrolyzer and fuel-cell cost curves are the third curve. Green-hydrogen production via PEM and alkaline electrolysis has fallen from approximately $5–8/kg of H₂ in 2020 to$3–5/kg in 2024, with continued projections of sub-$2/kg by 2030 in regions with cheap renewable electricity. This determines the cost of synthetic methanol, of which ARCANE is one downstream consumer; it also determines the cost of hydrogen-based long-duration energy storage, which is the alternative architecture to space-based solar for terrestrial firm-power applications. Green hydrogen at$1–2/kg makes SBSP a niche; green hydrogen above $4/kg, indefinitely, makes SBSP the lowest-cost dispatchable-power option for medium-density urban demand. The ENERA program treats the hydrogen cost curve as the principal terrestrial competitor and tracks it accordingly.

What ENERA is, technically

The architectural building blocks of an off-Earth-and-deep-Earth power system are not exotic. They are well-characterized. The integration is the open frontier, and it is what the ENERA research program pursues. The technical state of each building block is summarized below, with citations.

Space-based solar power and the in-orbit photon

Solar irradiance in space is a constant 1,361 W/m² at Earth distance — the “solar constant” measured to high precision by the SORCE / TIM instruments and reconstructed from historical data by Kopp and Lean.² On Earth’s surface, peak irradiance under clear-sky conditions at noon is approximately 1,000 W/m²; the 24-hour, all-weather average at populated mid-latitudes is closer to 200 W/m². A solar collector in geostationary orbit therefore receives roughly eight times more useful energy per unit area per year than an equivalent ground-based collector, with no diurnal cycle, no seasonal variation in the populated-latitude sense, and no atmospheric attenuation. This is the fundamental energy advantage of space-based solar; it has been understood since Glaser’s 1968 paper and has not been challenged since.¹

The architectural choices within SBSP fall into a few classes. The classical Glaser-era design uses concentrator-photovoltaic or thermal-engine collectors at geostationary altitude (35,786 km) coupled to large microwave transmitters beaming to ground rectennas of order 5–10 km in diameter. The CASSIOPeiA architecture, developed by Ian Cash and presented to the International Astronautical Federation, retains the geostationary orbit but uses a heliogyro-style rotating collector with retrodirective sandwich-tile microwave transmitters, eliminating the need for slip-ring power transmission between rotating and pointed parts of the structure. The MR-SPS architecture (modular, rotational, scalable) and the tethered-bolometer SPS architectures are alternative system topologies. None of these architectures has been built at scale; the architectural debate is, in the current period, being settled empirically by the demonstrator missions described below.

The defining engineering challenge of any SBSP architecture is structural: the solar-collection area for a gigawatt-class system is on the order of square kilometres, and the microwave transmitter diameter required to focus the beam onto a Earth-side rectenna is on the order of one kilometre. No structure of that size has ever been built in space, and the in-space assembly capability required is a research frontier in itself. Recent advances in deployable composite structures (DOLCE, deployed by Caltech in 2023; large-area phased-array deployments by AST SpaceMobile and Lynk in the LEO communications context) have made the structural problem tractable but unsolved at the scales SBSP requires. The ENERA program funds work specifically on the assembly question, which is the binding constraint on first-deployment timelines.

Microwave wireless power transmission

Microwave power transmission uses the same physics as a radar or satellite communication link, optimized for power density rather than information density. A phased array of solid-state amplifiers generates microwaves; the beam propagates through atmosphere with very low loss at carefully-chosen frequencies (2.45 GHz, 5.8 GHz, 10 GHz) where atmospheric absorption is minimal; the beam is received by a rectenna — a rectifying antenna array — which converts the microwave RF energy directly back into DC electricity at high efficiency.

The relevant efficiencies are characterized. DC-to-RF conversion at the transmitter, using modern gallium-nitride solid-state amplifiers, achieves 70–85%; magnetron tubes can exceed 90% but are heavier and less controllable.³ Free-space propagation through the atmosphere at 2.45 GHz is essentially lossless, even through clouds and rain; at 5.8 GHz, losses remain small; at 10 GHz, losses are characterizable but small under most conditions; at 24 GHz and above, rain attenuation becomes significant. RF-to-DC conversion at the rectenna, using GaAs Schottky-diode rectifiers, has demonstrated up to 86% in laboratory settings. The current outdoor terrestrial benchmark is the 73% RF-to-DC conversion ratio achieved by the U.S. Naval Research Laboratory’s Safe and COntinuous Power bEaming – Microwave (SCOPE-M) program in 2022, which transmitted 1.6 kilowatts of power over 1 kilometre at the U.S. Army Research Field in Blossom Point, Maryland — the most significant terrestrial microwave-power-beaming demonstration in nearly fifty years, led by Christopher Rodenbeck with Paul Jaffe as the power-beaming lead.³ End-to-end efficiency of an integrated SBSP system, including PV conversion losses, currently targets 8–15% sun-to-grid; theoretical maxima with mature technology are in the 25–35% range.

The defining engineering challenge of microwave power transmission is the phased-array synchronization problem at scale. A microwave beam from geostationary orbit to a kilometre-scale rectenna requires sub-degree phase synchronization across tens of thousands of free-flying transmitter elements over a kilometre-scale aperture. Pilot-tone retrodirective architectures, in which the ground rectenna emits a low-power synchronization signal that the orbital array uses to phase-lock its transmitters, are the current preferred architecture; the approach has been demonstrated at hundred-element scales and is unbuilt at kilometre scale. The ENERA program funds work specifically on retrodirective synchronization at scale, with the in-space demonstration of kilometre-scale phasing as the key milestone.

Laser wireless power transmission

Laser power transmission is the alternative architecture to microwave. A near-infrared laser at an eye-safe wavelength (around 1550 nm, where the atmospheric water-vapour absorption window aligns with the eye-safety regulatory framework) is collimated into a tight beam, propagates to a photovoltaic ground receiver at substantially smaller diameter than the equivalent microwave rectenna (5–10 m versus kilometres), and is converted to DC by a wavelength-tuned PV array. The advantages over microwave are smaller receiver footprint and substantially less spectrum-coordination work. The disadvantages are weather sensitivity (clouds essentially block the beam), pointing precision (sub-arc-second pointing required at orbital ranges), and lower DC-to-DC end-to-end efficiency than mature microwave systems.

The leading commercial proponent of orbital-to-ground laser power is Aetherflux, founded by Baiju Bhatt with substantial backing. Aetherflux’s planned Mission 1, scheduled for 2026 on an Apex Aries bus, will transmit approximately 1 kilowatt via near-infrared laser to a photovoltaic ground station — the first commercial space-to-ground laser power demonstration. The ENERA program does not pick a winner between microwave and laser at the architectural level. Both are likely to find their best applications in different contexts: microwave for gigawatt-class baseload power transmission to municipal-scale rectennas, laser for kilowatt-class targeted delivery to mobile platforms, off-grid sites, or dispersed industrial loads where the small receiver footprint is the binding advantage.

Surface fission

Solar power, whether terrestrial or space-based, has a fundamental limitation on the Moon and on Mars: the lunar polar craters never receive direct sunlight, and Mars dust storms can reduce surface insolation by more than 90% at the equator and more than 99% at the poles for periods of weeks. Solar alone does not anchor a Mars colony, and solar alone does not anchor a permanent lunar base in the polar resource regions. The complementary technology is surface fission.

The reference NASA program is the Fission Surface Power (FSP) project, jointly developed with the U.S. Department of Energy and the Defense-side Strategic Capabilities Office. The Phase 1 awards in 2022 went to Westinghouse, Lockheed Martin/BWX Technologies, and IX (a Space Nuclear Power Corporation–Intuitive Machines partnership) for 40 kWe-class lunar reactor concept development.⁴ Phase 2 was scheduled to begin in 2025, targeting a flight-ready demonstrator by the early 2030s. Westinghouse’s eVinci microreactor technology is the leading concept, leveraging a heat-pipe-cooled core architecture that has been adapted from terrestrial industrial-heat applications. The 40 kWe target — chosen as the minimum sustained surface-power level required to support a lunar habitat with associated rovers, communications, and life-support loads — is at the low end of the credible space-fission deployment range; the architecture is designed to be modular, with multiple 40 kWe units composing into hundreds of kilowatts of continuous surface power. NASA’s announced goal is six metric tons reactor mass at 40 kWe output, with a design lifetime of ten years of operation without human intervention.

The complementary research result is the Kilopower Reactor Using Stirling TechnologY (KRUSTY) demonstration of March 2018, conducted at the Nevada National Security Site by NASA and Los Alamos National Laboratory, which demonstrated a 1 kWe space fission reactor — the first U.S. space-reactor full-power test in over fifty years.⁵ KRUSTY proved the underlying physics of space-deployable fission and is the technical predecessor of FSP. The Russian RORSAT, TOPAZ, and BUK programs of the 1965–1988 period operated more than thirty space reactors in actual orbit, providing the historical record from which the current FSP program is partially derived. The U.S. SP-100 program of the 1980s and the SNAP-10A reactor flown in 1965 are the analogue U.S. heritage. Surface fission is, on this view, not a speculative technology but a deployed technology that has been dormant in the U.S. for forty years and is being revived for lunar and Mars applications.

For ENERA, surface fission is the answer to the night-and-storm problem. Not the only answer — battery storage, molten-salt thermal storage, and beamed power from orbit are all complementary — but the most reliable answer in the 10-year deployment window. The hybrid Mars surface-power architecture that the program targets is solar PV for daylight loads, fission for night-and-storm baseload, and beamed power from areostationary orbit for high-priority dispatchable loads. The hybrid is more expensive per kilowatt-hour than any single architecture but more reliable than any single architecture. Reliability is the binding constraint at off-Earth deployment.

Demonstrators in flight and in operation

The list of in-space and ground-tested results since 2018 is the load-bearing artifact that distinguishes ENERA from speculative free-energy schemes. Each entry below is concrete and citable.

Caltech SSPD-1 (January 2023 launch). The Caltech Space Solar Power Project, led by Ali Hajimiri, Harry Atwater, and Sergio Pellegrino, launched a 50-kilogram demonstrator on a SpaceX rideshare in January 2023. The MAPLE (Microwave Array for Power-transfer Low-orbit Experiment) experiment successfully transmitted microwave power wirelessly in space, with the signal detected at the Caltech ground receiver on May 22, 2023 — the first such in-space-to-ground power transfer in history.⁶ The DOLCE (Deployable on-Orbit ultraLight Composite Experiment) structural demonstrator, a 1.8 m × 1.8 m deployable composite array, was completely deployed over the Canadian Arctic on September 29, 2023, validating that lightweight phased-array architectures can be packaged in a small launch volume and deployed reliably in orbit. The ALBA (Advanced cell technoLogies and Bandgap Analysis) experiment characterized 32 different photovoltaic cell types in space across six months, identifying thin GaAs cells as the most robust to long-duration radiation exposure. The mission ended communications on November 11, 2023, with all three experiment classes meeting their primary objectives.

JAXA OHISAMA (2025). Japan Space Systems and Kyoto University, with Naoki Shinohara as the principal investigator, launched a 180-kilogram LEO satellite that transmitted approximately 1 kilowatt of microwave power to a 13-element ground rectenna array spanning approximately 600 m² near Suwa, Japan — the first orbital-to-ground microwave power transfer at the kilowatt scale.

AFRL Arachne / SSPRITE / SSPIDR. The U.S. Air Force Research Laboratory’s Space Solar Power Incremental Demonstrations and Research (SSPIDR) program, jointly with Northrop Grumman, demonstrated end-to-end solar-to-RF conversion in ground testing in late 2021 — a single integrated “sandwich tile” combining solar cells on one face with a microwave transmitter on the other, the architectural building block for scalable orbital power stations under the Glaser-class CASSIOPeiA-style designs.

Aetherflux Mission 1 (planned 2026). A LEO satellite on an Apex Aries bus transmitting approximately 1 kilowatt via near-infrared laser to a photovoltaic ground station — the first commercial space-to-ground laser power demonstration.

ESA SOLARIS preparatory programme (2022–2024). A €60 million preparatory programme commissioned two parallel concept studies (Thales Alenia Space + ENEL; Arthur D. Little + ENGIE), both of which concluded that SBSP is technically feasible with credible 2040-class deployment if launch costs continue falling. The ESA ministerial decision on full development funding was scheduled for late 2025.

China’s Zhuri roadmap. A 10 kW LEO demonstrator targeted for 2028, a 1 MW GEO demonstrator by 2030, scaling to 10 MW by 2035, gigawatt-class by 2040, and 2 GW by 2050. The Bishan ground facility in Chongqing is operational; Xidian University’s “Chasing Sun” full ground verification system reported leading microwave-power-transmission numbers in 2023.

Virtus Solis Phase 1 (planned 2027). A 100 kW in-space-assembled demonstrator in a Molniya orbit, jointly with Orbital Composites — 217 hexagonal photovoltaic tiles robotically assembled into a 28-metre array, beaming over 1 kilowatt to Earth as a precursor to a 200 MW commercial station by 2030.

KRUSTY (March 2018). The NASA / Los Alamos Kilopower Reactor Using Stirling TechnologY demonstrator achieved full-power operation at 1 kWe at the Nevada National Security Site — the first U.S. space-reactor full-power test in over fifty years.⁵

NRL SCOPE-M (April 2022). 1.6 kilowatt microwave power transmission over 1 kilometre at the U.S. Army Research Field in Blossom Point, Maryland — the most significant terrestrial microwave-power-beaming demonstration in nearly fifty years, with the rectenna RF-to-DC conversion at 73% efficiency at X-band.³

The cumulative weight of these results is the relevant data point. SBSP and wireless power transmission are no longer speculative. They are deployed-flight-experiment technology, with operational ground hardware, with the regulatory bars cleared in multiple jurisdictions, and with the cost-curve trajectory pointing toward terrestrial competitiveness within the next decade if launch-cost continues falling. The remaining work is integration, scale, and the regulatory-and-political-economy questions that the next subsection addresses.

Definitional bounds

Before moving to the deployment and integration questions, four exclusions are worth being explicit about, because the loose use of “space-based solar” and “wireless power” in the discourse has produced confusions that make the substantive arguments harder to have.

ENERA does not mean replacement of terrestrial renewables. The Earth-side energy system, for the foreseeable future, is dominated by terrestrial solar PV, wind, batteries, and (regionally) nuclear and hydro. SBSP is a complement to this, not a substitute. The economic case for SBSP is in applications where the terrestrial system has structural disadvantages — high-density urban demand without sufficient transmission corridors, baseload firm power without cheap long-duration storage, off-grid industrial sites, off-Earth deployment — rather than in head-to-head competition with utility-scale solar PV in regions with abundant land and transmission. The terrestrial transition does most of the work of decarbonization. SBSP serves the residual.

ENERA does not mean energy is the only constraint. Even with abundant power, there are mineral, water, semiconductor, transmission, workforce, and political-economic constraints that the energy system has to operate within. The lithium-and-cobalt question, the rare-earth question, the silicon-and-polysilicon question, the high-voltage-DC-cable supply chain — each of these is a binding constraint at sufficient scale, and ENERA does not pretend otherwise. The program funds work specifically on energy-system architectures that are robust to mineral-supply constraints, including the hydrogen-and-electrofuels pathways that dilute the need for direct electrification.

ENERA does not mean fusion is the answer. Fusion is a research bet. The ITER, Commonwealth Fusion Systems SPARC, Helion, and TAE Technologies programs have made substantial progress, and the achievement of net energy gain at the National Ignition Facility in December 2022 is a milestone. But fusion at commercial-deployment scale, with reactors being built at the rate of multiple gigawatts per year, is a 2040s-or-later proposition under any honest assessment. ENERA includes fusion as a long-arc research bet but does not depend on it for the deployable architectures of the next decade or two. The deployable architectures are solar, wind, fission, batteries, electrolyzers, and beamed power.

ENERA does not mean off-Earth deployment is imminent at scale. The Mars and lunar power architectures described above are research-and-demonstration programs over the next decade and operational deployments over the following decade. The first kilowatt-class lunar reactor will operate by the early 2030s under current Phase 2 timelines.⁴ The first megawatt-class SBSP demonstrator will operate by the early 2030s under current commercial roadmaps. The first gigawatt-class commercial SBSP station is a 2040s deployment under any honest assessment. ENERA is an architectural program, not a deployment program, and the deployments are downstream of the architecture work the program funds.

These exclusions are not throat-clearing. They are the load-bearing definitional choices that determine what the rest of the analysis is about, and they constrain what the ENERA program is allowed to claim.

The historical lineage

Energy systems are infrastructural and accumulate over generations. The current architecture is the product of identifiable historical decisions, and the proposed next architecture has to contend with the lessons that have been paid for by the previous ones.

Glaser and the 1968 origin

Peter Glaser’s 1968 Science paper “Power from the Sun: Its Future” is the foundational document for the field.¹ Glaser, then a senior engineer at Arthur D. Little, proposed a two-satellite system at geosynchronous altitude (35,680 km) generating electricity in orbit via photovoltaic arrays and transmitting it to Earth via microwaves. The 1973 U.S. patent on the Solar Power Satellite formalized the concept. The DOE/NASA studies of the late 1970s, conducted in the wake of the 1973 oil shock, evaluated SBSP as a candidate large-scale energy solution; the conclusion was that the physics was sound but the launch-cost economics did not support deployment at then-current Shuttle launch prices. The field went dormant from approximately 1980 through 2010, kept alive by a small academic community but unfunded at substantial scale by any major space agency.

The lessons of the Glaser era are several. The physics of space-based solar power has been understood for half a century; nothing in the underlying science has changed. The economics are sensitive to launch cost in a way that nothing else in the architectural design is sensitive to; reductions in launch cost flow through almost linearly to SBSP economics, and increases similarly. The political-economic question — who builds, owns, and operates a kilometre-scale orbital structure — was unresolved in the 1970s and remains unresolved today. The technical architecture has not converged on a single canonical design; the CASSIOPeiA, MR-SPS, and tethered-bolometer architectures all remain candidates.

The Soviet space-fission lineage

The Russian RORSAT, TOPAZ, and BUK programs of the 1965–1988 period operated more than thirty fission reactors in actual orbit, primarily on naval-radar reconnaissance satellites in low Earth orbit. The RORSATs (Radar Ocean Reconnaissance Satellites) used uranium-235 fuel and thermoelectric conversion at output levels of approximately 3 kWe. The TOPAZ program used thermionic conversion at higher efficiency. Several reactors re-entered uncontrolled when the satellite buses failed, producing the well-known Cosmos-954 contamination event over northern Canada in 1978 and the 1983 Cosmos-1402 break-up. The Soviet record demonstrated that space-deployable fission is engineerable at industrial scale; it also demonstrated that the safety case for low-Earth-orbit fission is unforgiving of failure.

The U.S. analogue is the SNAP-10A reactor flown in 1965 — the only U.S. space reactor ever flown in orbit — and the SP-100 program of the 1980s, which developed but never flew a 100 kWe reactor architecture. The cumulative U.S. and Soviet record is the body of operational evidence behind the current FSP program. KRUSTY in 2018 was the first U.S. space-reactor full-power test in fifty years.⁵ The discipline that the historical record imposes is conservatism about safety and about lifecycle planning; the lessons of Cosmos-954 and the SP-100 cancellation are foundational.

The terrestrial transition

The dominant historical fact in the global energy system over the last fifteen years is the cost decline of utility-scale solar PV and lithium-ion battery storage. The Lazard LCOE+ analysis tracks the full trajectory: utility-scale solar PV has fallen from approximately $359/MWh in 2009 to$61/MWh in 2024, an 83% decline over fifteen years.⁷ Onshore wind has fallen approximately 70% over the same period. Lithium-ion battery storage has fallen approximately 90%. The terrestrial deployment-rate has lagged the cost-curve, but the cost-curve is the underlying engine, and the cost-curve is continuing.

The lesson for ENERA is that the cost-curve dynamics of solar and batteries are the dominant historical fact and that any space-based architecture has to compete with the cost-curve, not against a fixed competitor. The Lazard 2024 finding that utility-scale solar PV LCOE rose for the first time on the high end was significant; the rise was driven by interest rates and supply-chain factors rather than by the underlying technology, and the long-run trend is expected to resume as those factors normalize. ENERA does not assume the trend reverses. It assumes the trend continues and treats SBSP as a complement rather than a substitute.

Mars-specific power architecture

Mars receives only 43% of Earth’s solar flux (~586 W/m² mean vs. 1,361 W/m²), and its eccentric orbit produces a roughly 45% perihelion-to-aphelion modulation. Worse, dust storms can reduce surface insolation by more than 90% at the equator and more than 99% at the poles for periods of weeks to months. The 2018 global dust storm — which observed Opportunity rover insolation drops below operational thresholds for over a month and ultimately ended the rover mission — is the historical reference point. Mars surface solar alone cannot anchor a Mars colony at the populated-base scale that durable human presence requires.

Three credible architectures compose into a workable Mars power system. First, Mars-orbit space-based solar power: a microwave SPS in Mars-areostationary orbit at approximately 17,000 km altitude remains fully illuminated during dust storms because dust does not extend that high; microwave or laser beaming to surface receivers compensates for dust losses. The atmospheric attenuation of microwaves in Mars’ thin CO₂ atmosphere is poorly characterized — this is one of the most leveraged open scientific problems for Mars settlement, and one no national programme is currently funding. Second, surface fission power: the FSP architecture scaled to Mars, with multiple 40 kWe units composing into hundreds of kilowatts of continuous surface power. The U.S. Department of Energy’s stated goal is reactor-on-the-Moon by approximately 2030; the Mars deployment is a 2030s timeline with the lunar deployment as the proving ground.⁴ Third, hybrid solar plus fission plus battery storage: surface solar farms for daylight loads, fission for night and storm-period baseload, lithium or flow battery storage for transient buffering, and orbital solar later in the decade for high-priority dispatchable loads. This is the most likely actual architecture; it is more expensive per kilowatt-hour than any single architecture but more reliable than any single architecture.

The ENERA program funds work on each of these legs. The Mars-orbit microwave-attenuation experiment is the single highest-leverage data point; it is funded as a small-satellite Mars-orbiter mission target for the late 2020s. The hybrid architecture analysis — what fraction of Mars surface power should come from each leg, at what cost, with what reliability profile — is the systems-engineering question that the program treats as an integration problem rather than as separate technology problems.

Three risk scenarios

Honest planning for the program requires honest enumeration of the failure modes. We name three. They are not exhaustive. They are the ones the design has to take seriously.

Scenario A — Launch-cost stagnation

The first failure mode is the launch-cost-stagnation scenario. SBSP economics are sensitive to launch cost in a way that nothing else in the architectural design is sensitive to. If Starship-class reusability does not deliver — if the cost-per-kilogram-to-LEO does not fall from current levels of ~$1,500 to ~$200 over the next decade — then SBSP economics remain marginal at best for terrestrial deployment, and the gigawatt-class commercial stations recede into the 2050s. The terrestrial deployment of SBSP is contingent on a launch-cost trajectory that is currently the responsibility of approximately three companies. Programmatic over-reliance on any single launch-cost trajectory is a strategic risk the ENERA program treats as a planning constraint, by funding architectures that scale gracefully to a wide range of launch-cost outcomes — including the low end where SBSP becomes the dominant Earth-side dispatchable-power technology, and the high end where SBSP remains a niche off-Earth-anchor technology.

Scenario B — Regulatory and beam-safety capture

The second failure mode is the regulatory scenario. Microwave beams of multi-megawatt density propagating through populated airspace are a regulatory event without precedent. The ITU spectrum-allocation question, the FAA / national-aviation airspace question, the public-acceptance question (the “death ray” framing in popular discourse, no matter how factually wrong, is a real political-economy variable), and the international-treaty question (the Outer Space Treaty’s restrictions on weaponizable space deployments are non-trivially adjacent to high-power-density beam systems) all have to be navigated. Beam-safety qualification frameworks for multi-megawatt beams over populated geographies are as much a regulatory problem as a technical one. Public characterization data — open, reproducible, publicly-checkable — is the load-bearing artifact for clearing the regulatory bar.

The risk is not that SBSP is unsafe. The physics says it is safe within standard safety envelopes; the SCOPE-M demonstration cleared the relevant safety thresholds at the 1.6 kilowatt scale.³ The risk is that the regulatory and public-acceptance frameworks are not built in time, or are built in ways that grandfather in incumbent-favoring constraints. The mitigation is open characterization, public data, and engagement with the spectrum-allocation and aviation-regulatory communities from the earliest design stages.

Scenario C — Successful staged deployment

The third scenario, which we treat as the base case if the technology and regulatory work are competent, is staged deployment in which kilowatt-class demonstrators in the 2020s validate megawatt-class commercial systems in the early 2030s, validate gigawatt-class commercial systems by the late 2030s, and compose into Mars-and-lunar architectures by the 2040s. This is the trajectory that the Caltech, JAXA, ESA, Chinese, and U.S. commercial roadmaps point at; it is the trajectory the FSP lunar deployment supports. It is also a long trajectory by tech-industry standards, and the ENERA program is explicit that this is a multi-decade architectural build-out rather than a fast-iteration product cycle.

What technical work bears on this

The reason ENERA appears on a research-company website at all, rather than in a journal of aerospace engineering, is that the technical work is coupled to the rest of the civilizational stack in ways that are not always obvious. We pull three threads back from the energy question to the broader technical agenda.

The first is that ENERA is the precondition for ARCANE. Closed-loop food production via electrolytic hydrogen and bioreactor electricity is not feasible at habitat scale without a reliable kilowatt-to-megawatt source. The energy system is downstream of nothing, and upstream of almost everything else in the off-Earth deployment architecture. The food-energy coupling is structural rather than incidental.

The second is that the orchestration of orbital solar arrays — phased-array synchronization, robotic assembly, fault recovery, thermal management at gigawatt-station scale — is exactly the multi-agent coordination problem that Economic Orchestration is built to handle, and the autonomous robotic assembly is the deployment surface for Physical Intelligence and Humanoid Robotics. A kilometre-scale phased array is not a structure that can be assembled by any human or human-supervised process at deployment scale; it has to be an autonomous-assembly architecture, with hundreds to thousands of robotic agents coordinating across the assembly task and across the long-term operational task. The ENERA program is, in this sense, the largest single deployment surface for the agentic-systems and physical-intelligence layers of the broader Apik research agenda.

The third is that the Cognitive Computing investments in ultra-low-power edge inference directly determine how much of an orbital array’s local control loop can run on a fraction of its harvested power. The orbital-power-to-onboard-compute ratio is the binding constraint on autonomous-control sophistication for orbital infrastructure; reductions in inference energy per token, per inference call, per unit of decision-making, flow directly through to the ENERA program’s deployable autonomy budget.

The summary, drawn back from the technical layers to the energy question, is this. A civilization that has not solved the energy problem is a civilization that has not solved any of the other problems at scale. ENERA is the energy leg of the broader civilization-stack architecture. Without it, the other legs do not have the substrate on which to operate. With it, the architectural envelope that the rest of the program operates within is dramatically larger than current assumptions allow.

Open questions in the field

The research-program agenda. We name nine. These are the questions the program is funded to address.

1. Phased-array synchronization at kilometre scale. Distributed coherent beam-forming over tens of thousands of free-flying nodes requires sub-degree phase synchronization. Pilot-tone retrodirective architectures are proven at hundred-element scales but unbuilt at kilometre scale.
2. Thermal management of high-density solar-to-RF tiles. Caltech identified electrothermal coupling as MAPLE’s principal degradation mode.⁶ Rejecting kilowatt-per-m² waste heat in vacuum at gigawatt-station scale is unsolved.
3. Robotic in-space assembly of structures larger than 1 km. NASA’s OSAM-1 program, the leading public effort, was canceled in 2024. This leaves a serious capability gap that private actors (Virtus Solis, Orbital Composites) are now stepping into.
4. Wireless power transmission to Mars surface receivers from Mars orbit. Atmospheric attenuation of microwaves at Mars CO₂ pressure with variable dust loading has never been characterized experimentally. This is the single most-leveraged ENERA experiment.
5. Beam-safety qualification framework for multi-megawatt beams over populated geographies. This is as much a regulatory and ITU-spectrum-allocation problem as a technical one. Public characterization data is the load-bearing artifact.
6. Radiation-hardened, low-mass thin-film photovoltaics. Atwater-lab GaAs thin films and quantum-dot architectures are the current frontiers; the goal is sub-1 kg/m² tiles that survive a decade in GEO.
7. Long-duration lunar fission demonstration at 40 kWe class. The FSP Phase 2 deployment by the early 2030s is the load-bearing milestone for the surface-fission leg.⁴
8. Hybrid Mars surface-power systems engineering. What fraction of Mars surface power from each leg, at what cost, with what reliability profile, under what dust-storm-frequency assumptions.
9. Long-duration energy storage at multi-day timescales. Either iron-flow batteries, hydrogen, molten-salt thermal, or geological storage; the architecture matters less than the cost-and-reliability frontier.

Each of these is a multi-year research effort. None of them is solved. All of them are tractable. The ENERA program funds work on each, in collaboration with the laboratories named above and with the open scientific community that is the natural home of this work.

Connections

ENERA does not stand alone. The power it delivers is the precondition for ARCANE — closed-loop food production via electrolytic hydrogen and bioreactor electricity, neither of which is feasible at habitat scale without a reliable kilowatt-to-megawatt source. ENERA’s power supports the medical, scientific, and life-support loads that INTEGRITISSUE gates on. On the AI side, the orchestration of orbital solar arrays — phased-array synchronization, robotic assembly, fault recovery, and thermal management — is exactly the multi-agent coordination problem Economic Orchestration is built for, and the autonomous robotic assembly is the deployment surface for Physical Intelligence and Humanoid Robotics. Cognitive Computing investments in ultra-low-power edge inference directly determine how much of an orbital array’s local control loop can run on a fraction of its harvested power.

Where to read further

ARCANE treats the food-production layer that ENERA’s power enables. INTEGRITISSUE treats the biological-resilience layer that ENERA’s power supports. The Agentic Systems, Physical Intelligence, and Economic Orchestration research pillars treat the autonomous-control and orchestration substrate. The manifesto provides the broader architectural framing.