Planetary Scale Energy Systems

Planetary-scale energy means supplying a civilisation's power at the scale of a whole planet, and then asking what the next scale up would even look like. The subject contains two things that do not belong in the same room, and this brief's first job is to keep them apart. One is a cost problem with published numbers. The other is a detectability research programme with published null results, and it is not an engineering trajectory.

Established A planetary energy system is measured in four quantities, in this order. Levelised cost of electricity in dollars per kilowatt-hour, which is the only number that decides deployment. Lifecycle greenhouse-gas intensity in grams of carbon dioxide equivalent per kilowatt-hour, which decides desirability. Capacity factor, which decides how much storage or overbuild is required. And waste heat, the thermodynamic floor, which becomes binding only very far beyond present use. Any argument about planetary-scale energy that does not land on those four is a rhetorical frame rather than an engineering claim.

Frontier The framing under test is that energy capture can be scaled to planetary and stellar levels. At the terrestrial and orbital end that is a cost question and the costs are known. At the stellar end it is a question that has been turned into an observational search, and the search has run: five million objects examined, seven unconfirmed candidates, no detections. Those are different kinds of statement and the page flags them differently throughout.

Established One sourcing limit is declared here rather than papered over. Absolute global primary-energy figures — exajoules, terawatts, the total terrestrial solar resource — could not be obtained for this rewrite; the statistical agencies were unreachable. This brief therefore builds its spine on ratios and costs, which are fully sourced, rather than on absolute totals it would have to invent. That is a narrower page and a more defensible one, and where a number is missing the page says so instead of reaching for a remembered figure.

Established The terrestrial baseline is the number every other option in this brief is measured against, and it is low. NASA's Office of Technology, Policy and Strategy assembled the comparison for 2050 using NREL projections: terrestrial renewable electricity in 2050 at $0.02 to $0.05 per kilowatt-hour, with land-based wind without storage having both the lowest cost and the lowest emission intensity of every technology NREL tracks. Terrestrial renewables sit at 13 to 43 grams of CO₂ equivalent per kilowatt-hour. Frontier The same table gives coal at 486 and natural gas at 1,001 grams — and the gas figure is higher than the report's own coal figure and out of line with the usual lifecycle literature. It is carried here as the report states it and flagged rather than silently corrected, because quietly repairing a source's number is worse than showing the reader where it looks wrong.

Established The fossil share, in the same report's sourcing, is the reason any of this is urgent. Renewables and nuclear together supplied 40% of US electricity in 2021; the electric power sector was 25% of US greenhouse-gas emissions in 2020; the Energy Information Administration projects 44% of US electricity still fossil-fired in 2050; and the International Energy Agency holds that net zero requires cutting fossil fuels from about 80% of global energy today to slightly over 20% by 2050. Frontier That last ratio is the actual planetary-scale energy problem, and it is a deployment and finance problem rather than a physics one.

Established Space-based solar power is the load-bearing case in this brief, and it is unusually well evidenced because the agency that would gain most from a positive answer published a negative one. NASA's OTPS study of January 2024 costs two reference designs, each delivering 2 gigawatts to the grid starting in 2050. Reference design one, an innovative heliostat swarm: 11.5 square kilometres of solar panel, 5.9 million kilograms of system mass, 99% capacity factor, and a levelised cost of $0.61 per kilowatt-hour. Reference design two, a mature planar array: 19 square kilometres, 10 million kilograms, 60% capacity factor — requiring five such systems to match the first design's output — and $1.59 per kilowatt-hour. Established Against the 2050 terrestrial baseline that is 12 to 31 times more expensive for the first and 32 to 80 times for the second.

Established The cost structure is more informative than the totals, and it is where this brief's central finding comes from. Total lifecycle cost is $276 billion for the first design and $434 billion for the second. They require 2,321 and 3,960 launches respectively. Launch is 71% and 77% of lifecycle cost; manufacturing is 22% and 18%. Lifecycle emissions are 26 and 40 grams of CO₂ equivalent per kilowatt-hour, of which launch accounts for 64% and 72%. The baseline launch assumption is $100 million per launch, about $1,000 per kilogram, with a 15% bulk discount. Established And the single most striking line in the whole study: twelve of every thirteen launches serve only in-orbit refuelling of the tugs, not payload.

Frontier The study's own competitive case is a conditional, and it should be read as one. Seven improvements together: launch to $50 million per launch, about $500 per kilogram and $425 with a block discount; solar cell efficiency from 35% to 50%; first-unit servicer cost from $1 billion to $100 million; first-unit debris-removal vehicle from $500 million to $50 million; the manufacturing learning curve improved by five percentage points; hardware life from 10 to 15 years; and electric propulsion for the transfer from low Earth orbit to geostationary orbit. The result is $0.04 per kilowatt-hour for the first design, a 95% cost reduction, and $0.08 for the second, a 93% reduction, with emissions down to 3.87 and 4.33 grams. Frontier Individually, direct-to-geostationary launch saves 42 to 47%; electric orbital transfer saves 63 to 69% of cost and 54 to 63% of emissions; a fifteen-year life saves 26%; the cell-efficiency jump saves 25 to 26%. That is not a roadmap. It is a list of seven things each of which would individually be a major achievement, all of which must land together.

Established The emissions figures carry a caveat the report states and this brief repeats: they do not include upper-atmosphere effects, which the report assumes are worse than producing the same emissions at the surface. A launch-dominated emissions profile with the atmospheric chemistry of the launches left out is a number with a known direction of error.

Established And the calibrating sentence about the demonstration everyone cites. The same NASA report characterises Caltech's SSPD-1 as the first successful electricity beaming demonstration from space to ground, in June 2023, conducted “at a scale that is orders of magnitude below what is baselined for the systems studied in this report”. Established That is a third-party assessment of a demonstration and it is the fairest single sentence available on it. The demonstration is real; it does not bear on the economics.

Established Now the finding that reorients the whole subject: the beaming half was solved fifty years ago. The measured record, all from the JPL and Raytheon-era programme: end-to-end DC-to-DC efficiency of 54.18% with RF-to-DC conversion at 78.67% plus or minus 1.1%, in 1975. A high-power reception-conversion array achieving more than 80% collection-conversion efficiency and up to 30.4 kilowatts of DC output, also 1975. A laboratory system at 54% DC-to-DC with a subsystem transporting over 30 kilowatts, 1976. Free-space transmission approaching 54% at laboratory scale. And rectennas designed for more than 85% RF-to-DC conversion. Established Nothing about space-based solar power is blocked on transmission physics. It is blocked on dollars per kilogram to orbit and on assembling a five-point-nine-million-kilogram structure.

Speculative The 1970s reference-system studies give the scale of “planetary” properly, and it is sobering. The NASA and Department of Energy architecture covered satellites, space construction, transportation, ground stations and operations control. Sizing tradeoffs put a 2.45 gigahertz configuration at a 1.53 kilometre transmitting antenna delivering 5.05 gigawatts into a 6.8 kilometre-diameter rectenna, and a 5.8 gigahertz configuration at a 0.75 kilometre antenna delivering 2.72 gigawatts into a 5.8 kilometre rectenna. The microwave system requires 101,552 power amplifiers under coherent phase control. Speculative Kilometre-class apertures, six-figure amplifier counts and multi-kilometre ground footprints — per satellite. That is the honest measure of the phrase.

Established One parameter in the modern study is ahead of the flight record and should be flagged. The 2024 analysis takes 35% solar cell efficiency as its baseline and 50% as the improvement case. The space-qualified figures in the NASA literature this brief could reach are more modest: Spectrolab multijunction production cells at 21.5%, 25.1% and 26.8%; a manufacturing-technology programme targeting 24 to 26%; a 1990s status review at 22 to 26%; and a 1981 projection of 20% for two-junction and 23 to 24% for three-junction devices. The study's baseline is already ahead of those, and its improvement case is a research target rather than a product.

Frontier The live positions in this subject sort by which of the four measuring quantities they are arguing about, and several of them are not in conflict at all — they are about different halves of the system.

Established Position one: space-based solar power is cost-prohibitive and technically infeasible today. This is the baseline conclusion of the study cited throughout section 2, stated by its own authors: feasible by 2050, economically uncompetitive under current assumptions. Frontier Position two, from the same report: it can be made cost-competitive by 2050 given a specific bundle of seven improvements. The report presents this as a conditional rather than a forecast, and this brief carries it the same way. The two positions are the same document and are not in tension; treating the conditional as the finding is how the study gets misquoted.

Established Position three: microwave power beaming is a solved engineering problem. At kilowatt scale this is simply true and has been since 1975 — 54% end to end, above 80% at the rectenna, 30 kilowatts delivered. Frontier At gigawatt scale it is unproven, and the unproven part is coherent phase control across a hundred thousand amplifiers on a kilometre-class aperture rather than the conversion efficiency of any individual element.

Established Position four, and the most consequential: launch cost is the dominant term and everything else is secondary. Within the 2024 study's assumptions this is not an opinion — launch is 71 to 77% of lifecycle cost and 64 to 72% of lifecycle emissions, and twelve of every thirteen launches exist only to refuel tugs. Frontier The interesting consequence is that the most powerful single lever on space solar economics is not a solar technology at all: electric orbital transfer alone saves 63 to 69% of cost, because it attacks the refuelling launches rather than the payload ones.

Speculative Position five: kilometre-class geostationary apertures with of order a hundred thousand coherently phased amplifiers are buildable. This is the 1979-to-1981 reference-system position and it has never been tested at any scale. Nothing in the intervening forty-five years has built a phased array of that size anywhere, in orbit or on the ground.

Frontier Position six: a civilisation's energy use is thermodynamically obliged to reappear as mid-infrared waste heat, so large energy supplies are searchable. This is the premise of the G-hat infrared survey programme, resting on the argument that life has potential for exponential growth until checked by resource limits and that intelligence implies overcoming such limits. The same programme derives a maximum galaxy-colonisation timescale below 10⁹ years. Speculative Position seven: partial Dyson spheres exist and are findable in current survey data. This has been tested. A search across five million objects in Gaia DR3, 2MASS and WISE, with a convolutional neural network filtering false positives, produced seven candidates, all around M dwarfs, all requiring further investigation rather than constituting detections.

Established Position eight: monolithic Dyson spheres are physically impossible and only swarms are coherent. The theoretical treatment states it plainly — a solid sphere is dynamically unstable under gravity and radiation pressure and mechanically unstable to buckling. Speculative The same treatment flags a problem nobody has worked on: the radiative feedback of the enclosing structure on the central star's own structure and luminosity is unexplored. Building a swarm changes the star you are harvesting, and that question belongs to Stellar Engineering.

Handwave Position nine, at the far tail: stellar harvesting could multiply the energy available to a future civilisation several-thousand-fold. One analysis argues that harvesting stars of roughly 0.2 to 1 solar masses out to several tens of megaparsecs, over a horizon of order 10¹¹ years before accelerating expansion carries the rest of the universe out of reach, would do this. Handwave It is carried because the directive is to name hypotheses rather than omit them, and it is flagged at the weakest level available. The author line on the source could not be confirmed, so no author is named here.

Handwave And position ten, which this brief's predecessor asserted and which has no support at all: that humanity currently sits at about 0.7 on the Kardashev scale. No source obtainable for this rewrite supports any specific figure. It is a back-of-envelope convention whose derivation is essentially never shown, and it is corrected in section 14 rather than repeated here.

Established The binding bottleneck is transport, and it has been for fifty years. The two halves of “energy capture at planetary scale” have completely different maturities. Conversion and beaming: solved and measured — 54% end-to-end DC-to-DC microwave transfer with above-80% rectenna collection, demonstrated in 1975, at 30 kilowatts. Getting the collector there: unsolved, and the reason everything else fails — 5.9 to 10 million kilograms of system mass, 2,321 to 3,960 launches, launch at 71 to 77% of cost, and twelve of every thirteen launches spent on refuelling.

Frontier The second bottleneck is orbital assembly, and this brief does not own it. A 5.9-million-kilogram structure spanning 11.5 square kilometres of collector is the largest orbital-assembly demand anyone has costed, and it is the reason this subject cannot be separated from Space-Based Manufacturing. Manufacturing is 18 to 22% of lifecycle cost in the study, which is the second-largest term after launch and is usually ignored in favour of arguing about cells.

Frontier Third, capacity factor, which is the quantity that makes the orbital case interesting at all. The heliostat-swarm design achieves 99% of the year; the planar-array design achieves 60% and needs five copies to match the first's output. Established That contrast is the honest statement of what space buys: not more sunlight per square metre so much as sunlight that does not stop. Frontier It is also why the fusion dependency is the one this brief records — fusion is the only candidate on this map that changes the capacity-factor term without a storage build-out.

Frontier Fourth, on the terrestrial side, a set of bottlenecks this brief deliberately does not quantify. Grid-scale storage, long-distance high-voltage and superconducting transmission, and continental-grid megaprojects are real constraints on any planetary energy system, and the research for this rewrite could not reach sources that characterise them. They are owned by Energy Storage Revolutions, Superconducting Infrastructure and Energy Corridors, and this page names them and hands them over rather than asserting figures it cannot source.

Speculative And fifth, the thermodynamic floor, which is not currently binding and eventually is. Every joule used on a planet becomes heat that must be radiated. That is the constraint that makes megastructure civilisations searchable in the mid-infrared, and it is the only bottleneck in this brief that no engineering can remove. At present human energy use it is nowhere near binding, which is why it appears fourth on the list of measuring quantities rather than first.

Frontier This brief records one dependency, on FR-I-12, and the reason is specific rather than gestural. Every planetary-scale energy scenario in the evidence base is a cost-and-capacity-factor argument. Cost is dominated by transport and assembly; capacity factor decides how much storage or overbuild the system needs. Fusion is the only candidate that changes the capacity-factor term without a storage build-out, which makes it the one result on this map whose arrival would rewrite this brief's arithmetic rather than adjusting it.

Frontier All fusion physics, milestones and cost claims are deferred to that brief and to Commercial Fusion, and this page asserts none of them. No fusion evidence was gathered for this rewrite and none is stated here. That is a deliberate boundary: a planetary-energy brief that recited fusion milestones would be duplicating another page's adjudication with worse sourcing.

Frontier A second dependency is real and runs to orbital industry rather than to physics. Space-Based Manufacturing owns orbital assembly, and nothing in the space-solar column gets built without it. Established The number that makes the dependency concrete is 5.9 million kilograms — the mass of the cheaper of the two reference designs, and the largest orbital-assembly demand anyone has put a price on.

Established And a dependency that is entirely outside this map: launch price. The single quantity that sets this brief's answer is about $1,000 per kilogram to orbit at a baseline of $100 million per launch, with the competitive case at roughly $500 per kilogram. It is not a research result, it is a market price, and no laboratory produces it. A brief that treated launch cost as a technology to be developed rather than a price to be observed would be miscategorising the constraint that dominates its own cost model.

Established The decisive experiments in this subject were mostly run in the 1970s, and that is the most useful thing this section can say. Microwave power transfer was measured end-to-end at 54.18% DC-to-DC with 78.67% RF-to-DC conversion; a reception-conversion array reached above 80% collection-conversion at 30.4 kilowatts of DC output; rectenna designs reached above 85% RF-to-DC. Repeating them proves nothing that is not already known.

Speculative The experiment that has not been run is the one at scale. Coherent phase control across of order 100,000 amplifiers on a kilometre-class transmitting aperture, holding a beam onto a multi-kilometre rectenna, is the untested step, and it is an array-engineering experiment rather than a power-conversion one. Nothing in the fetched record demonstrates phased-array coherence at anything approaching that element count.

Speculative The second untested step is orbital assembly of a multi-million-kilogram structure, which is not a power experiment at all. It is the reason this brief points at Space-Based Manufacturing for the capability and at Space-Based Solar Power for the architecture.

Established Third, the demonstration that has been run and is routinely over-read. Caltech's SSPD-1 beamed detectable power from orbit to ground in June 2023. NASA's own characterisation is that it was conducted at a scale orders of magnitude below what is baselined for the systems in its study. Established It is a real first and it is not an economic data point, and the two statements are compatible.

Frontier Fourth, and a correction to this brief's own predecessor: Japan's beaming demonstration is a plan, not a result. A January 2024 document records that Japan plans to test beaming energy from space to Earth in 2025. That is a forward-looking statement in a document written before the date it refers to, and no source obtained for this rewrite reports an outcome. The earlier version of this page stated it as accomplished; it is corrected here rather than carried.

Established And fifth, at the far end of the subject, the experiment that has already returned a null. The search for waste-heat signatures of stellar-scale energy use is an observational programme and it has run at scale: five million objects across three surveys, filtered by a convolutional neural network, yielding seven candidates, all M dwarfs, none confirmed. The stellar end of this topic has an experimental programme, and its result so far is nothing.

Established The engineering requirements for the orbital case are specifiable because two of them have been costed in detail. Collector area of 11.5 or 19 square kilometres; system mass of 5.9 or 10 million kilograms; a transmitting aperture at kilometre class; a ground rectenna at 5.8 to 6.8 kilometres in diameter; and 101,552 power amplifiers under coherent phase control in the reference architecture. Delivered power, 2 gigawatts to the grid.

Established The requirement that dominates every other is the launch campaign. 2,321 to 3,960 launches per system, at $100 million each, with twelve of every thirteen serving in-orbit refuelling. That is not a spacecraft requirement; it is a logistics requirement of a kind nothing in spaceflight history approaches, and it is the term the study's improvement case attacks hardest — electric orbital transfer saves 63 to 69% of cost precisely by removing refuelling flights.

Established The component requirement most often quoted is also the one furthest ahead of the flight record. Photovoltaic conversion at 35% baseline and 50% improved, against space-qualified multijunction production cells in the fetched NASA literature at 21.5% to 26.8%. Established The gap between a study's baseline assumption and a manufacturer's production figure is a real engineering requirement and is usually presented as an incremental one. It is not: the improvement case asks for roughly double the demonstrated production efficiency.

Established And the servicing requirement, which the study prices and most discussion omits. The competitive case needs first-unit servicer cost to fall from $1 billion to $100 million and first-unit debris-removal vehicle cost from $500 million to $50 million — tenfold reductions in two vehicle classes that barely exist. A gigawatt-class structure in geostationary orbit with a ten-to-fifteen-year hardware life is a servicing problem before it is a power problem.

Frontier The boundary that matters most on this page is with the megastructure briefs, and it is drawn hard. Dyson Swarms owns swarm architecture, construction, materials budget, orbital mechanics and stability. This brief mentions Dyson swarms only as the asymptote of the energy-capture argument and as a detectability programme, and then hands off. Everything about sphere instability, optimal radii and construction belongs on that side of the line.

Speculative And with the star itself. Stellar Engineering owns anything that acts on the star — including the open problem this brief names in one sentence and does not pursue: the radiative feedback of an enclosing structure on the central star's own structure and luminosity is unexplored, which means nobody knows what harvesting a star does to the star.

Established Within the practical half, the adjacencies are the ones that own the terms in the cost model. Space-Based Solar Power owns the architecture in detail; this brief carries the levelised-cost and launch-share numbers because they are the planetary-energy argument. Space-Based Manufacturing owns orbital assembly. Wireless Energy Transmission owns beaming as a technology; this page carries the 1975 efficiencies because they are what settle the capture-versus-transport question. Frontier Lunar Energy Infrastructure owns the one case where the orbital-solar economics genuinely differ, because the market is off Earth and the terrestrial baseline does not apply.

Frontier On the terrestrial side this brief is a customer rather than an owner. Energy Storage Revolutions, Superconducting Infrastructure, High-Temperature Superconductors and Energy Corridors own storage and transmission; Commercial Fusion owns the terrestrial fusion case. Speculative And Interstellar Archaeology owns the search side of the waste-heat argument, which is where the stellar end of this topic actually produces data.

Established The strongest institutional fact in this brief is who wrote its central document. An agency with a clear institutional interest in a positive answer on space-based solar power commissioned a cost study and published 12-to-31-fold and 32-to-80-fold cost gaps. Established Interest running against the finding is the strongest form a negative result can take, and it is why the numbers in section 2 are load-bearing rather than contestable.

Established The second institutional observation is about the shape of the field's own literature. Publications on space-based solar power nearly doubled between 2018 and 2022, concentrated in China, the United States, the European Union, Japan and Russia, with studies, design concepts and technology development identified in seven countries. Activity is rising while the economics have not moved, which is a fact about strategic positioning rather than about cost curves.

Frontier Third, a governance requirement specific to beamed power: the beam has a ground footprint measured in kilometres and a regulator. A 5.8 to 6.8 kilometre rectenna is a land-use decision, a spectrum allocation and a public-safety case simultaneously, and the 1970s reference studies treated ground stations and operations control as first-class architecture elements rather than as details. That framing has largely dropped out of modern advocacy and should not have.

Speculative Fourth, and about this subject's rhetoric rather than its engineering. The Kardashev scale is roughly sixty years old, is regarded by its own reviewers as oversimplified, and remains in use chiefly because it emphasises detectability — the reason the G-hat infrared survey revived interest in it. Handwave Used as a classification of what a civilisation could be detected doing, it is a serviceable tool. Used as a ladder a species climbs on a schedule, it is a rhetorical frame with no measurement behind it, and institutions that adopt the second reading end up planning against a number nobody computed.

Frontier The concrete ethical questions here are about beams and land, not about civilisational destiny. A multi-kilometre rectenna is a large land taking; a gigawatt-class microwave beam through the atmosphere is a public-safety and spectrum question; and the beam's failure modes matter more to people underneath it than its efficiency does. The 1975 measurements establish that the power can be delivered; they say nothing about who lives under the receiver.

Established Second, a real environmental caveat that the central study itself flags. Space-based solar's lifecycle emissions are 64 to 72% launch, and the reported figures exclude upper-atmosphere effects, which the report assumes are worse than surface emissions. A low-carbon technology whose carbon is almost entirely in a launch campaign whose atmospheric chemistry has been left out of the accounting is not yet a clean-energy result, and saying so is the honest position.

Frontier Third, the opportunity-cost question, stated without moralising. $276 to $434 billion delivers 2 gigawatts from orbit at $0.61 to $1.59 a kilowatt-hour. The same money spent on land-based wind, which the underlying comparison identifies as having both the lowest cost and the lowest emission intensity of every technology tracked, delivers considerably more at two to five cents. Speculative The defensible pro-orbital argument is not cost but capacity factor and land use — 99% of the year, and no competition for surface area — and advocates would do better making that argument than the cost one.

Established Fourth, and about this page's own conduct. An earlier version of this brief stated that humanity sits around Kardashev 0.7 and that Japan demonstrated space-to-ground beaming in 2025. Neither has any support in the sources obtained for this rewrite; the first has no derivation anyone shows, and the second is a plan recorded in a document written a year before the date. Both are corrected in section 14. Energy is a subject where confident round numbers travel further than the evidence behind them, and a page in this position should be conspicuous about which of its numbers it can defend.

Established The civilizational claim that survives contact with the evidence is smaller and more useful than the one usually made. Abundant clean energy is an engineering, finance and deployment problem rather than a physics barrier, and the ratio that defines it is concrete: cutting fossil fuels from about 80% of global energy to slightly over 20% by 2050 in the net-zero pathway, against a projection that 44% of US electricity is still fossil in 2050.

Speculative The orbital case, if the seven improvements ever landed together, would add something terrestrial systems structurally cannot: near-continuous output. A 99%-capacity-factor source needs no storage and no overbuild, and storage is one of the two largest unpriced terms in every terrestrial decarbonisation plan. Speculative That is the strongest version of the space-solar argument and it is not the version usually made, which is about sunlight being brighter above the atmosphere.

Speculative Beyond the planet, the honest statement is that this topic stops being engineering and becomes astronomy. The waste-heat argument makes large energy supplies detectable in principle. The search has been done at scale — five million objects, seven unconfirmed M-dwarf candidates, no detections — and the best-developed theory says monolithic spheres cannot exist at all and that the effect of harvesting a star on the star is unexplored. Handwave So the defensible sentence is not “energy capture scales to stellar levels”. It is: we have looked, at scale, for anyone who has done it, and found nothing — and we do not yet know what harvesting a star does to the star.

Speculative And one implication that runs the other way and is worth keeping. The reason the stellar end of this subject is searchable at all is thermodynamics: a large energy supply must reappear as mid-infrared waste heat. That makes energy use the one civilisational property that cannot be hidden, which is a genuinely interesting fact about the universe and is the most defensible thing the Kardashev framing ever contributed.

These horizons track levelised cost, launch price and capacity factor, because those are the three quantities the evidence base actually constrains:

• 10 yr: Established Expect terrestrial renewables to keep setting the baseline this brief measures against, with land-based wind at the low end on both cost and emissions. Frontier Expect more space-solar demonstrations at scales orders of magnitude below the studied systems, and more national programmes: publications nearly doubled between 2018 and 2022 and several states have announced activity. Frontier The number to watch is not a demonstration but a price: launch at roughly $500 per kilogram is the hinge on which the whole cost model turns, and it is a market observation rather than a research milestone.
• 25 yr: Frontier This is the window the 2024 study actually models, with 2 gigawatts to the grid starting in 2050 under both reference designs. Frontier The competitive case requires seven simultaneous improvements totalling a 93 to 95% cost reduction; the honest forecast is that some will land and the bundle will not. Speculative A single space-solar system delivering commercial power at a competitive price in this window would require the launch and orbital-transfer terms to move first, and those are the two furthest from anyone's laboratory.
• 50 yr: Speculative Beyond the modelled horizon, the plausible shape is that orbital solar finds a market where the terrestrial baseline does not apply — off-Earth demand rather than grid supply — which is a much smaller and much more defensible proposition. Handwave Any forecast of planetary-scale orbital power in this window is forecasting a launch industry, an orbital assembly industry and a servicing industry simultaneously, none of which this brief can evidence.
• 100 / 250+ yr: Handwave Beyond useful forecasting, and the megastructure end belongs to Dyson Swarms and Stellar Engineering rather than here. Speculative The one durable statement is the thermodynamic one: whatever a civilisation does with energy at that scale reappears as waste heat, which is why the only evidence anyone has about it is a null from a survey of five million stars.

• Depends on One dependency, and the reason for it is arithmetic rather than affinity. This brief waits on FR-I-12 because every planetary-scale energy scenario in the evidence base reduces to a cost-and-capacity-factor argument, and fusion is the only candidate on this map that changes the capacity-factor term without a storage build-out. All fusion physics, milestones and cost claims are deferred to that brief and to Commercial Fusion; none is asserted here, and no fusion evidence was gathered for this page. Two further dependencies are recorded in prose rather than as edges because they are not results: orbital assembly, which Space-Based Manufacturing owns and without which a 5.9-million-kilogram structure is unbuildable, and the launch price of roughly $1,000 per kilogram, which is a market observation rather than a research output.
• Enables The enabling reach is broad in principle and unevidenced in practice, so no typed enabling edge is claimed. A planetary energy system at the costs in section 2 enables nothing that terrestrial renewables at two to five cents a kilowatt-hour do not already enable more cheaply. The one genuine enabling case is off-Earth, where the terrestrial baseline does not apply and orbital or lunar power competes against nothing — which is why Lunar Energy Infrastructure is the adjacency where these economics actually change sign.
• Adjacent Space-Based Solar Power owns the orbital architecture in detail; this brief carries its levelised-cost and launch-share numbers because they are the planetary-energy argument itself. Wireless Energy Transmission owns beaming as a technology. Energy Storage Revolutions, Superconducting Infrastructure, High-Temperature Superconductors and Energy Corridors own the terrestrial storage and transmission terms this brief names and does not quantify. At the far end, Dyson Swarms owns swarm architecture and construction, Stellar Engineering owns anything acting on the star including the unexplored radiative-feedback problem, and Interstellar Archaeology owns the waste-heat search that is the only place the stellar end of this topic produces data.

Handwave “Humanity sits at around Kardashev 0.7.” No source obtainable for this rewrite supports any specific figure, and the earlier version of this page asserted one. Speculative The scale itself is roughly sixty years old, its own reviewers describe it as oversimplified, and it survives chiefly because it emphasises detectability rather than because it measures anything. A back-of-envelope convention whose derivation is essentially never shown is not a measurement, and stating it as one was the single largest defect in this brief's predecessor.

Established “The problem with space-based solar power is beaming the energy down.” It is not, and it has not been since 1975. End-to-end DC-to-DC microwave transfer was measured at 54.18%, RF-to-DC conversion at 78.67%, rectenna collection-conversion above 80% at 30.4 kilowatts of output, with rectenna designs reaching above 85%. Established The bottleneck is transport: 5.9 to 10 million kilograms, 2,321 to 3,960 launches, launch at 71 to 77% of lifecycle cost, and twelve of every thirteen launches existing only to refuel the tugs.

Established “Caltech's 2023 demonstration showed space solar is nearly viable.” It showed that power can be beamed from orbit to the ground, which is a genuine first. NASA's own characterisation of it is that it was conducted at a scale orders of magnitude below what is baselined for the systems in its cost study. Frontier A first is not an economic data point, and the two claims are compatible rather than in tension.

Frontier “Japan demonstrated space-to-ground power beaming in 2025.” The source available for this rewrite is a January 2024 document recording that Japan plans to test beaming energy from space to Earth in 2025. Established That is a forward-looking statement written before the date it refers to, and no outcome is reported in any source obtained here. The earlier version of this brief stated it as accomplished, which is corrected rather than repeated.

Frontier “Space solar is 12 to 80 times too expensive, so it is dead.” That is the baseline conclusion and it is not the whole finding. The same study models a combined-improvement case reaching $0.04 and $0.08 per kilowatt-hour — 95% and 93% reductions. Frontier What makes that a conditional rather than a forecast is that it needs seven improvements simultaneously, each of which would individually be a major achievement, including halving launch cost and roughly doubling the demonstrated production efficiency of space solar cells.

Established “Solar cells at 35 to 50% efficiency are the near-term baseline.” The space-qualified multijunction production figures in the NASA literature reachable for this rewrite are 21.5%, 25.1% and 26.8%, with programme targets of 24 to 26%. The study's 35% baseline is already ahead of those, and its 50% case is a research target rather than a product. A study assumption is not a component specification.

Established “A Dyson sphere is a shell around a star.” A solid shell is dynamically unstable under gravity and radiation pressure and mechanically unstable to buckling; the coherent version is a swarm. Speculative And the question nobody has worked on is what the swarm does to the star: the radiative feedback on the central star's own structure and luminosity is unexplored, which Stellar Engineering owns.

Speculative “Dyson swarms have been detected.” They have not. A search across five million objects in Gaia DR3, 2MASS and WISE produced seven candidates, all around M dwarfs, all requiring further investigation and none confirmed. Frontier The stellar end of this subject is a search programme with a null result, and reporting candidates as detections is the standard error.

Handwave “A Kardashev Type I trajectory is an engineering programme.” It is a rhetorical frame. Established The engineering programme is the one in section 2: levelised cost, capacity factor, emission intensity and launch price, and it is measured in cents per kilowatt-hour rather than in civilisational tiers. A brief that let the second stand in for the first would be replacing a solvable problem with a slogan.

Established And one thing this page does not claim. Grid-scale storage, high-voltage and superconducting transmission and continental-grid megaprojects are genuine constraints on planetary energy, and the research for this rewrite could not reach sources characterising them, so no figures for them appear here. They are named, handed to the briefs that own them, and left unquantified. A gap stated is content; a gap filled from memory is a defect.