The Atom in the Building

Made reliable, solar and wind multiply their minerals.

Made reliable without fossil fuel, solar and wind:

• take about 760 times coal’s critical minerals for the same delivered power;

• emit 200 to 400 g of CO₂ a kWh over their life, a quarter to a half of coal’s and in the range of gas, once their batteries are counted;

• move about 420,000 tonnes of rock for each terawatt-hour for the batteries alone, the same order as the 500,000 tonnes of coal a coal plant burns;

• throw away about 19,000 tonnes of equipment for each terawatt-hour, the same order as coal’s 50,000 tonnes of ash, but as composite, glass and chemical waste that is harder to recycle.

Made reliable the way it is actually done, with fossil plants behind them, they bring coal and gas along with them.

The battery and overbuild figures are this study’s model, built on the published numbers above. Every assumption is stated so it can be checked: a 50-50 mix of solar and wind; batteries at 2 kg of critical minerals, 7 kg of mass and 60 to 140 kg of CO₂ a kWh, replaced every fifteen years; and rock moved at about 300 tonnes for each tonne of lithium, 200 for copper and 10 for graphite. They leave out transmission. Even the 12-hour case assumes a continent-wide high-voltage grid, and every new line is more copper, aluminium, steel and cleared corridor.

2.9 Weighed against coal and oil, measure by measure

Every row is for the same electricity, delivered when it is needed. Solar and wind are taken made reliable without fossil fuel, with two weeks of batteries and 50% overbuild, as the fair comparison requires.

Made reliable and counted whole, solar and wind are worse than coal on seven of these ten measures and match it on two, the rock mined and the waste thrown away.

Coal’s remaining two measures are a record of how coal was used, not of coal. Its smoke and its carbon came from burning it with poor science:

• Filters, scrubbers and catalysts. A modern plant’s fabric filters take out more than 99% of the particulates, its scrubbers about 98% of the sulphur, and its catalysts most of the nitrogen oxides; mercury comes out with them. Many of the plants behind the death rate in the table ran with few or none of these.

• Deaths fall when the controls go in. A study of 480 US coal plants found that around the Keystone plant in Pennsylvania deaths from its pollution averaged more than 600 a year before scrubbers were installed and fell below 100 a year after.

• Carbon can be caught. Captured from a boiler’s flue and returned to the ground, coal’s life-cycle carbon falls to roughly 150 to 250 g/kWh. That is already about even with solar and wind made reliable.

• Not burned at all. Properly processed, every carbon fuel can run smokeless with all its carbon returned, so long as it is not burned in a flame. Coal is gasified in a closed vessel, its sulphur and ash taken out, and the clean gas fed to solid-oxide fuel cells. Or its carbon is fed straight to a carbon fuel cell. Either way the oxygen reaches the fuel through the ceramic and the air never touches it, so the exhaust is carbon dioxide and steam alone, with nothing to separate. Coal gasification with solid-oxide cells has been shown to capture all its carbon dioxide, the sulphur sequestered with it, at efficiency well above a boiler’s. Laboratory carbon fuel cells turn about 80% of the carbon’s energy into electricity and yield a pure stream of carbon dioxide. This is the same principle as the community’s own building plant, whose cell keeps its gas from the air and sends its carbon back to the ground.

Of all the carbon fuels, natural gas is the easiest to return. That is why the Constitution makes gas, not coal, the bridge.

• It arrives clean. It is a gas, its sulphur and water taken out at the field, with no ash and no gasifier needed.

• It goes into the cell as it is. The solid-oxide cell reforms methane inside itself, so the building’s plant is the whole chemical works.

• It carries the most hydrogen for its carbon, four atoms to one. For each gigajoule it makes about 50 kg of CO₂, against about 70 for oil and 88 for bituminous coal, so it leaves the least carbon to return for the energy it gives.

• Its carbon goes back the way it came, down qualified wells into the formations that held it.

• It can give up its carbon as a solid. Split by heat in the district pyrolysis works, methane becomes hydrogen and solid carbon, with no carbon dioxide at all. The carbon goes to the farms as biochar and to building materials, and the hydrogen comes back as fuel.

Processed with the best of science and run in a cell rather than burned, coal is smokeless and its carbon goes back to the ground entire; only the mining and processing remain. Its last two measures then fall below those of solar and wind made reliable, and it keeps about a tenth of their land and under a seven-hundredth of their critical minerals.

That is the density principle at work. Coal is denser than wood, solar or wind, and when it was handled badly the harm came from the handling.

This is the sense in which this paper holds that solar and wind are as harmful to the environment as coal and oil. Their harm takes a different form: mines, tailings, fenced and spaced land, lost habitat and landfill, rather than smoke. It is no bridge that ends. It is renewed with every rebuilding for as long as they are used, and over 250 years at the scale of twelve billion people it is a comparable claim on the earth. For the objective of reducing the footprint of man, it is worse: no other way of making power takes so much land from every other species.

Only nuclear is low on every line at once. That is why the Constitution keeps solar and wind out of the communities, and why the bridge leads to the atom.

3. What the building asks of a reactor

3.1 The load

The NV Build papers size the plant on the public building, the largest in the community: two fuel-cell modules, 650 kW installed, drawing about 300 kW for the building, 60 for the greenhouse and 160 for its data hall. The building’s exhaust at 356 to 384 °C fires the absorption chillers, its heat is banked in the 284 m³ pool, and the supercritical-water waste tube runs at 650 °C. Each of the community’s 1,982 buildings carries its own plant in a utility storey of seven grids.

A reactor that replaces the fuel cell must therefore:

• deliver from about 100 kW to 1 MW of electricity, the public building at the top of the range and the apartment buildings well below it;

• deliver its heat hot enough to drive the absorption cooling and, better, the 650 °C waste tube directly;

• fit the utility storey’s bays and floor loading;

• run for years without being opened, and be changed as one sealed unit.

3.2 Every unit of heat used

A fission plant turns roughly a third of its heat into electricity; two units of heat come with every unit of power. At a central station that heat is thrown into a river or the sky. In the building it is the point. High-temperature gas-cooled and heat-pipe reactors run their outlets at about 650 to 750 °C, hot enough to fire the chillers, heat the waste tube directly, warm the water and the greenhouse and, through the pool, carry the building through the night. The rule of The Building Data Centre holds unchanged: no heat leaves the building while any use remains. This is the efficiency a central station can never have, and the reason the Constitution says “never a central station”.

When the reactor heats the waste tube, the methane the tube makes no longer feeds a fuel cell. It becomes feedstock: hydrogen, carbon for the farms as biochar, and chemicals.

3.3 What exists today

Reactors of about the right size and kind are being built now.

• Radiant’s Kaleidos, a 1 MWe helium-cooled reactor on TRISO fuel, left California on August 13, 2026 for the DOME test bed at Idaho National Laboratory, where its tests are to end in a 150-hour continuous full-power run. It is built to travel by road, sea or air.

• Four microreactors, from Antares, Valar Atomics, Deployable Energy and Aalo Atomics, reached criticality at Idaho under DOE’s Reactor Pilot Program between early June and July 4, 2026. Each must still pass NRC review before commercial use.

• Heat-pipe designs of a few megawatts, such as Westinghouse’s eVinci, are in licensing.

• Kilowatt-scale cores exist. NASA’s KRUSTY test of 2018 ran a 1 kWe-class core, and NASA’s lunar fission programme aims at about 100 kWe. A sealed reactor at the apartment building’s scale of a few hundred kilowatts is within physics; it is an engineering, fuel and licensing problem.

The building reactor is therefore one step smaller than today’s smallest commercial designs, at the public building about the same size. That step is the first research objective.

4. Where society stands

4.1 Two generations lost

After Three Mile Island in 1979 and Chernobyl in 1986, almost no new reactors were ordered in the United States for a generation, and the industry’s supply chain, skills and cost discipline decayed. Germany closed its last three reactors in April 2023. Over the same decades the world built coal at a pace that has cost millions of lives in air pollution. The fear was of radiation; the harm came from what was built instead.

Fukushima shows the pattern exactly. The United Nations Scientific Committee (UNSCEAR) found radiation health effects among the public “unlikely to be discernible”, and Japan has recognized one worker’s death as caused by radiation. More than two thousand disaster-related deaths in Fukushima Prefecture followed the evacuation itself, most among the elderly and ill moved from hospitals and homes. The lesson for a reactor in every building is not that radiation is harmless; it is that measured, published risk must govern, not dread.

4.2 The turn now under way

• Law. The NRC published its final Part 53 rule in March 2026, a risk-informed, technology-inclusive licensing path for advanced reactors, with design approvals expected in 18 months or less. In May 2026 it proposed Part 57 for microreactors: no defined emergency planning zone where an accident cannot exceed 1 rem at the site boundary; remote monitoring, remote and autonomous operation; reactors built and fuelled at the factory under a manufacturing licence; and security sized to consequences. The ADVANCE Act of 2024 extended the Price-Anderson liability framework to 2065.

• Fuel. Centrus has delivered 900 kg of HALEU to DOE, and the Savannah River Site has approval to downblend weapons-origin HEU into about 3.1 tonnes of HALEU. Supply remains the tightest constraint on every small reactor.

• Opinion. Gallup’s March 2026 poll found 46% of Americans wanting more emphasis on nuclear, its highest reading, the only source to gain since 2021, while support for more solar and wind fell. Most Americans still resist a plant in their own area.

The last point is the hardest for this paper. A reactor in the building is the plant in one’s own area, at the closest possible range.

5. The levels of issue, and how each is met

The transition fails if any one level fails. Each level is listed with what must become true, and how the community system meets it.

1. Physics and engineering. A sealed core of 100 kW to 1 MW that shuts itself down and cools itself with no power, no operator and no water (walk-away safety), and stays shut through fire, flood and quake. Met by: TRISO and heat-pipe designs already claim this; Bureau 13’s research enterprises take the building-scale core as a standing objective, and a community offers the first fleet buyer.

2. Fuel. HALEU, 5% to 19.75% uranium-235, in volume. Met by: weapons HEU first (section 6), then new enrichment, then recycled fuel. Below 20% enrichment uranium is not usable for a weapon without re-enrichment.

3. Scale and safeguards. At full build the United States holds about 11 million building plants, world area 1 about 28 million, the world at twelve billion about 258 million. Each is a place where fissile material sits. Met by: – no plutonium ever enters a building core: plutonium of any grade can make a weapon, so it is burned only in national reactors and later in the fuel enterprises’ own recycling plants; – every core is fuelled at the factory and never opened on site; – spent TRISO fuel is among the hardest of all fuels to reprocess; – every core is continuously monitored, its location, temperature, power and seal published, with SI watching for any anomaly and accountable humans deciding; – cores move only between the building and the fuel enterprise’s plant, under a published custody record.

4. Siting where people live. No licensing framework yet contemplates a reactor beneath an apartment floor. Met by: the gates of section 7. Reactors stand first in industrial and logistics buildings beside the transfer yards, then in the public buildings, and in apartment buildings only after a published operating record and at a size whose worst accident gives no meaningful dose beyond its own utility storey. The utility storey becomes the shielded room: fire-rated compartments already, mass added below the core.

5. Security and cyber. Met by: the consequence-based security of Part 57, a core that cannot be made to melt or to release its fuel, controls that fail safe, and a cybersecurity standard Bureau 23 publishes with Bureau 15 verifying it.

6. Waste. Met by: spent cores return whole to the fuel enterprise. For long-term disposal the communities already drill and qualify deep formations to return carbon; the same skills serve deep borehole disposal in qualified formations kilometres down. Later, spent fuel is recycled in fast-spectrum reactors at the fuel plants, which burn the long-lived actinides and leave waste whose radioactivity falls to that of the original ore in a few centuries rather than many thousands of years.

7. Decommissioning. Met by: the core is a module; the building outlives many cores. Ending a reactor is lifting out a sealed unit.

8. Cost. Microreactors cost far more today, per kilowatt, than the stations they would replace. Met by: the same argument the NV Build papers make for the fuel cell. Ten thousand identical plants a year, the volume of five communities a year, is what moves a machine down its learning curve, and decentralisation is what makes the cheap machine possible. A community orders cores as it orders fuel cells: one design, built 1,982 times.

9. Liability and insurance. Met by: the national framework (Price-Anderson to 2065) at first; later the operating record of millions of reactor-years, which insurers price as they price every other risk.

10. People and skills. Met by: the schools teaching what radiation is and is not, from the first years; Business Enterprises in reactor custody, maintenance and monitoring; the fuel enterprise as a career path.

11. Acceptance. Met by: choice, data and ownership. – Each building’s radiation readings, inside and out, are published live, as its water and air are. – A building’s residents hold the first right to decline a reactor on their roof, and see the record of those who did not. – The community holds the reactor’s title, and the residents are its owners through the community: the reactor is theirs, not a utility’s.

12. Will. The deepest issue. A society that started down this road in the 1950s turned back for two generations; the next 250 years may hold their own Fukushima. Met by: the Constitution naming the destination and the gates, so that a bad year slows the crossing but does not abandon it; Bureau 13’s standing objective; and a published checkpoint every decade that measures where the bridge stands.

6. Swords into fuel: the weapons as feedstock

6.1 What exists

At the start of 2026 the Federation of American Scientists counted about 12,187 nuclear warheads: about 9,745 in military stockpiles and about 2,442 retired and awaiting dismantlement, nearly all of them American and Russian.

The International Panel on Fissile Materials counted, at the start of 2025:

• about 1,240 tonnes of highly enriched uranium, of which about 1,100 tonnes sit in or are available for weapons;

• about 570 tonnes of separated plutonium, of which about 140 tonnes sit in or are available for weapons. Russia holds about 88 tonnes and the United States about 38.

6.2 What has been done before

From 1993 to 2013 the Megatons to Megawatts programme blended down 500 tonnes of Russian weapons uranium, the equivalent of about 20,000 warheads, into nearly 15,000 tonnes of reactor fuel. For years up to a tenth of all US electricity came from it. It ran on commercial terms: a buyer that wanted fuel made the disarmament pay for itself.

6.3 What the remaining weapons uranium is worth

Blended down to 19.75% with depleted uranium, 1,100 tonnes of weapons HEU makes about 6,150 tonnes of HALEU, exactly the fuel the building reactors need. Burning 40% to 50% of its uranium-235 gives about 9,000 to 11,000 terawatt-hours of heat, of which about 3,200 to 3,900 TWh is electricity at a third’s efficiency and the rest heat the building uses.

At the assumed 1.2 to 1.6 TWh a year for one community, that is 2,200 to 2,900 community-years: the first cores of several hundred communities for a decade. It is not the world’s fuel for long. It is the startup fuel for the first generation of building reactors, available when new HALEU enrichment is still scarce.

The 140 tonnes of weapons plutonium would give about 1,100 TWh of electricity if fully fissioned. Under the rule of section 5 it never enters a building. It is burned in national reactors as mixed-oxide fuel, or later in the fuel enterprises’ fast reactors.

6.4 The fifty-year target

The paper plans on every warhead dismantled within the first fifty years, by 2081, and its uranium and plutonium in the fuel cycle. Whether that happens is decided by nations, not by NewVistas. NewVistas takes no political stance; its participants vote and hold office like anyone. What the communities offer is what made Megatons to Megawatts work: a steady commercial buyer, in this case for fuel made from weapons, with a use that turns every tonne into homes, schools and farms.

7. The crossing, stage by stage

The 250-year bridge, from the fuel cell to the atom in the building.

The bridge ends when the last building’s gas is shut off. That is within the 250 years if each gate is passed on time, and well before it if acceptance comes faster than this study assumes.

7.1 Fuel for the long age

Once the bridge is crossed, fuel is not the limit. The oceans hold about 4.5 billion tonnes of dissolved uranium, renewed from the rocks as it is drawn. Thorium is about three times as common in the earth’s crust as uranium. With breeding and recycling, fission can carry a world of twelve billion for many thousands of years on less land than a single farm belt’s towns occupy today.

8. The rule proposed

The study proposes this rule for the building plant:

The crossing. The building plant changes from gas to fission in stages, each opened by a published gate: first the industrial and logistics buildings, then the public buildings, then the apartment buildings as each building’s residents choose. Every core is fuelled at the factory, sealed, never opened in the community, and returned whole to the fuel enterprise; no plutonium enters a building core. Each building’s radiation readings are published live. The heat of the core is used in the building, none leaving while any use remains, and its waste tube is heated by the core directly. Fuel made from the uranium of dismantled weapons is preferred while it lasts. Bureau 23 publishes the gates and the record against them each decade; Bureau 13 holds the building-scale core as a standing research objective until it is licensed for every building.

Sources

• Lovering, Swain, Blomqvist and Hernandez, Land-use intensity of electricity production and tomorrow’s energy landscape, PLOS ONE, 2022.

• Breakthrough Institute, Updated Mining Footprints and Raw Material Needs for Clean Energy, April 2024.

• Our World in Data, Death rates from energy production per TWh (Markandya and Wilkinson 2007; Sovacool et al. 2016). Lifecycle CO₂: IPCC AR5 (2014) medians.

• IRENA and IEA-PVPS, End-of-Life Management: Solar Photovoltaic Panels, 2016. Wind blades: Liu and Barlow, Waste Management, 2017.

• Neutron Bytes, Four micro reactors hit criticality milestones by July 4th, July 8, 2026.

• Radiant, Kaleidos has left the building, August 2026.

• American Nuclear Society, NRC unveils Part 53 final rule, March 26, 2026.

• Beveridge & Diamond, NRC Proposes New Licensing Framework for Microreactors (proposed Part 57), June 2026.

• POWER, Centrus completes 900 kg HALEU delivery to DOE; ANS, Savannah River Site could produce 3.1 MT of HALEU.

• Gallup, Less support for solar, wind energy; more for nuclear, April 22, 2026; NucNet, US support for greater emphasis on nuclear is highest to date.

• Bulletin of the Atomic Scientists, Counting the dead at Fukushima, March 2026; UN finds ‘no adverse health effects’ from Fukushima disaster (UNSCEAR 2020/2021); Fukushima Daiichi nuclear disaster casualties.

• IEA, The Role of Critical Minerals in Clean Energy Transitions (2021), kg/MW and lifetime-adjusted t/TWh, with Bright New World and US DOE Quadrennial Technology Review (2015) steel, concrete and copper, as tabulated by the World Nuclear Association, Mineral Requirements for Electricity Generation.

• Öko-Institut, Rare earths: a bottleneck for future wind turbine technologies?, 2011 (160 to 200 kg of rare earths per MW, direct drive).

• Metals Focus and the Silver Institute, via FXStreet, Silver’s solar paradox, June 2026.

• Wiser et al., Assessing wind power operating costs in the United States, LBNL, 2019.

• pv magazine USA via Ground News, the cost gap between recycling and disposal, June 2026; Yale E360 via Mother Jones, April 2023 (under 10% recycled; landfill and hazardous-waste rules).

• Liu and Barlow, Wind turbine blade waste in 2050, Waste Management, 2017; WindEurope call for a Europe-wide landfill ban on blades.

• van Zalk and Behrens, The spatial extent of renewable and non-renewable power generation (Energy Policy, 2018), median power densities, via Surface power density; WHO, Household air pollution and health.

• Shaner, Davis, Lewis and Caldeira, Geophysical constraints on the reliability of solar and wind power in the United States, Energy & Environmental Science, 2018, via Carnegie Science.

• Reuters/LSEG via Zawya, Germany’s weak winds trigger record surge in gas-fired power, November 2024.

• Battery materials: Battery Design, key minerals in a battery; battery manufacturing emissions (Ecoinvent via ADEME), Nos Gestes Climat.

• Loss, Will and Marra, Estimates of bird collision mortality at wind facilities in the contiguous United States, Biological Conservation, 2013; Hayes, BioScience, 2013, via phys.org; Frick et al., Fatalities at wind turbines may threaten population viability of a migratory bat, Biological Conservation, 2017.

• The Wildlife Society, Ungulates turn away while humans embrace solar; USDA ERS, Agricultural land near solar and wind projects, 2024; Sonter et al., Renewable energy production will exacerbate mining threats to biodiversity, Nature Communications, 2020, via phys.org.

• Henneman et al., Science, 2023, via Scientific American, Coal power kills a staggering number of Americans; Power Engineering, Raising the bar for pulverized coal (over 99% particulate and 98% sulphur removal); coal with carbon capture, IPCC AR5 (2014), Annex III.

• AIChE 2009, Coal power generation with high efficiency and near-zero emissions (coal gasification with solid-oxide cells, all CO₂ captured); Lawrence Livermore National Laboratory, Direct carbon conversion (about 80% efficiency, pure CO₂ stream); Integrated gasification fuel cell with CO₂ near-zero emission and its challenges.

• US EPA, GHG Emission Factors Hub, 2024: natural gas 53.06, distillate oil 73.96, bituminous coal 93.28 kg CO₂ per million Btu.

• Federation of American Scientists, Status of World Nuclear Forces, early 2026.

• International Panel on Fissile Materials, The 2025 update on fissile material stocks, June 2026.

• Megatons to Megawatts Program; Centrus, Megatons to Megawatts.

• NewVistas: the Constitution (November 4, 2026); NV_Build, The Buildings of the Community (October 23, 2026).

• Calculations: firmed systems a 50-50 mix of solar and wind, batteries 2 kg of critical minerals, 7 kg and 60 to 140 kg CO₂ a kWh, 15-year life; world at twelve billion 130,000 communities at 150 MW; panels about 55 t a MW, capacity factor 20% for 25 years; blades 13 t a MW, wind at 33% for 25 years; coal about 0.5 t a MWh and ash a tenth of it; fission energy 82 TJ per kg of U-235; HEU at 90% blended to 19.75% with 0.2% tails; 35% conversion to electricity; community draw 1.5 to 2.0 kW a person, 92,160 people.