From the EditorThis week two desks found that the answer changed when the unit did.
Our cover story, from The Power Bill, reprices the firm power a US buyer can contract in 2026 per kilowatt that a grid operator will count, not per kilowatt on the nameplate. PJM's ratings for its 2029/30 auction count a reactor at 96% of installed capacity, a gas combined cycle at 75% and tracking solar at 7%. On that basis a nuclear restart and a new combined cycle, tied on nameplate, separate by about 22% in the restart's favour, and solar-plus-storage moves from second-cheapest to seventh. The cheapest options are also the ones not for sale: all three restarts are contracted, and a turbine ordered today is delivered in 2031. Among the briefings, Joule finds that a kilowatt-hour of US storage now has two prices, one for cells that qualify for the investment credit and one for cells that do not, and that the gap sits at the value of the credit.
Neither desk found new data. Each divided published costs by a different denominator, and we show the division so readers can check it.
Twelve features and eight briefings from across the network.
— The Ecliptic, Editorial Desk
Cover Story · Power & Energy · The Power Bill
The Power Bill — The Accredited Kilowatt Reorders the Firm-Power Stack — October 8, 2026
Priced per kilowatt that a grid operator will count, a nuclear restart costs about $2,000-4,000, a new combined cycle at quoted prices $3,200-3,470 and enhanced geothermal $6,100-7,800, while solar-plus-storage goes from second-cheapest on nameplate to seventh in PJM; the options that win on cost are sold out, and the ranking turns more on how hard a gas plant runs and on which market does the counting than on the price of gas
Which firm megawatt is cheapest per unit of reliability? This report puts the options a US buyer can contract in 2026 on one basis: capital cost per kilowatt of accredited capacity, using PJM's ELCC class ratings for the 2029/30 auction and MISO's indicative seasonal values, and levelised cost at one financing rate (7.7%, pre-tax) with stated capacity factors and fuel prices. Per accredited kilowatt a nuclear restart costs about $2,000-4,000, a combined cycle at the quoted $2,400-2,600/kW costs $3,200-3,470, an uprate $5,300-6,200, enhanced geothermal $6,100-7,800 and a new reactor $10,300-19,300 (The Power Bill estimates). The ranking changes in three places when nameplate becomes accredited: the restart and the combined cycle stop being a tie, solar-plus-storage falls from second to seventh in PJM, behind geothermal, and a four-hour battery ranges from $1,100 to $3,700 depending on the market. The gap between geothermal and gas is most sensitive to the gas plant's capacity factor, then to geothermal capital cost; gas price ranks third. The cheapest options are also the ones already sold.
Read the Graphic Edition →
Flagship Joule
Joule — Eight Reactors for $100 Billion: The Budget Prices No Learning, and Financing Decides the Power Price — October 3, 2026
The US–Korea nuclear framework budgets about $10,300–10,700 per kW of overnight cost for six AP1000s and two APR1400s, above an independent estimate for the next pair of AP1000s and more than twice the nth-of-a-kind figure; the same reactors cost about $104/MWh on sovereign-style capital and about $193/MWh on commercial capital, and the $20 billion reserve covers a fifth of the budget against a US record of overruns above 100%
At the end of September 2026 the United States and South Korea announced a non-binding $120 billion framework to build up to eight large reactors in the United States: six Westinghouse AP1000s and up to two Korean APR1400s, with $100 billion for overnight construction cost and $20 billion…
Flagship Beyond Solar
Beyond Solar — The £431 Megawatt-Hour — October 2, 2026
Over four full years the world's only long-running tidal array spent more on operations than it earned, at five certificates per megawatt-hour — about £431 of operating cost per MWh against the £72 the government's cost assumptions imply — and nine-tenths of that gap is reliability, paid for once in vessel campaigns and again in energy that was never generated
HEL-0030 found that the operating cost per megawatt-hour of the one commercial tidal array roughly doubled in a half-year with one turbine ashore, and asked whether tidal's cost gap is a capital problem or an availability problem. This report answers with the owner's own accounts — three annual…
Flagship The Power Bill
The Power Bill — The Calendar Is a Capacity Resource — September 28, 2026
PJM's planned-maintenance window reopened on 12 September under a rule that counts Wednesdays, and five days later the RTO was in emergency operations at 85% of its summer peak forecast with more than 35 GW of planned outages it had no right to recall — the planned-outage stock is a quantity the resource-adequacy models schedule away and the operating rules do not
Builds on CAL-0031, which found that PJM's two September 2026 emergencies hit the same wall — load plus outages of about 167 GW — from opposite sides, and asked how much margin the autumn maintenance window actually leaves. This report answers by reading the rule that sets the…
Flagship Star Power
Star Power — Commercial Traction Does Not Follow Capital — September 17, 2026
Three of fusion's five billion-dollar companies hold no plant site and no offtaker, while three companies below USD 700m hold site or utility commitments. The agreements track who the counterparty is, not how much money the developer has, and in six of nine cases the counterparty is also a shareholder.
TRP-0016 and TRP-0018 left one question open: does commercial traction in fusion follow capital or precede it? This report answers it by building a public register of every identifiable plant-site and offtake agreement held by a fusion developer and setting it against the capital table of the seventeen…
Flagship Parameter
Parameter — Energised Capacity: Half of an Approved Megawatt Is Drawing Load — October 6, 2026
The energised-capacity series already exists in grid-operator and utility filings. In Texas 9,456 MW of large load is approved to energise and 4,316 MW is observed drawing; in Virginia thirteen years of billing data put a mature data centre at 63% of its contract capacity. The gap between announced and energised has a second gap inside it, and it does not close.
PAR-0017 found that every capital-intensity figure for the AI buildout divides by announced capacity, and proposed building the energised series instead. This report assembles it from what is already public: ERCOT's monthly large-load observations, Dominion Energy's September 2026 data centre forecast filed with PJM, Georgia Power's commission-mandated quarterly…
Flagship Parameter
Parameter - The Six-Year Assumption - August 27, 2026
The depreciation schedule that carries AI capex rests on a residual-value tail its own owners are spending billions to compete away
Major cloud operators moved server and accelerator useful life from three or four years to six, cutting collective annual depreciation by roughly USD 18 billion and, on one 2024 estimate, from USD 39 billion to USD 21 billion — a 46% reduction in reported expense on an unchanged…
Flagship Engines of Tomorrow
Engines of Tomorrow — The Launch Price That Kills a Nuclear Rocket — September 17, 2026
DRACO was cancelled because launch got cheap. Solving for the launch price at which nuclear thermal propulsion stops paying shows the class is not dead. Its smallest viable mission got much larger, and the cislunar mission DRACO was named for was the first one priced out.
MACH-0013 and MACH-0019 recorded that DARPA cancelled its DRACO nuclear thermal rocket because falling launch costs weakened the return on its R&D, and both asked the same question: at what launch price does a nuclear thermal stage stop paying for itself? This report answers it. We built a…
Flagship Delta-V
Delta-V — Capacity Per Object: The Starlink V3 Transition — September 15, 2026
A V3 satellite carries roughly ten times the capacity of the satellite it replaces, which could shrink the broadband fleet by nine-tenths. The operator's own filings and guidance point the other way — ten times the satellites for a hundred times the bandwidth — and the hazard moves from object count to mass, automation dependence and the upper atmosphere.
DLV-0018 ended on an open question: does the V3 transition raise or lower the number of objects in orbit for a given capacity? This report answers it from primary disclosures — the Starlink V3 specification, the second-quarter 2026 earnings release and call, the Gen2 and Gen3 FCC record,…
Flagship Delta-V
Delta-V — One Operator Is Party to Ninety-Four Percent of Conjunctions — August 31, 2026
SpaceX's registration statement discloses 8,900 satellites, roughly 75% of every active maneuverable object in orbit, and over 1,000 automated collision avoidance maneuvers a day. At that share, space traffic coordination has quietly become one company's internal scheduling problem.
SpaceX's Form S-1 registration statement, submitted to the SEC in March 2026 and now public, contains an operational disclosure that has not been read for what it implies about space traffic management: as of December 31, 2025 the company operated over 8,900 Starlink satellites accounting for approximately 75%…
Flagship Chain Reaction
Chain Reaction — The Five-Year Clock Has Not Started — October 3, 2026
The only drafted plan to retire Bitcoin's quantum-breakable signatures takes five years from an activation that has no date and two prerequisites that are not activated either. Meanwhile the emergency workaround fell 79% to USD 67 a transaction in one September week — and on-chain it already weighs more than a NIST lattice signature would. The constraint on Bitcoin's quantum migration is the calendar and the block, not the qubit count
How long would Bitcoin need to retire its quantum-breakable signatures, and what binds that timetable? This report reads the two draft Bitcoin Improvement Proposals that define the migration, the March 2026 Google Quantum AI resource estimate, NIST's post-quantum standards and the first no-fork quantum-safe transaction, and converts them…
Flagship Chain Reaction
Stablecoins After the GENIUS Act
The economics of a regulated, interest-banned, Treasury-backed dollar
Does federal legalization make stablecoins a bigger or merely a more durable business? This report builds a transparent reserve-income model for the US payment-stablecoin market after the GENIUS Act, triangulating issuer disclosures (Circle Q1 2026, Tether Q1 2026 BDO attestation), regulator filings, and on-chain data. The headline finding:…
The Maple Grid · 2026-10-08The Maple Grid — Briefing — October 8, 2026
1. Ontario bought 600 MW of winter capacity outside its own auction 2. Churchill Falls now has a minority government on one side and a three-vote majority on the other 3. NB Power's twenty-year…
Joule · 2026-10-08Joule — A Kilowatt-Hour of US Storage Has Two Prices, and the Tax Credit Sets the Gap — October 8, 2026
1. The benchmark's $241/kWh is a 2024 China-cell cabinet with duties already in it; rebuilt from 2026 quotes, the imported block is about $145–181/kWh 2. The duty on a Chinese storage cell is 28.4%,…
Faraday · 2026-10-08Faraday — Briefing — October 8, 2026
1. **Closing the pack ledger.** The fleet study's bus figures imply a lifetime-average charge-utilization loss near 5.6%, against the under-2% quoted. The build-time capacity and resistance spread of the same packs would show how…
Engines of Tomorrow · 2026-10-06Engines of Tomorrow — The Solid-Motor Capacity Map, Class by Class — October 6, 2026
1. PAC-3 is the one interceptor motor with a second motor house, and the second source has no date 2. THAAD has the steepest ramp and a single boost-motor supplier 3. Standard Missile has…
Chain Reaction · 2026-10-06Chain Reaction — Briefing — October 6, 2026
1. The CFTC's crypto "proposal" is an advance notice — and it draws the line at leverage, not at spot 2. The SEC's self-custody permission is written to expire, one asset at a time…
The Reality Gap · 2026-10-05The Reality Gap — Critical Is Not Generating: The 2026 Microreactor Register — October 5, 2026
1. Five reactors are critical; one has run above 100 watts 2. The one electricity demonstration has three power ratings and no published output 3. Four of the five critical machines are not the…
The Gas Gauge · 2026-10-05The Gas Gauge — Turbine Slots Cap the Bull Case, Not the Agency Outlook — October 5, 2026
1. **Three manufacturers hold about 220 GW against 52 GW a year of output.** The firm part of that is 157 GW, or three years, which is the quoted lead time. Reservations add the…
Joule · 2026-10-05Joule — Briefing — October 5, 2026
1. FERC approved the price of PJM's backstop and sent back the question of who posts $1.5 million a megawatt behind it 2. Washington is pursuing transmission by two routes, and only one of…