1cFE · Techno-economic reference · Jul 2026

LCOE lever stacks for six fusion archetypes.

Every fusion architecture has a different route to ultra-low-cost electricity — and a different set of things that must all go right to get there. This reference maps the techno-economic levers behind the most favorable LCOE case of six representative fusion archetypes, spanning confinement families (magnetic, inertial, magneto-inertial) and fuel cycles (D-T, D-³He, p-¹¹B), against an ambitious ~1 ¢/kWh cost target.

The levers were surveyed account-by-account across the standard fusion cost-accounting structure (CAS10–90 and the CAS22 reactor-equipment sub-accounts) and cross-checked against the 1costingfe costing framework. Each lever is placed under its dominant cascading effect — reduced CAPEX, increased availability, increased net electric output, reduced OPEX, reduced fuel-cycle cost, or revenue offsets — because that cascade, not the account it lives in, is what determines how it moves LCOE.

How to read An archetype's best case = the Universal stack (levers common to all six: financing, regulatory, siting, fleet, and operations) + its archetype-specific levers.

Note: In practice, not all of these levers are likely to be realized simultaneously. This exercise simply identifies the maximum set of favorable assumptions that could improve a given reactor's LCOE.
§UUniversal stackapplies to all six archetypes
Reduced financing & schedule cost
  • FOAK→NOAK maturity: contingency → 0, all FOAK-anchored coefficients to NOAK values
  • Construction-duration compression (modular schedule de-risking; parallel module fab, serial site assembly)
  • Concessional financing / LPO loan guarantees; long-term PPAs lowering equity share of WACC
  • Favorable regulatory environment + Standard Design Certification / pre-licensing (indirect fraction 0.20→0.10; shorter licensing critical path; insurance 3× lower; owner & O&M relief)
  • Digital twin virtual commissioning (construction/pre-op runway −0.25–1.5 yr)
  • Factory acceptance testing shrinking on-site commissioning
  • Site banking of pre-approved sites
Reduced CAPEX — siting & fleet
  • Brownfield siting at retired coal/nuclear (land, permits, buildings, interconnect, cooling-water reuse) — turbine reuse only for Rankine concepts
  • Fleet-standardized n-of-a-kind buildings; shared RH tooling suite; fleet lifetime-spares pooling; owner-staff amortization across co-located units
  • Modular factory-build with on-site assembly only; AI-aided project controls / 4D scheduling
  • Subterranean/berm reactor building; low-seismic site selection
  • AI-native software-defined I&C
Increased availability / CF
  • Real-time digital twin predictive operation (availability +0.02–0.05)
  • Condition-based / performance-based maintenance and higher fluence-limit materials
  • Capacity factor pushed to 85–90 % (the target all availability hardware serves)
Reduced OPEX
  • Robotic maintenance + AI predictive ops staffing reduction (fuel-specific floors in archetype blocks)
  • Plant lifetime extension 60–80 yr (modest at commercial discount rates)
Revenue offsets
  • Process-heat co-product sales (district heat / HTSE H₂ / desal) — siting-contingent
  • Decommissioning bond reform; state-backed insurance
ACompact MFE, steady-stateD-T
Reduced CAPEX (machine-scale cascade)
  • HTS ≥20 T high-field magnets (P_fus ∝ B⁴) → compact envelope → cube-law reduction in containment/shield/crane-bay volumes, 2–3× smaller structure per MWe
  • High REBCO learning rate, < $50/kAm
  • LN2 / HTS-warm cryoplant ($200M-class → ~$30M), enabled by warm-magnet architecture
  • NOAK heating systems, aspirational cost + wall-plug efficiency per type: NBI $7.06 → $2.5M/MW, η_source 0.60 → 0.75 (photodetachment neutralizer replacing gas cell); ECRH gyrotrons $5.0 → $3.0M/MW, η 0.50 → 0.55 (multi-stage depressed collectors recovering spent electron-beam energy); ICRH $4.15 → $1.5M/MW, η_source 0.70 (solid-state amplifiers replacing tetrodes, ~50 % source-cost cut)
  • Alpha-heating dominance (Q > 20) minimizing installed auxiliary heating
  • Self-cooled high-T breeder (FLiBe/PbLi) collapsing coolant + breeder loops (−60 % primary-coolant term)
  • B₄C-W composite shield (30 % thinner at same dose)
  • NOAK vessel unit cost via standardized sector fabrication ($0.72 → $0.35–0.50/m³)
Increased net electric (denominator)
  • Topping cycle η_th → 0.60 with >700 °C salt coolant (or sCO₂ Brayton fallback at η 0.47)
  • η_couple 0.833 → 0.92–0.95 (three-strap ICRF, steered 1 MeV NBI, resonant ECRH)
Increased availability
  • Liquid-metal divertor surfaces (Li/Sn-Li): erosion cap removed, core_lifetime ~5×
  • Liquid/flowing blanket: no discrete solid-FW replacement, core_lifetime ~3×
  • High fluence-limit FW material (>20 MW-yr/m²) where solid surfaces remain
  • Sector-replacement vessel + robotic in-port welding (availability +0.05)
  • Advanced divertor geometry (Super-X/snowflake) buying CF for modest capex
Reduced fuel-cycle cost & risk
  • TBR ≥ 1 via Be/Be₁₂Ti multipliers + ⁶Li enrichment → zero external tritium exposure
  • Cheap ⁶Li enrichment (Hexium laser / ζ-V₂O₅)
  • DIR + membrane separation (fuel_recovery → 0.999, smaller tritium plant)
  • Burn efficiency >50 % via spin polarization / selective pumping (inventory & plant sizing)
  • NOAK tritium startup price ($30k → $10k/g)
Reduced OPEX
  • DT staffing to <90 FTE (om_cost $52 → ~$45M/yr)
Revenue offsets
  • Medical radioisotopes (Mo-99, Lu-177) in blanket flux — traded against TBR margin
  • Hg → Au transmutation via (n,2n) in blanket flux — traded against TBR margin (realistic ~$2–5M/yr; matrix flags as economically marginal)
BLaser IFED-T
Reduced driver CAPEX & recirculating power
  • Driver wall-plug efficiency improvements
  • Fast / shock ignition: e_driver → ~30 % / ~60 % of central-hotspot (fast ignition adds PW-ignitor line ~$500M/MJ)
  • Fuel magnetization (seed field) in a magnetized-ICF variant → suppressed thermal conduction lowers required compression energy (matrix magnitude of 5–10× is MIF-anchored; treat as upper bound here)
  • Improved driver–target coupling efficiency → more of each delivered joule absorbed into the implosion, shrinking required e_driver (and thus stored energy and recirc power) at fixed yield
Reduced recurring target cost (the archetype's dominant lever set)
  • Direct drive: hohlraum eliminated from target_unit_cost and lower wall-plug demand per unit yield
  • Target tolerance relaxation: wetted-foam (0.3×) → mass-produced injection-molded capsules (0.02×)
  • Automated target inspection replacing bespoke cryogenic precision QA
Increased net electric
  • sCO₂ Brayton or topping cycle on chamber heat (η_th up, CAS23/26 down)
Increased availability
  • Liquid-wall chamber (liquid blanket): FW self-refreshes, no discrete replacement cycle
  • Standardized replaceable-component interfaces for laser-line and final-optics swap (CAS72 stream)
Reduced fuel-cycle cost
  • TBR ≥ 1 + ⁶Li enrichment (chamber blanket); DIR fuel recovery; burn-efficiency gains via target physics; NOAK tritium startup price
Grid interface
  • Buffer batteries/flywheels smoothing pulsed output + switchgear derating (0.7×); SST/HVDC option
Revenue offsets
  • Medical radioisotopes (Mo-99, Lu-177) in chamber-blanket flux — traded against TBR margin
  • Hg → Au transmutation via (n,2n) — traded against TBR margin (marginal per matrix)
Gate notesNo magnets → HTS, cryoplant, LN2 levers N/A. Brownfield turbine reuse only if Rankine chamber cycle chosen.
CPulsed inductive FRCD-³He · Helion-class
Reduced fuel-cycle cost (the archetype's make-or-break lever)
  • High efficiency He-3 breeding + consumption
Increased net electric (DEC channel)
  • Inductive DEC (driver-in-reverse) >85%
  • Hybrid DEC + topping cycle on the thermalized fraction (+10–15 % overall η)
Reduced driver / pulsed-power CAPEX
  • Solid-state switch technology (IGBT/SiC → advanced solid-state, 0.4× switch cost)
  • Large-scale inductive energy storage as a potentially cheaper replacement to capacitor banks
  • Cap-bank component-level NOAK costs (bank + switches + charging split)
  • (Note: η_pin already 0.93–0.95 for pulsed-magnetic compression — at ceiling, no lever room per the matrix)
Reduced BOP & installation CAPEX (small-module advantage)
  • Skid-mounted modular BOP (per-module p_net ≤ 50 MW)
  • Buffer batteries/flywheels + switchgear derating for pulsed output; SST/HVDC grid integration
Reduced nuclear-overhead CAPEX (low-neutronicity dividend)
  • Risk-informed structural margins (reduced activation, no large tritium inventory)
  • Thin shield / minimal RH / low fuel-handling & licensing baselines fully credited
Increased availability
  • Standardized replaceable-component interfaces for formation electrodes & cap-bank modules (the CAS72 replacement stream)
Gate notesNo SC-magnet cryoplant levers (non-SC excluded automatically); no target factory (in-situ FRC formation); no blanket/TBR levers.
DSteady-state mirror / linearp-¹¹B · aneutronic
Increased net electric (DEC-dominant conversion)
  • X-ray PV brem harvesting, implemented as a topping cycle alongside thermal power conversion
  • Electrostatic DEC at aspirational $0.31M/MW base, if applicable
Reduced CAPEX
  • HTS high-field mirror coils (low 1.73 coil markup; compact linear machine)
  • LN2 / HTS-warm cryoplant ($200M-class → ~$30M)
  • ICRH at NOAK $4.15 → $1.5M/MW, η_source 0.70 (solid-state amplifiers replacing tetrodes, ~50 % source-cost cut; ferrite-tunable antenna on the coupling side)
  • Alpha channeling via ICRH-band waves → alpha-heating dominance, minimizing installed auxiliary heating power (enters the model as low p_input)
  • η_couple 0.833 → 0.88–0.92 toward the mirror ceiling
  • Skid-mounted BOP if module size ≤ 50 MW
Reduced nuclear-overhead CAPEX (aneutronic dividend)
  • Risk-informed structural margins (explicitly justified for aneutronic: no T inventory, no high-fluence activation)
  • Minimal shield (0.1× scale) + B₄C-W composite thinning; near-zero RH, decom, licensing baselines fully credited
  • No blanket, no tritium plant, no cryo-distillation — structural absences locked in
Reduced fuel & OPEX
  • B-11 at industrial NOAK $75/kg (chemical distillation at scale)
  • Burn efficiency gains against the bremsstrahlung power-balance constraint
  • pB11 staffing to <45 FTE (om_cost $24 → ~$17M/yr); mirror om_scale 0.85 (linear-machine ergonomics)
Revenue offsets
  • Ac-225 targeted-alpha-therapy production via the DEC alpha channel (8.7 MeV alphas uniquely suited; $100M–$1B+ market)
Siting flexibility
  • Dry cooling with minimal penalty (DEC-dominant → small thermal share exposed to the η_th hit) → inland/arid siting freedom
Gate notesNo divertor-erosion or blanket-lifetime levers (no breeding blanket); medical-isotope neutron co-products N/A.
EMagLIFD-T
Reduced driver CAPEX & recirculating power
  • Fuel magnetization (seed field) — the defining lever: suppressed thermal conduction cuts required driver energy 5–10×, propagating through the entire e_driver-scaled driver cost
  • η_pin 0.15 → 0.30–0.40 via solid-state pulsed power (SiC/GaN, Marx→LTD topology, low-inductance buswork)
  • Inductive energy storage (0.25× cap-bank $/J at multi-MJ scale)
  • Switch technology + cap-bank component splits at NOAK
  • DPSSL preheat-laser efficiency gains (0.10 → 0.15–0.20) on the laser_preheat term
Reduced recurring target cost
  • Relaxed-tolerance / mass-manufactured liner-target production
  • Automated target/liner inspection
Increased availability
  • Standardized replaceable-component interfaces for cap banks, etc.
  • Liquid-metal chamber walls where architecture permits (self-refreshing FW, RTL-remelt accounting)
Reduced fuel-cycle cost
  • TBR ≥ 1 + ⁶Li enrichment; DIR fuel recovery → 0.999; burn efficiency / spin polarization; NOAK tritium startup price
Grid interface & BOP
  • Buffer batteries/flywheels + switchgear derating for pulsed output
  • Skid-mounted BOP if module size ≤ 50 MW
Revenue offsets
  • Medical radioisotopes (Mo-99, Lu-177) in blanket flux — traded against TBR margin
  • Hg → Au transmutation via (n,2n) — traded against TBR margin (marginal per matrix)
Gate notesSeed coils are minor — no SC cryoplant levers; sCO₂/topping applies to chamber thermal cycle.
FZ-pinchD-T · shear-flow / Zap-class
Reduced driver CAPEX & recirculating power (the archetype's headline stack)
  • η_pin 0.15–0.20 → 0.30–0.40 (SiC/GaN solid-state switches replacing spark gaps/thyratrons; Marx→LTD; low-inductance buswork) + inductive energy storage
  • Cap-bank component-level NOAK costs (bank/switch/charging split)
Reduced CAPEX (inherent simplicity fully credited)
  • No magnets, no cryoplant, no laser — structural absences locked in
  • Compact simple cylindrical geometry: cube-law-small buildings, NOAK vessel unit cost via standardized fabrication, minimal primary structure
Increased availability
  • Improved instability mitigation via shear flow controls
  • Flowing liquid-metal wall/blanket: FW and electrode-adjacent surfaces self-refresh, eliminating the erosion-driven replacement cap (core_lifetime multiple)
  • Standardized replaceable-component interfaces for electrodes & pulsed-power modules (CAS72 stream)
Reduced fuel-cycle cost
  • TBR ≥ 1 + ⁶Li enrichment (liquid blanket doubles as breeder); DIR fuel recovery; burn efficiency / spin polarization; NOAK tritium startup price
Grid interface & BOP
  • Buffer batteries/flywheels + switchgear derating for pulsed output; SST/HVDC option
  • Skid-mounted BOP for ≤50 MW modules (+ installation & aux knockdowns)
Increased net electric
  • sCO₂ Brayton or topping cycle on blanket heat (liquid-salt blanket enables >700 °C path)
Revenue offsets
  • Medical radioisotopes (Mo-99, Lu-177) in flowing-blanket flux — traded against TBR margin
  • Hg → Au transmutation via (n,2n) — traded against TBR margin (marginal per matrix)
Gate notesSheared-flow in-situ formation → target-factory and target-cost levers N/A (manufactured_target=False zeroes C220108); if a target-based Z-pinch variant, import the MagLIF target block instead. Self-cooled-breeder lever folds into the flowing liquid wall entry above.