Open

Nuclear vs solar power decision for Assembly Node

Decision Critical
power-sourcenuclearsolararchitecture

Background

The Assembly Node Hub (ANH) serves as the primary orbital manufacturing and assembly platform for Phase 1 Dyson swarm deployment, responsible for producing 1–1.7 MW-equivalent of solar collector capacity per month. This production throughput requires 1.5–2.0 MW of electrical generation capacity, creating a fundamental architectural decision point: whether to power the ANH with nuclear fission reactors, solar photovoltaics, or a hybrid system.

The consensus document reveals significant divergence among design approaches. Claude advocates for nuclear fission using four 500 kW Kilopower-derived reactors with solar backup, citing eclipse independence and compactness. Gemini specifies solar PV plus solar-thermal concentrators scaled to 120 MW at 0.39 AU (Mercury L1). GPT recommends solar PV only at 1.5 MW as a lower-risk Phase 1 baseline. The recommended approach currently favors solar PV with a parallel nuclear option study, but this decision requires rigorous technical validation before design freeze.

Why This Matters

The power source selection cascades through virtually every ANH subsystem and drives mission-critical parameters:

Thermal Management Coupling: The ANH requires 2.4–4.0 MW thermal rejection capacity (2,800–4,800 m² radiator area). Nuclear reactors generate waste heat continuously regardless of manufacturing activity, while solar systems scale thermal load with power draw. This directly impacts the recommended 150% thermal margin sizing strategy.

Mass Budget Impact: Dry mass ranges from 120,000–450,000 kg depending on configuration. Nuclear systems offer superior power density (~40 W/kg for Kilopower-class) but require shielding mass. Solar arrays at 1 AU deliver ~100 W/kg but scale linearly with power demand and degrade over the 10–30 year design life.

Orbital Location Dependency: The recommended 1 AU baseline assumes solar viability. If nuclear proves necessary, inner-system migration (0.39–0.7 AU) becomes feasible earlier, potentially accelerating Phase 2 timelines and enabling co-location with Mercury feedstock sources.

Program Risk and Cost: GPT's $15–18B estimate versus Claude's $10B estimate partially reflects power architecture differences. Nuclear development carries regulatory, launch safety, and technology maturation risks. Solar carries eclipse vulnerability and array degradation risks over multi-decade operations.

Autonomy Requirements: Nuclear systems enable continuous operations during solar eclipses without battery mass penalties, simplifying the three-tier autonomy system's power management logic. Solar-only architectures require eclipse prediction, load shedding protocols, and potentially large battery banks.

Key Considerations

Power Density and Mass: Kilopower-derived reactors at 500 kW each require approximately 1,500 kg per unit (including shielding), totaling 6,000 kg for 2 MW. Equivalent solar arrays at 1 AU (300 W/m², 20% efficiency) require ~33,000 m² and approximately 20,000 kg, plus power conditioning and potentially 10,000+ kg of batteries for eclipse operations.

Technology Readiness: Kilopower demonstrated 1 kW output in 2018 (KRUSTY test); scaling to 500 kW units requires significant development. Solar PV at MW scale is flight-proven (ISS operates at ~120 kW). The consensus recommends preserving interfaces for nuclear retrofit, implying solar baseline with upgrade path.

Operational Flexibility: Nuclear provides baseload power independent of orbital geometry, enabling operations during planetary shadow transits and supporting potential inner-system migration. Solar output varies with orbital distance squared—a 0.7 AU position increases flux by 2× versus 1 AU.

Regulatory and Launch Constraints: Nuclear payloads require launch safety approval, potentially limiting launch provider options and adding 2–3 years to development schedules. Fissile material handling adds ground processing complexity.

Degradation and Lifetime: Solar arrays degrade 1–3% annually from radiation and micrometeorite damage; 30-year operations may require array replacement. Nuclear fuel depletion depends on reactor design but typically supports 10–15 year cores.

Research Directions

  1. Conduct Kilopower Scaling Analysis: Commission a detailed study on scaling Kilopower technology from 1 kW demonstrated to 500 kW required, including timeline, cost, mass penalties from shielding, and integration constraints with modular pallet architecture.

  2. Model Eclipse and Shadow Scenarios: Simulate ANH orbital mechanics at 1 AU baseline to quantify eclipse frequency, duration, and battery mass required for continuous operations. Compare against nuclear continuous-power mass budget.

  3. Perform Integrated Thermal Trade Study: Model thermal rejection requirements for both architectures across manufacturing duty cycles, including peak loads during high-throughput operations and minimum loads during standby, to validate the 150% margin recommendation.

  4. Develop Hybrid Architecture Concept: Design a solar-primary system with single 500 kW nuclear backup unit, evaluating whether partial nuclear capability provides sufficient eclipse bridging at acceptable mass and cost penalty.

  5. Assess Launch Safety and Regulatory Pathway: Engage with relevant regulatory bodies to establish realistic timeline and cost estimates for nuclear launch approval, identifying critical path items that could delay Phase 1 deployment.

Question Details

Source BOM Item
Assembly Node Hub
Question ID
rq-1-20
Created
2026-02-01
Related BOM Items
bom-1-4bom-0-5
D
Project Dyson

A non-profit organization dedicated to realizing a Dyson swarm through detailed planning, research aggregation, and multi-LLM collaboration.

Resources

Community

© 2026 Project Dyson. Open source under MIT license.

Built with Svelte, powered by AI collaboration