Terrestrial renewables solved generation.
They haven't solved power.
Three structural constraints — intermittency, land, and storage — cap the ceiling of ground-based solar and wind. Each finding below is sourced. The numbers are current.
Intermittency costs grid operators $47B annually in storage and backup capacity.
Wind and utility solar deliver capacity factors of 25–35%. Every gigawatt of nameplate capacity requires a parallel gigawatt of dispatchable backup — gas peakers, pumped hydro, or battery banks priced at $350–600/kWh installed. The storage problem doesn't disappear at scale; it compounds.
LBNL Grid Integration Study, 2024. Includes curtailment losses, spinning reserve, and battery procurement.
Utility solar requires 75× more land per TWh than space-based collection.
A 1 GW terrestrial solar farm occupies 10,000–12,000 acres at average U.S. insolation. The same output from an orbital platform requires a 10 km² rectenna — sited on marginal land, dual-use with agriculture, and transmitting at night when terrestrial panels are dark.
NREL Land Use Data, 2023. Orbital rectenna footprint per Jaffe & McSpadden (2023), IEEE Spectrum.
Long-duration storage adds $180–240/MWh to levelized cost — eliminating terrestrial solar's price advantage.
4-hour lithium battery storage handles daily cycling. 100-hour storage for seasonal balancing — the real grid reliability problem — remains commercially unproven above 50 MW. Flow batteries and hydrogen round-trip at 45–60% efficiency. Space solar delivers baseload without the storage stack.
NREL ATB 2024, long-duration storage scenario. Flow battery round-trip per Form Energy technical disclosure.
The solution isn't a better battery or a larger farm. It's moving the collector above the atmosphere.
The engineering is settled.
The launch cost wasn't — until now.
Every subsystem in the Harvest architecture has been demonstrated in terrestrial or low-orbit experiments. The constraint was always $/kg to GEO. Starship changes the denominator.
Orbital Collector Array
Thin-film GaAs photovoltaic tiles bonded to a deployable tensegrity truss. Each 100m × 100m module launches folded, self-deploys at GEO. Thermal management via selective emitter coating — no coolant fluid required.
Microwave Transmission
Phased-array transmitter with 1-meter element spacing achieves ±0.003° pointing accuracy. Beam shaping limits edge power density to 1 mW/cm² — below ICNIRP occupational exposure limits at all rectenna boundaries.
Safety & Interference
Beam automatically defocuses within 200ms on loss of ground pilot signal. ITU coordination under WRC-27 agenda item 1.18 already underway. No ionospheric heating at 2.45 GHz — confirmed by NRL propagation modeling.
Ground Rectenna
Semi-transparent mesh rectenna mounted 3m above ground — allowing full agrivoltaic use beneath. Schottky diode array converts 2.45 GHz to DC at 85% efficiency. Modular 100m × 100m panels — installable without grid outage.
Grid Integration
DC output feeds directly to co-located AC inverter substation. Dispatchable at 1-minute ramp rate — faster than combined-cycle gas. Compatible with existing 345kV and 500kV transmission infrastructure. No new long-haul lines required.
Harvest Technical Reference Architecture v4.2, Feb 2026. All values subject to final PDR confirmation.
Levelized cost crosses grid parity
at the 12th launch.
The economics hinge on one variable: $/kg to GEO. At Starship's projected $200/kg fully reusable, the 24-platform constellation delivers power at $54/MWh — competitive with combined-cycle gas, without fuel price risk, without carbon liability.
LCOE model: 8% WACC, 25yr asset life, $200/kg launch cost at scale. Starship full reuse assumption per SpaceX Starship User's Guide v2. Grid parity threshold: $90/MWh (U.S. wholesale baseload average, EIA 2025).
Grid parity threshold row. Launch cost basis: $200/kg Starship fully reusable. Regulatory: ITU WRC-27 coordination assumed complete by launch 4.
FCC Part 25 (space station) + Part 18 (industrial microwave) filings complete for demonstration orbit slot. ITU coordination under WRC-27 Agenda Item 1.18 addresses SBSP spectrum allocation globally. FAA launch licenses follow standard commercial launch vehicle procedures — no novel regulatory category required. The path is mapped; it is not blocked.
Request the Full
Technical Brief.
The brief covers mission architecture in full, beam safety modeling, rectenna site selection methodology, launch manifest, financial model with sensitivity analysis, and regulatory timeline. 47 pages. No NDA required for initial distribution.