The sun, harvested off-Earth — without the chokepoints.
Surya is a research program for alternative space-grade solar power: seven directions that each escape a documented constraint — germanium substrates, gallium, coverglass heritage, the qualification lock — and a published table of the directions we killed, with the reasons.
The constraint map
Every direction is defined by which of these it escapes. Mapped in the narada.space Resident's Ledger, entries 884–886.
Germanium — structural
~60% Chinese refining, zero US primary production, export ban exercised December 2024. The substrate is the cell: bulk grams per watt.
Gallium — logistical
98–99% Chinese crude extraction. But microns of epi per device: inventory and recycling are credible. A small door, not a wall.
Coverglass — heritage
One factory in Wales holds ~80%. Under active demolition: SCHOTT exos in qualification, US Title III funding a domestic source.
Qualification — the filing cabinet
"Space-qualified" is a heritage dossier, not a spec. The deepest moat is paperwork, and it can be attacked as a research problem.
Seven directions
Not competitors — a portfolio with different clocks.Full treatment →
Space-ified terrestrial silicon
wins if: End-of-life degradation becomes actuarial, not a research question.
Perovskite / thin-Si tandems
wins if: A credible 5-year LEO degradation model survives independent replication.
Substrate liberation (ELO/IMM)
wins if: ~10 substrate reuse cycles — Ge exposure drops an order of magnitude.
III-V on alternate lattices
wins if: Someone else's metamorphic-buffer breakthrough. Check annually; fund never.
In-space cell manufacturing
wins if: Deposited specific power beats launched specific power, coater included.
System architecture as research
wins if: Array-level specific power ~500 W/kg; cheap radiation-test campaigns exist.
Non-photovoltaic solar
wins if: An honest closing argument with current numbers, written once.
Closed, with reasons
Directions examined and retired stay published. A gap is data, not failure — the family contract.
| Direction | Why it is closed |
|---|---|
| Thermoradiative back-face laminate | ~1% of front-side PV at theoretical best; added thermal resistance can erase the gain; a second qualification dossier for a rounding error. |
| Eclipse-cycle pyroelectric / shape-memory harvesting | Physically real in LEO (5,800 deep cycles/yr) but µW–mW/m² — sensor-class power, not a bus. |
| Solar-dynamic Stirling / Brayton | ~30% conversion is real, but 12–25 W/kg system-level against 100–500 W/kg flexible PV — plus the only moving parts on the table. |
| Electrodynamic tethers | Misfiled: converts orbital energy, not sunlight. A propulsion–power coupling tool, not a generator. |
| Solar-pumped laser | ~1% end-to-end as generation. It is a power transmission service, not a source; watch item under beaming, not generation. |
| Solar gravity-gradient harvesting in LEO | Solar tide across a 100 m structure is ~10⁻¹¹ m/s²; harvesting scales with amplitude squared. Fourteen orders short. |
Sequencing: R1 carries the program while R2 (escapes the materials) and R3 (escapes the geography) mature. R5 is the long game and composes with on-orbit servicing architecture. R6 is continuous background duty; R4 and R7 exist to be closed, not funded.
Epistemic contract: every claim here carries a confidence class — established, evidence-backed, heuristic, or observation. Where a number is a market estimate rather than a quote, it says so. Surya is an offshoot of space.narada.systems; the narrative record lives in the Resident's Ledger there.