One shaft. Not a city. Not a rocket pad.

A yawing solar collector.

Two thousand metres. A sun-facing wall sized from the equation. 300 MW AC at 1,000 W/m² and 35% sun-to-AC. Storage at the podium shifts energy. It does not create it.

Height: 2,000 m
Face, from the equation: 857,143 m²
Wall, full height: 429 m
Nameplate: 300 MW

Clear studies

Twenty seconds. The coast shaft yaws toward the sun, then the beam crosses the crown. The stills are the same object, sharper than the earlier studies.

Coast yaw, then the crown light path. One shaft. No city.

Concept sizing, not a measured plant.

Sharp study of a coastal shaft with a brass sun-facing wall, a parabolic crown, salt tanks, and the ocean behind. — Coast. Salt tanks and a dry hall. The sea is only the horizon.
Concept sizing, not a measured plant.

Sharp study of the same shaft inland, with dry-cooling fans, salt tanks, and an empty desert. — Inland. Dry coolers beside the tanks. No seawater.
Concept sizing, not a measured plant.

Looking up the sun-facing photovoltaic wall of the coastal shaft to the crown. — The lit face, from the ground. The wall is the collector.
Concept sizing, not a measured plant.

Close study of the crown: parabolic dish, concentrators, a prism, and multi-junction cells. — Crown. Mirror, concentrators, prism, cells. The path is the design point.
Concept sizing, not a measured plant.

Track the face

Drag the wall to yaw the shaft. Drag the sun across the day. The crown only delivers while the sun stays inside the acceptance angle.

Concept sizing, not a measured plant.

P = I × A_face × η

A_face = 300×10⁶ / (1,000 × 0.35) = 857,143 m²

Wall 429 m wide if it runs the full 2,000 m. The brief rounds the face to 857,000 m². Annual energy uses the equation, which is what lands on the published gigawatt-hours. Nameplate stays 300 MW.

E = A_face × H_beam × η

H = 2,800 kWh/m² → E = 840 GWh

CF = E / (300 MW × 8,760 h) = 0.320

Storage shifts energy. It does not create it. The pose above is not this annual total.

Concept sizing, not a measured plant.

Same face, different beam

Annual energy changes with direct beam. The face and the 35% do not. Alaska is on the chart so the miss is visible. It is not a site.

Annual energy, same face

Gigawatt-hours from E = A × H × η.

Concept sizing, not a measured plant.

Against the 300 MW nameplate

Capacity factor = E / (300 MW × 8,760 h).

Concept sizing, not a measured plant.

Site	Beam	Energy	CF	Call
Atacama	2,800 kWh/m²	840 GWh	0.320	Keep
Superb desert	2,500 kWh/m²	750 GWh	0.285	Keep
Mojave	2,400 kWh/m²	720 GWh	0.274	Keep
Alaska	1,085 kWh/m²	326 GWh	0.124	Do not build

Mojave and Sonoran

Atacama

Namib

Northwest Arabian Peninsula

Australia

Tibetan Plateau and Gobi edge

Keep the shaft only where direct beam is about 2,400 kWh/m² or better. The list is a siting rule. It is not six measured resource years.

Concept sizing, not a measured plant.

Two builds

Coast adds a dry hall and a closed seawater heat sink. Inland does not. A 100 MW hall draws 876 GWh a year.

Coast

Antofagasta · Baja · Namibian shore · Pilbara

Shaft, salt tanks, dry server hall.
Closed freshwater loop.
Titanium seawater exchangers.
Diffused warm return. Coolant never meets the ocean.

Inland

Mojave · Empty Quarter · Gobi

Shaft, dry cooling, salt storage.
No wet cooling.
No seawater loop.
No hall counted against this pattern.

96%One shaft here covers this share of a 100 MW hall (876 GWh/year). The desert case at 750 GWh is about 85%.

192%Two shafts. Enough annual energy for the hall, including the night, only if storage shifts the surplus.

0.67 MtOn the order of CO₂ avoided if this shaft displaces coal. The 750 GWh case is 0.6 million tonnes. Flat PV already does that per terawatt-hour.

Molten salt · shape of one clear day

Generation follows the sun. Delivered salt output is flatter and runs later. Totals match (7.60 = 7.60). Storage shifts energy. It does not create it.

Concept sizing, not a measured plant.

Concept study of a coastal shaft with salt tanks, a dry hall, and a low tree ring. — Coast study. Hall and exchangers stay onshore. The sea is a heat sink, not a coolant dump.
Concept sizing, not a measured plant.

Closed loop

Step through the coast loop. The hall water and the seawater do not share a pipe.

Air stays in the hall. The coolant loop is fresh water, closed.

Concept sizing, not a measured plant.

Diagram of a dry server hall, a closed freshwater loop, and titanium seawater exchangers with no mixing. — Reference diagram. Fresh water stays closed. Seawater is the other loop.
Concept sizing, not a measured plant.

Light path

Mirror, concentrator, prism, cells. The path is the design point. A missing layer is an incomplete crown, not a new efficiency.

1
Parabolic mirror
Aims beam into the concentrator.
2
Compound parabolic concentrator
Holds a wide acceptance angle. Tracking has to stay inside it.
3
Spectral prism
Splits the beam by wavelength.
4
Multi-junction cells
Each band lands on a matched junction.

Layers follow the switches in Track. With the CPC in, acceptance is the angle you set. Without it, the model holds the crown to ±6°. Both numbers are assumptions, not a measured optic.

Cutaway of a crown with parabolic mirrors, compound parabolic concentrators, a prism, and multi-junction cells. — Optical study of the crown. Plant output on this page stays on the 35% face equation.
Concept sizing, not a measured plant.

Tree ring

Native plants for dust and erosion. Nothing tall on the sun-facing wall. No lawn. No irrigated forest.

Plan of a circular podium with a native scrub ring and the collector wall marked do not shade. — Algarrobo or tamarugo. Only on an existing seep. The ring stops before the wall so the face stays clear.
Concept sizing, not a measured plant.

Ring on

Algarrobo or tamarugo. Only on an existing seep.

Mojave: mesquite, acacia, palo verde. Atacama coast: algarrobo or tamarugo, and only on an existing seep.

Toggle the ring in Track. The plan compass opens a gap toward the wall.

Concept sizing, not a measured plant.

What this is not

One shaft does not beat Three Gorges. Matching about 90 TWh at the desert case takes about 120 shafts.

Not in this concept: Tesla fleet yard
Not in this concept: Optimus ring
Not in this concept: Splashdown lagoon
Not in this concept: Starship
Not in this concept: Ozone platform
Not in this concept: A claim that one shaft beats Three Gorges

Capital to match ~90 TWh / year

The 120-shaft count uses the 750 GWh desert case. 36 GW × $1.61/W is about $58 billion.

Concept sizing, not a measured plant.

Published comparison, desert case: 120 shafts × $5 billion is about $600 billion, against about $58 billion for 36 GW of flat PV at $1.61/W. At 840 GWh a shaft, ~90 TWh takes about 107 shafts, about $536 billion at the $5 billion scenario. One shaft is not Three Gorges.

Concept sizing, not a measured plant.

Safety line

The shaft inherits solar’s published floor, about 0.02 deaths per TWh, against about 25 for coal. It does not beat operating solar, wind, or nuclear.

Published accident and air-pollution rates

Deaths per TWh, log scale. Context only. Not a shaft-specific study.

Concept sizing, not a measured plant.

One desert shaft, 750 GWh, displacing coal avoids on the order of 0.6 million tonnes of CO₂ a year. This site’s order-of-magnitude figure is 0.67 million tonnes. Flat PV already does that per terawatt-hour. Nothing here is an operating record.