Series: Mining & Materials • Part 14 of 14
Scaling Civilization: Playing in Terawatts
The story so far: We dug the first clean pit and shaped it into a lake. We taught rocks to confess, printed sunlight, melted without smoke, moved mountains with batteries, moved products not dirt, made light from sand, snapped factories together, built objects up to supercomputers, closed every loop, and designed towns to love their lakes. Now we zoom out: how many terawatts can we build — calmly, quickly, beautifully?
Today’s mission
Define a terawatt in atoms, land, ships, crews, and weeks — not slogans.
Publish pre‑calculated scenarios for PV, storage, steel, glass, copper, and compute loads.
Show the clone math: factories that build factories until sunlight is our default fuel.
What a terawatt means (and why we’ll build many)
Terawatt cheat sheet (PV‑centric)
| Quantity | Planning value | Notes |
|---|---|---|
| Annual energy / TWp | ~1.6–2.0 PWh/yr | Climate & tilt dependent |
| Average power | ~180–230 GW | From energy ÷ 8,760 h |
| 12 h storage pair | ~2.2–2.8 TWh | Avg GW × 12 |
| Area (ground mount) | ~16–22 k km² | 1.6–2.2 ha/MW |
| PV modules mass | ~45–60 Mt | ~45–60 t/MW |
Ranges keep us honest across latitudes, trackers, and BOS design.
The simple why
- Electrons ≫ fuels: we’d rather move wires than mountains.
- Clean heat: furnaces and kilns listen to electricity (Parts 4–6, 9).
- Predictable load: compute & factories give us the steady baseload that storage loves (Parts 10–12).
Clone math — factories that build factories
Seed → snowball (PV factories, 1 GW/yr each)
| Calendar point | Factories alive | PV capacity/yr | Comment |
|---|---|---|---|
| Month 0 | 1 | 1 GW/yr | Seed factory (Part 3) |
| Month 12 | 4 | 4 GW/yr | First clones (Part 10) |
| Month 24 | 16 | 16 GW/yr | “Snowball” cadence |
| Month 36 | 36–64 | 36–64 GW/yr | Crew & pod limited |
| Month 60 | 150–250 | 150–250 GW/yr | Regional clusters online |
We cap growth with people/pods, not imagination; quality stays boring and high.
Clone kit bill (per 1 GW/yr PV factory)
| Pod | Count | Avg load | Shell area |
|---|---|---|---|
| Power PP‑20 | 3 | ~60 MW | — |
| Water WP‑500 | 2 | — | ~180 m² each |
| Heat HP‑20 | 1 | — | ~400 m² |
| Line pods | 12 | — | ~1,200 m² each |
| Controls + People | 1 + 3 | — | QA + labs |
This is the same Lego grammar we used across the series (Part 10).
How do we avoid a quality cliff while scaling fast?
Pods carry the skill; sites carry the concrete. Every pod is tested at the seed shop, serialized, scanned at setdown, and commissioned with a script. We scale the boring part — checklists — not risk.
Atoms per terawatt (what we actually move and melt)
PV hardware per TWp (ground mount)
| Item | Per MW | Per TW | Notes |
|---|---|---|---|
| Modules (mass) | ~45–60 t | ~45–60 Mt | Glass+frame (Part 9) |
| Mount steel/Al | ~60–100 t | ~60–100 Mt | Galv. steel + Al rails |
| Copper | ~1.2–2.0 t | ~1.2–2.0 Mt | Strings → inverter |
| Glass area | ~5,000 m² | ~5,000 km² | Low‑iron (Part 9) |
| Area | 1.6–2.2 ha | 16–22 k km² | Trackers, spacing |
Per‑TW totals spread across regions and years; we ship shapes (Part 8), not dirt.
Factories to feed that TW
| Line / Campus | Unit output | Units for 1 TW | Notes |
|---|---|---|---|
| Solar glass campus | ~1 Mt/yr | ~45–60 | Feeds modules & façade |
| Mini‑mills (steel) | ~1 Mt/yr | ~60–100 | Sections + coil (Part 5) |
| Al extrusion plants | ~0.2 Mt/yr | ~100–200 | Rails, frames |
| Copper refinery/EW | ~0.5 Mt/yr | ~3–5 | Busbars, cables |
| PV factories | ~1 GW/yr | ~1,000 | Or 200 @ 5 GW/yr clusters |
These units are pods in disguise (Part 10). We multiply calmly, not chaotically.
“Isn’t that a lot of steel and glass?”
Yes — which is why we make them with electrons (Parts 4–6, 9). The mod‑kit mini‑mills and glass lines exist to digest this exact workload, powered by the PV we already made (Part 3).
Land, water & neighbors (room for birds and ballgames)
Land math (context, not excuses)
- Per TW: ~16–22 thousand km² of PV meadows.
- Share of global land: ~0.01–0.02% (order‑of‑magnitude context).
- Dual‑use: PV fields as meadows, grazing, pollinator corridors (Part 13).
Water & lakes
- Process loops: 85–95% recycle in plants (Part 12).
- Lakes: seasonal buffers + trails + habitat (Part 13).
- Storms: bioswales + wetlands before the lake.
Storage & stability (keep the lights politely on)
Rules we actually use
- PV‑min (MWp) ≈ Avg MW × 5.14 (5.5 PSH, 85% DC→AC) — see Parts 3, 10–12.
- Storage (MWh) ≈ 12 h × Avg MW for calm operations.
- Overbuild: 1.5–2.0× PV to share with neighbors and shorten clone cycles (Part 10).
Example pairings (pre‑calculated)
| PV size | Avg power | 12 h storage | Where it fits |
|---|---|---|---|
| 1 TWp | ~180–230 GW | ~2.2–2.8 TWh | Regional grid |
| 100 GWp | ~18–23 GW | ~220–280 GWh | Nation‑scale hub |
| 10 GWp | ~1.8–2.3 GW | ~22–28 GWh | Mega‑campus + city |
Storage can be batteries, thermal, pumped, or fleet packs (Part 7). We pick the calmest mix.
Why does compute make storage easier?
Racks run 24/7 at steady power (Part 11). That stable appetite lets PV+storage operate predictably; waste heat warms blocks and homes (Parts 9, 12–13). A calm grid is a cheap grid.
Shipping & flows (move shapes, not mountains)
TEU & rail (sanity checks)
| Bundle | Per 100 MWp | Per 1 TWp | Notes |
|---|---|---|---|
| Solar farm kit | ~1,000–1,600 TEU | ~10–16 M TEU | Distributed across regions |
| Rail steel | ~6 kt / 50 km | Scales with corridors | Electrified (Part 8) |
| Modules | Ship short distances | Local finishing | We build near demand |
We avoid global module caravans by cloning factories (Part 10). Atoms stay near their destiny.
Trucks, rail, ropeways
- Mega vans (200 t): 3–5 MWh packs, flywheel peaks (Part 7).
- Rail spine: 0.04 kWh/t‑km planning (Part 8).
- Conveyors/ropeways: where roads don’t make sense (Part 8).
Crews & training (jobs with clean hands)
People per clone (typical)
- PV factory 1 GW/yr: ~300–500 FTE
- Glass line: ~250–400 FTE
- Mini‑mill 1 Mt/yr: ~600–900 FTE
- Compute 20 MW hall: ~80–150 FTE + support
Training spine
- Each campus ships a People Pod first: safety, clinic, classroom (Part 10).
- Digital twins for lines; practice on virtual steel before hot steel.
- Apprenticeships tied to pods: electricians, riggers, controls, QA.
Roadmaps (2, 5, 10‑year — pick your speed)
Two‑year “Kick”
- Clone PV to ~16 GW/yr (from 1 GW seed).
- Stand up 4–8 glass lines, 4–8 mini‑mills.
- Deploy 5–10 GWp PV meadows at mines & towns.
- Start 2–3 lake towns (Part 13).
Five‑year “Lattice”
- 150–250 GW/yr PV capacity in three regions.
- 20–30 glass campuses; 20–30 mini‑mills.
- Regional storage to ~0.5–1.0 TWh.
- 10–20 towns; first coastal hub.
Ten‑year “TW Habit”
- ≥1 TW/yr PV clone rate across continents.
- Glass and steel output paced to PV needs.
- Compute halls heat whole districts (Part 11).
- Campus loops so boring they’re invisible (Part 12).
“Is this just curves on a slide?”
No: every number here traces back to pods and plants we’ve already laid out — PV lines (Part 3), furnaces (Parts 4–6), logistics (Part 8), glass (Part 9), clone kits (Part 10). It’s a construction plan, not a mood.
Pre‑calculated global scenarios
Scenario A — 1 TWp/yr buildout for 10 years
| Metric | Value | Notes |
|---|---|---|
| PV added (10 y) | 10 TWp | Even cadence |
| Annual energy @ 1.7 PWh/TW | ~17 PWh/yr | Once installed |
| 12 h storage paired | ~22–28 TWh | At full effect |
| Steel for mounts | ~600–1,000 Mt | Over the decade |
| Glass | ~450–600 Mt | Module glass only |
| Copper | ~12–20 Mt | Arrays to inverters |
These decade totals call for dozens of glass campuses and mini‑mills — exactly our kit (Parts 5, 9).
Scenario B — 5 TWp/yr “sprint” (years 5–10)
| Metric | Value | Notes |
|---|---|---|
| PV added (5 y) | 25 TWp | Clone fever |
| Annual energy @ 1.7 PWh/TW | ~42.5 PWh/yr | From the sprint alone |
| 12 h storage paired | ~55–70 TWh | Regionally dispersed |
| PV meadow area | ~0.4–0.55 M km² | Dual‑use land |
“Sprint” requires matured pod supply and trained regional crews (Part 10).
Scenario C — Balanced lattice (electric industry + towns)
Assume a region targets 500 GWp PV, industry anchored by 5 steel mini‑mills, 5 glass lines, 2 compute halls.
| Item | Planning value | Comment |
|---|---|---|
| Avg power | ~90–115 GW | From PV |
| Storage (12 h) | ~1.1–1.4 TWh | Battery + thermal mix |
| Steel output | ~5 Mt/yr | Local beams/coil |
| Glass output | ~5 Mt/yr | Modules + façade |
| Compute | ~40 MW | District heat anchor |
| Lake towns | ~4–8 | Each 5–25k people (Part 13) |
This is one tile in a world lattice. Copy, rotate, paste.
Tap‑to‑open Q&A
“Where do the materials come from — do we have enough?”
We sized clean mines‑as‑factories in earlier parts: ore is sorted (Part 2), smelted without smoke (Parts 4–6), and shipped as shapes (Part 8). Steel and glass dominate PV hardware mass; both are easy to scale with electricity. Copper needs care but is measured in single‑digit Mt per TW — manageable with recycling (Part 12).
“Won’t land be the bottleneck?”
Dual‑use PV meadows, rooftops, parking, canals, and brownfields add up. At ~16–22k km²/TW ground‑mount, we’re talking hundredths of a percent of land — arranged thoughtfully around towns and habitats (Part 13).
“How do we keep this pleasant to live next to?”
Electric motion, enclosed lines, covered conveyors, quiet yards, dark‑sky lighting, public dashboards (Parts 7–9, 12–13). We design for birds, ballgames, and bedtime.
“What’s the hardest part?”
People. Which is why we ship People Pods first, over‑invest in training, and let pods carry expertise so local teams can build careers without leaving home (Part 10).
Appendix — Cheats, conversions, & cross‑links
Quick conversions we used
| Thing | Rule of thumb | Used in |
|---|---|---|
| PV energy per TWp | ~1.6–2.0 PWh/yr | All scenarios |
| PV area | 1.6–2.2 ha/MW | Land tables |
| Storage pairing | 12 h × Avg MW | Storage tables |
| Rail energy | 0.04 kWh/t‑km | Logistics (Part 8) |
| E‑truck (site) | 0.25 kWh/t‑km | Campus flows (Part 7) |
Cross‑links (this series)
- Part 1 — Lakes & first hole: water buffers and future parks.
- Part 3 — Solar seed factory: where the snowball starts.
- Parts 4–6 — Furnaces & metals: electrons, not smoke.
- Part 8 — Transport: ship value, not dirt.
- Part 10 — Lego factories: pods & ports.
- Part 12 — Circular loops: “waste” with a job.
- Part 13 — Towns: life around the lake.
Final note: We never asked for permission from physics — only clarity. Pick a rock, sort it, melt it with sunlight, ship shapes, stack parts, and tell the lake you’ll be back with a boardwalk. That’s the plan. Let’s build.