Ocean Thermal Energy Conversion (OTEC) is the dusty cousin of renewable energy: invented in 1881 by d'Arsonval, demoed in Cuba by Georges Claude in 1930, and never scaled past a few megawatts. The premise is gorgeous in its simplicity — the tropical ocean surface sits around 26°C, while water at 1000 m depth hovers near 4°C. That 22°C gradient is a heat engine waiting to happen. So let's push it: can we run all of North America (≈4 TW average) from the Gulf of Mexico and Caribbean?

The Carnot ceiling is brutal. Maximum theoretical efficiency is η = (T_hot − T_cold)/T_hot = (299 − 277)/299 = 7.4%. Real closed-cycle OTEC plants (using ammonia as working fluid in a Rankine loop) achieve about 3% net after parasitic pumping losses. Compare to a coal plant at 40% — OTEC is moving 13× more thermal energy per watt delivered.

The water flow problem. To extract 1 MW of net electric power, you need to dump roughly 33 MW of heat into the cold reservoir. With a 22°C ΔT and seawater's specific heat (4 MJ/m³·K), each cubic meter of cold water absorbs only ~88 MJ as it warms to surface temperature. That means ~0.4 m³/s of cold water per net MW, plus a similar volume of warm surface water.

Scale to 4 TW: you need 1.6 million m³/s of cold water pumped from 1 km depth. For context, the Amazon River discharges 209,000 m³/s. We'd need to pull eight Amazons up a kilometer of pipe, continuously, forever.

Pipe engineering gets absurd. A single 10 m diameter cold-water pipe (CWP) at 2 m/s flow carries 157 m³/s. We need ~10,000 such pipes — or equivalently, a fleet of ~1,000 plants each rated 4 GW with 10 CWPs apiece. Each CWP must survive currents, biofouling, and the fact that polyethylene at 1 km depth deals with 100 bar external pressure. Current state of the art: Makai's 1.55 m HDPE pipe tested in Hawaii. We're asking for a 40× scale-up, ten thousand times over.

Now the thermodynamic kicker — we'd cool the ocean. The Caribbean and Gulf together hold roughly 10²² J in the top 100 m above 4°C. Running at 4 TW (net) means dumping 130 TW of waste heat into the surface mixed layer. Earth's net radiative imbalance is currently ~460 TW globally. We'd add 28% to the planet's heat-disposal load regionally, in one basin.

Actually — it's subtler. OTEC moves heat from surface to depth, then radiates it skyward via evaporation. Net effect on the basin: surface cooling, deep warming, vigorous artificial upwelling. The upwelling brings nutrient-rich deep water to the photic zone, which is either a fisheries bonanza or an algal-bloom catastrophe depending on stratification dynamics.

The hidden win: that cold deep water, after passing through the condenser, is still only ~10°C. Pipe it ashore and you get free district cooling for every coastal city in the tropics. Honolulu's SWAC system already does this at small scale. At continental scale, you'd eliminate ~15% of US electricity demand (air conditioning) before you even count the OTEC generation.

Verdict: physically possible, economically ruinous (~$30 trillion in pipes alone at current $/kg HDPE), and ecologically transformative in ways we cannot model. But the Caribbean as North America's thermal battery? The physics doesn't say no.