Ton of refrigeration to Petawatt

Before mechanical refrigeration, buildings were literally cooled with ice. A "ton of refrigeration" was the cooling you got from melting one ton of ice per day. When compressor-based AC arrived in the early 1900s, the ice-based unit stuck because the entire industry — contractors, building codes, ductwork sizing — was built around it. Telling a building owner "you need 200 tons of cooling" was intuitive when they used to order 200 tons of ice. The unit survived because switching costs exceed inconvenience costs.

Roughly 1 ton per 400–600 sq ft of office space, depending on climate, occupancy, glazing, and internal heat loads (computers, lights, people). A 50,000 sq ft office needs 80–125 tons. Data centers are extreme: they need 1 ton per 200–300 sq ft because of server heat. A single rack of GPU servers can require 5–10 tons of cooling alone. The Trump Tower in New York has about 2,600 tons of installed cooling capacity.

When outdoor temperatures exceed 45°C for months, every building runs AC at maximum capacity simultaneously — there is no "shoulder season." Dubai alone has over 1.5 million tons of district cooling capacity. These plants chill water at a central facility and pipe it underground to hundreds of buildings, achieving 40–50% better efficiency than individual rooftop units. The Pearl-Qatar plant in Doha runs 130,000 tons — cooling an entire artificial island. Without district cooling, the electrical grid in Gulf states would need to be 30–40% larger just to handle dispersed AC compressors.

The district cooling plant at The Pearl-Qatar in Doha has about 130,000 tons of refrigeration capacity — enough to cool a small city in one of the world's hottest climates. Dubai's district cooling network exceeds 1.5 million tons total across multiple plants. For a single building, the Venetian Macao resort has roughly 16,000 tons. These megascale systems use chilled water loops distributing cooling across kilometers of underground pipes.

A typical 40,000 sq ft supermarket needs 80–150 tons: roughly 40–60 tons for the sales floor AC, and another 40–90 tons for refrigerated cases, walk-in coolers, and freezers. The frozen food aisle alone can require 20–30 tons. Open-top refrigerated cases are notoriously wasteful — they dump cold air into the store, which the AC must then remove. Modern stores with glass-doored cases can cut refrigeration load by 30–40%.

It's a time trick. A petawatt laser concentrates a modest amount of energy (maybe 100–500 joules) into a pulse lasting 10–100 femtoseconds. Dividing a few hundred joules by 10⁻¹⁴ seconds gives you 10¹⁵–10¹⁶ watts — surpassing the Sun's 3.8 × 10²⁶ W is still far off, but these lasers do exceed total human power consumption by 100,000×. The catch: the total energy delivered is only enough to heat a cup of coffee.

Primarily for nuclear fusion research (compressing fuel pellets), particle acceleration (laser wakefield acceleration can produce electron beams rivalling billion-dollar synchrotrons), medical isotope production, and probing extreme states of matter found in stellar cores. The ELI (Extreme Light Infrastructure) project in Europe uses petawatt lasers to recreate conditions found in supernovae, helping astrophysicists study cosmic explosions in a lab.

Solar flares can briefly release 10–100 PW of electromagnetic radiation. The Chicxulub asteroid impact (the one that killed the dinosaurs) delivered roughly 4 × 10²³ watts during the few seconds of impact — about 100 million petawatts. Gamma-ray bursts top everything at 10²⁵–10²⁶ PW, briefly outshining the entire observable universe. Even supernovae "only" sustain about 10³⁶ PW for a few seconds at peak.

Building one costs $50–500 million. Operating costs are surprisingly modest per shot — each pulse uses only a few hundred joules (less than lifting an apple one meter), but the capacitor banks and cooling systems draw megawatts of continuous power. The NIF facility costs about $350 million per year to operate. Individual shots are "cheap" in energy terms but the infrastructure to achieve them is staggering.

In theory yes, but in practice current petawatt lasers are terrible weapons. They fire one pulse every few minutes to hours, require warehouse-sized buildings of equipment, and deliver total energy equivalent to a firecracker. Military-grade laser weapons focus on sustained power (100–300 kW continuous beams), not ultrashort pulses. A petawatt laser is a precision scientific scalpel, not a blunt instrument — brilliant for physics, useless for destruction.