Don't Boil the Ocean: A Resource Allocation Study

An order-of-magnitude thought experiment. Revised October 6, 2026.

When someone says “Don't boil the ocean,” they are saying: don't consume every available asset in the business to accomplish this one goal or project. A project can succeed at its assigned task while leaving the organization unable to do anything else.

Naturally, we have commissioned a study to determine precisely how much trouble this advice is saving us. The calculations are scientific. The proposal should probably have stayed in the meeting notes.

Eliminate the largest candidates

Working from largest to smallest by water volume, we first remove the bodies that fail our resource screen. The conclusion below applies to known conventional fuel stocks and a practical project using current technology; it is not a claim that heating them violates physics or that renewable energy could never accumulate over an unlimited time.

Volumes use NOAA's ocean and marginal-sea definitions. Rows represent separate candidates, not volumes to add together. Even the Mediterranean, the smallest here, exceeds the combined useful heat available from the study's global coal, oil, gas, and standing-wood benchmarks. Renewable alternatives would need a separate design for space, power, time, and heat loss; none is established here.

Procurement has confirmed that there is one Earth in stock. The request for additional Earths remains open.

First candidate with enough fuel: Lake Superior

Lake Superior is the first body on the descending list that passes an energy-inventory test, using coal, oil, or natural gas. That is our stopping point. The smaller Great Lakes—Michigan (4,920 km³), Huron (3,540 km³), Ontario (1,640 km³), and Erie (484 km³)—do not need separate studies to make the business point. Volumes: EPA's Great Lakes physical data.

“Enough fuel” is a necessary condition, not an engineering plan. No practical method for bringing the entire lake to a simultaneous boil is demonstrated here. The following calculation heats a fixed water inventory to 100°C at surface pressure; deep water would not boil in place at that temperature.

Use 5°C as an approximate annual, volume-weighted starting temperature for Superior. This is an analyst estimate, not a measured full-lake climatological mean; Great Lakes thermal research documents the importance of full-column structure. Assume density 1,000 kg/m³ and constant heat capacity 4,180 J/(kg·°C).

Water volume = 12,100 km³
Water mass   = 12,100 × 10^12 kg
Temperature rise = 100 − 5 = 95°C

Q = mass × specific heat capacity × temperature rise
  = 1.21 × 10^16 × 4,180 × 95
  = 4.805 × 10^21 joules
  ≈ 4,805 EJ ≈ 1.335 million TWh

Heat per m³ of water ≈ 110.3 kWh

The calculation excludes heat leakage, inflows, ice melting, and salinity or elevation corrections. A ±3°C starting-temperature error changes the result by approximately ±3.2%.

Vaporizing the water as well would require about 6.7 times this energy, using NIST's latent heat of vaporization. We stop at the boiling point. Phase Two has been canceled by Finance.

The fuel requisition

Prices below are illustrative USD assumptions, not quotations. Costs cover fuel only; the electricity row uses an assumed tariff. Construction, transport, operation, and environmental heat loss are excluded. Each row is an alternative source.

Assumptions: coal 25 MJ/kg at 800 kg/m³ and 80% efficiency, $100/tonne; oil 38.6 GJ/m³ at 85%, $1/liter; gas 38.6 MJ/standard m³ at 90%, $0.25/m³; wood 15 MJ/kg at 500 kg/m³ and 80%, $100/tonne; liquid hydrogen 8 GJ/m³ at 80%, $4/kg; nuclear 45 GWd/tonne uranium at 90% direct heat, 9,600 kg uranium/m³ pellets, $2,000/kg fabricated enriched uranium; electricity $0.05/kWh at 100% heater efficiency. Gas is measured at standard conditions; wood is solid volume. Sources: EIA fuel heat contents, DOE hydrogen properties, and IAEA's illustrative nuclear burnup.

The per-liter price looks manageable. This is how the lake gets you.

What would be left for everyone else?

Compare Superior's requirement with documented global stocks, using the same heat efficiencies:

Benchmarks: EIA coal, end-2021: 1,161 billion short tons; OPEC oil, end-2024: 1,567 billion barrels; EIA gas, January 2021: 7,299 trillion cubic feet; FAO wood, 2020: 557 billion m³; NEA/IAEA uranium, January 2023: approximately 7.9 million tonnes. Uranium assumes 4.5% enrichment, 0.711% natural feed, and 0.25% tails, requiring 9.22 kg natural feed per kg enriched fuel. These inventories are historical benchmarks, not all material in Earth's crust. Coal grade, wood moisture, and recovery losses make the comparisons approximate.

Wood has the advantage of being familiar and the disadvantage of running out before we finish. Lake Superior has declined the wood-fired option.

Renewable power avoids a finite fuel pile but still needs equipment, space, and time. At an assumed average 50 W/m² of useful solar heat, Superior's idealized 100-year heating bill requires approximately 30,400 km² of collectors, before environmental heat losses. This is an area-and-energy estimate, not evidence that a century of heating would bring the actual lake to a boil.

The business lesson

A project does not become sensible merely because the organization has enough assets to attempt it. Superior could consume most of the world's known oil or gas stock under these assumptions and still require an extraordinary heating system. “We have the resources” is a poor substitute for “we can afford to dedicate them to this.”

In a business, the equivalent resources are cash, people, equipment, leadership attention, and time. If one goal absorbs them all, payroll, customers, maintenance, and the next opportunity must wait outside while the project celebrates its milestone.

Don't boil the ocean: choose a goal that leaves the business able to do its other work.

The committee recommends a cup of tea. It meets the boiling requirement, uses existing technology, and leaves enough budget for biscuits.