may be of interest

[rf...@sonic.net]

A letter in the Financial Times of Feb. 5, 2026.  While not directly addressing water issues, it has been established that AI data centers require copious amounts of water.

_______________

"...By tethering the US grid to a massive 'petrotech' expansion of gas-fired power to meet AI demand - a nearly 50 percent increase in current capacity if all projects go ahead - we aren't lowering costs, we're importing systemic volatility into every American household's utility bill.

If gas power is overbuilt to meet AI demand that does not materialize, then consumers will ultimately end up carrying the can.

As AI drives a record surge in domestic gas demand, it will compete directly with the US push to increase liquefied natural gas exports.  This 'double-pull' on supply - exporting it abroad while increasing domestic use for AI - will inevitably exert upward pressure on domestic gas prices.  In fact, the Energy Information Administration, which compiles official energy statistics for the US government, is already forecasting a 33 per cent price surge by 2027.

Furthermore, GE Vernova's* own 21 per cent spike in turbine costs proves that the AI boom is inflating the capital cost of the grid itself.  We are building out a 30-year dependency on a volatile commodity at the exact moment its capital costs are peaking.

The 'petrotech' build-out may offer a short-term power fix for Big Tech that socializes the risk and privatizes the profits.  Instead of 'averaging out' costs, this build-out forces the public to subsidize the infrastructure for a private AI boom while remaining hostage to volatile gas markets.

-Jenny Martos, Project Manager, Global Oil and Gas Plant Tracker, Global Energy Monitor - Richmond, VA U.S.

-within italics added

*Not disclosed but presumably a General Electric-related operation.