Read this brilliant piece on 23 June and Britain's near-miss grid crisis 🧵 dispatches.kilowatts.io/p/a-… Summary: → Gas & nuclear outages (partly because of hot weather) → Solar great, then gone → Wind forecasts way off → Demand forecasts too national → Interconnectors saved us — barely Gas and nuclear had huge outages that day — some sudden. Solar was brilliant, then died with the sun, exactly as it always will. We need way more batteries, especially in homes, to fill that gap. Wind forecasts were miles out — not because wind is unpredictable, but because the central models used to plan the system are. Demand forecasts were even worse: a single national number can't capture where power is actually needed. London fell 472MW that evening while the model expected it to rise 1,138MW. That's not a rounding error, that's a system flying blind on geography. Interconnectors saved us that night — but only after frantic phone calls, not because the market was built to use them well. With zonal pricing they'd work with the grid instead of against it every single evening. The real lesson of 23 June: a grid built on one national price for everywhere is guessing at a national scale about problems that are actually local. The crisis wasn't "not enough power" — it was too little power in exactly the wrong place, behind wires that were already full, while other regions had headroom to spare. A single price can't tell a battery in Kent to discharge and a battery in Scotland to wait — it just tells everyone the same story at the same time. Zonal pricing fixes that: it makes location visible in the price itself, so flexibility shows up where the grid actually needs it. Pair that with batteries everywhere — homes, street corners, substations — and a genuinely digital grid that dispatches by the second rather than buying its safety margin in one panicked burst at 2pm, and near-misses like this stop being near misses at all. We have the technology. We just need the market design to use it.