When the power goes out after a storm, you expect a truck within a few hours. Someone climbs a pole, swaps a fuse or a trash-can-sized transformer, and you’re back on the couch by dinner. That’s about as much of the grid as most of us ever see.<\/p>\n
The transformer sitting in the substation that feeds your whole side of town is a different animal. It’s the size of a house, it’s full of oil, and if an earthquake cracks it, the Idaho National Laboratory says you’re looking at five years, maybe more, before a replacement shows up. No one had ever put one through a real earthquake to see what breaks first, and that’s a strange gap for something this expensive.<\/p>\n
So this fall, INL did exactly that. On September 22 and again on October 13, a 230,000-volt transformer, filled with its oil and still wearing its porcelain bushings, radiators and conductors, got bolted to the biggest outdoor shake table on Earth, about 10 miles east of the UC San Diego campus, and rattled around like it was sitting on top of a fault line.<\/p>\n
A three-phase, 230-kilovolt power transformer, the kind that steps high-voltage transmission power down before it heads toward a city. INL put the weight at 500,000 pounds when it first lined the unit up, and Jon Bender of WEGAI Research, the Boise engineering firm on the project, has described it as a decommissioned transformer that can still be energized. So it’s a retired piece of real grid hardware, not a mock-up.<\/p>\n
If you’re not sure what a transformer does, here’s the short version: electricity travels long distances at very high voltage because that wastes less energy, and a transformer knocks that voltage down (or up) using two coils of wire wrapped around an iron core. Everything else on the machine, the oil, the radiators, the ceramic bushings, exists to keep that core cool and keep 230,000 volts from jumping to somewhere it shouldn’t.<\/p>\n
The table it sat on is the Large High-Performance Outdoor Shake Table<\/a>, which UC San Diego runs for the National Science Foundation. The steel platform measures 40 feet by 25 feet and can carry 20 meganewtons, which works out to about 4.5 million pounds, so a half-million-pound transformer was frankly a light day. It moves in six degrees of freedom, meaning it can slide back and forth, side to side and up and down, and also pitch, roll and yaw, all at the same time. And because it has no roof, you can put something as tall as a transformer with its bushings sticking up on it without hitting a ceiling.<\/p>\n That last point matters more than it sounds like it should. Indoor shake tables have height limits, and transformer bushings are tall, skinny and exactly what everyone wanted to watch.<\/p>\n You may be thinking that a half-million-pound steel box full of oil is about the last thing an earthquake could hurt. And the tank itself usually does fine. The problem is the bushings, the hollow porcelain insulators that carry the copper or aluminum conductors out through the top of the tank to the overhead lines. According to Bender, the committee that writes seismic substation standards has spent decades flagging bushings as the most common earthquake casualty in a substation.<\/p>\n Why do they break? Resonance, basically. Every object has a frequency it likes to vibrate at, and that’s the same effect an opera singer uses to crack a wine glass. A steel tank full of oil has one, and the bushings bolted to the turrets on top of it have their own. When an earthquake happens to hit a frequency where the tank and the bushing swing in sync, the motion at the top of that porcelain gets amplified instead of damped, and porcelain doesn’t bend. It shatters, and you’ve lost the transformer.<\/p>\n INL’s answer, developed with WEGAI and the University at Buffalo, is a decoupler that bolts in where the bushing meets the turret and shifts the bushing’s natural frequency away from the tank’s. The team has said it’s meant to be cheap and retrofittable on transformers already in the ground, and it’s pursuing a patent. We don’t have a price yet. The September 22 runs were built around watching that decoupler do its job on a real bushing, with the industry guests<\/a> INL invited shuttled out to the site each morning to watch.<\/p>\n I’ll admit the decoupler is what I find pretty clever here. Utilities have argued for years about how to protect bushings, and a tunable bracket that goes in without touching the transformer itself is the kind of fix that could actually get installed rather than studied.<\/p>\n The October 13 session was the one INL billed as the nation’s first large-scale shake test of a base-isolated transformer. Base isolation is what’s done to hospitals and some high-end office towers in California: instead of bolting the building straight to the ground, you set it on bearings that let the ground move underneath while the building mostly stays put.<\/p>\n Bender has described the hardware here as a triple friction pendulum system from Earthquake Protection Systems, a California company whose bearings sit under a lot of those hospitals. The idea is simple enough to explain in a sentence: the transformer rests on sliders shaped like shallow dishes, so when the ground jerks sideways, the machine swings gently like a puck in a bowl instead of taking the hit head on. INL’s own release says this hadn’t been tried on a full-size 230-kilovolt unit before, though smaller tests have hinted it could take a lot of the load off the bushings.<\/p>\n That’s a big claim for a grid that, as far as I can tell, hasn’t had much of a plan for transformers beyond bolting them down harder and hoping.<\/p>\n Quite a bit, frankly. INL hasn’t published any accelerations from the runs, hasn’t said which earthquake records the table played back, and hasn’t said whether anything cracked. The release also skips the transformer’s power rating and who built it, which is a slightly odd omission for a test whose whole point is that the hardware was real. INL says the results are headed for the DOE’s Office of Scientific and Technical Information and for peer-reviewed journals, so we’ll get numbers eventually, and the stated goal is to feed them into industry standards.<\/p>\n The funding came through two DOE offices, the Office of Cybersecurity, Energy Security, and Emergency Response<\/a> and the Office of Electricity, and INL says dozens of utilities contributed. Bjorn Vaagensmith, the INL researcher leading the project, framed the goal as protecting “a critical, hard-to-replace piece of equipment,” and the five-year replacement figure is why a lab that normally spends its time on a reactor built to break fuel on purpose<\/a> is suddenly interested in substation hardware.<\/p>\n You can see why. When a Georgia reactor dropped offline<\/a> this month because of an electrical jolt on the grid side, the plant shut itself down the way it’s designed to and kept all its hardware. A substation that loses its transformer in a quake doesn’t get that kind of out, and there isn’t a shelf of spare 230-kilovolt units waiting for the day one earthquake hits several substations at once. INL says that scenario could leave a town dark for months, and there’s no truck for that. Chile and Japan, both of which have been through this for real, are watching the data along with the West Coast utilities.<\/p>\n INL’s Chandu Bolisetti called getting a full-size transformer onto that table “miraculous” back when the team first lined it up, and the lab announced the completed test<\/a> on October 14, 2026, the day after the last isolation run.<\/p>\n","protected":false},"excerpt":{"rendered":" When the power goes out after a storm, you expect a truck within a few hours. Someone climbs a pole, … <\/p>\nBushings are where transformers die<\/h2>\n
Then they put the whole thing on a pendulum<\/h2>\n
What we still don’t know<\/h2>\n