South Korea’s second largest earthquake has now been linked by two different studies to the injection of water deep below the surface at a new geothermal power plant.

Perched on South Korea’s southeast coast and far from grinding tectonic plates, Pohang is an unlikely spot for a big earthquake. Before the geothermal plant’s two wells were drilled, there had never been an earthquake there of any significance, says Kwanghee Kim, a seismologist at Pusan National University in Busan, South Korea, and lead author of one study. But while Kim was monitoring the aftermath of an unrelated earthquake in 2016, he began to detect rumbles from Pohang. That prompted his lab to deploy eight temporary seismic sensors at the site, which were finally in place on 10 November 2017. He expected any quakes to be small—after all, the largest previous quake tied to enhanced geothermal power, in Basel, Switzerland, was just 3.4 in magnitude.

It took only 5 days to be proved wrong. “The Pohang earthquake was larger than any predicted by existing theories,” Kim says. Although some initial measures placed the source of the quake several kilometers away from the plant, Kim’s network revealed that the earthquake, and several of its foreshocks, all began right below the 4-kilometer-deep well used to inject water into the subsurface to create the plant’s heating reservoir. Indeed, it appears likely that the well’s high-pressure water lubricated an unknown fault in the rock, causing it to slip and triggering the quake, Kim says.

A second paper, by European scientists who used regional seismic data, reinforces the South Korean team’s results, in particular its shallow depth. That study also points out that an earlier 3.1-magnitude earthquake also took place near the well bottom, increasing the odds of a common source. Satellite measures of shifts in the surface after the November 2017 quake support that idea, says Stefan Wiemer, the second study’s lead author and director of the Swiss Seismological Service in Zurich. It’s clear the locked fault was storing energy that was waiting to be released, Wiemer says. “If that fault would have gone next Tuesday or 50 years from now, we’ll never know.”

The article notes that scientists had previously concluded that injecting water underground for geothermal purposes was okay (since it reduced use of fossil fuels) while doing the same for fracking (to obtain and use fossil fuels) was bad.. The data here actually suggests just the reverse, since fracking has never produced an earthquake as large as the 5.5 magnitude Pohang quake.