In 2022, a data center in Houston came to Gridmatic with an ambitious ask: they wanted to begin matching their energy consumption with zero-emission power sources on an hourly basis, in a fashion similar to the 24/7 clean energy goal announced in 2020 by Google.

And so over the course of the next year, the energy modeling company did just that. They used artificial intelligence to forecast the data center’s energy needs and the availability of clean energy to identify ideal sources of energy and avoid the over-buying that is typically associated with time-matching energy strategies. By the end of the year-long pilot project, the data center was successfully matching 80% of its power consumption with zero-carbon energy sources on an hourly basis, said David Miller, vice president of business development for Gridmatic.

The pilot was even more successful than anyone the team had hoped: the initial goal was to match about 60% of the data center’s hourly energy needs with clean energy resources, Miller said. But they did not ultimately find affordable resources to match all of the data center’s energy needs.