This kind of tech was actually investigated several years prior, at PNNL, where they experimented with using electric water heaters to provide grid frequency r

This kind of tech was actually investigated several years prior, at PNNL, where they experimented with using electric water heaters to provide grid frequency regulation: https://availabletechnologies.pnnl.gov/technology.asp?id=287 https://availabletechnologies.pnnl.gov/technology.asp?id=287 Soon after, I worked on an team that, inspired by PNNLs work, created a research demand-response system using a variety of real and simulated loads, including electric water heaters. IIRC we used the current Texas wind output as our input power signal. At least at the time, our finding was that a collection of smart loads could replace natural gas peakers for frequency regulation on the demand side, so much that the service (therefore economic opportunity of) frequency regulation would be met with a small percentage of actual loads. The real long term and large scale value of dispatchable loads was thought to be the ability to adapt to the intermittency of renewables. I think that is still the case.