Bench Talk

  Back in May 2015, Tesla released the Tesla Power Wall a 7kWh home battery. The idea being that it can be connected to the DC bus in a home and charge from renewables, mainly solar panels and wind turbines, when the production is high but the use is low. Later, they can be discharged during the evening peak times when load demand exceeds supply. Large energy storage packs are needed to bridge the gap between renewables production and load consumption. The interesting thing is that Tesla already produces a large battery that is on stand by for most of the day, in the form of electric vehicles (EVs). There has been a recent push to try and find a way to use electric vehicle batteries as power grid balancing batteries. If EVs could be used to balance the grid, it could lower the cost that electrical companies spend on spinning reserves, while further monetarily incentivizing EVs by having the electric companies compensate them for the spinning reserves provided by their EV’s battery pack. However, figuring out how to build a communications infrastructure that could grow and shrink based on available vehicles is complicated. Fortunately it is similar to the problems faced by microgrids. One of the more interesting control schemes attempting to solve this problem is cooperative control of multi-agent systems.

  The idea behind cooperative control is that if every agent, EVs in this case, is balanced with its neighbors, then the whole system would be balanced, much in the way a school of fish or crowd of people move together by responding to the movement of their neighbors. The further challenge that EVs present is that they interface with many users, each with their own set of needs. For the commuter, it is worth it to have their car at 70% state of charge (SOC) at the end of the day in exchange for additional money; the road tripper wants a full battery. There have been several different solutions to this challenge. The approach used by the researchers at the North Carolina State University to solve this problem was to give each owner two variables to set: pick up time, and desired SOC at pick up time. Tying those together with the car’s current SOC, a number is produced that then can be balanced with its neighbor. This is the kind of ingenious solution that allows us to continue moving forward in innovation and opportunity while being responsible with our resources. Products like the PowerWall will enable consumers to be part of this solution without owning an EV. Turns out that the second generation of the PowerWall is expected this year, as Elon Musk revealed recently, with a “further step change in capabilities.”