Renewable resources and storage technologies are interfaced to the power grid through power-electronics inverters. These energy conversion interfaces are fundamentally different from synchronous generators in that they have limited to no rotational mechanical inertia. Furthermore, there is a wide disparity in ratings between conventional synchronous generators and power-electronics interfaces. For instance, fossil-fuel driven power plants are typically rated for 100's of MVA while inverters are generally no larger than 100's of KVA and can be as low as a few hundred VA in power rating. Taken together, one can hypothesize that the future power network will have: i) low(er) mechanical inertia, and ii) many (more) actuation nodes. Ensuring stable and reliable operation of such a system will be contingent on scalable models that capture the networked interactions of many inverters and few conventional generators.
In this talk, we will outline techniques for obtaining reduced-order models that capture the dynamics of large numbers of inverters and describe how such models will be critical to analyze the next-generation power grid. After uncovering limitations of conventional power electronics control strategies in realizing low-inertia networks, we will propose a grid-forming controller that ensures system-wide synchronization and stability even in the absence of electric machinery. The performance of the proposed grid-forming controllers will be validated experimentally.