The grid problem first. A microgrid — a campus, a hospital, a solar-plus-storage community — can disconnect ("island") during a main-grid outage and run on its own. The hard part is the reverse: when the main grid returns, the microgrid has to rejoin it. Reconnect while the two are out of phase and you get a violent current transient that can trip breakers or damage equipment. The maneuver has to be seamless.
ABB's grant US10044192B2, "Participation factor based method for resynchronization of microgrids" (issued August 7, 2018; inventors Ravindra Singh and Dmitry Ishchenko), is a documented method for doing it. The "participation factor" language points to the mechanism: when a microgrid has several controllable sources, the method assigns each one a share of the correction needed to align frequency and phase with the incoming grid — so they act in concert rather than working against each other. The CPC tags H02J 3/383 (renewable sources to a network) and H02J 3/387 (parallel operation / synchronization) confirm the domain.
Map it to deployment and this is the unglamorous enabler of resilient distribution. The marketing story is "the microgrid keeps the lights on during an outage." The engineering reality includes the reconnection at the end, and a clean resync is what makes the whole islanding feature safe to deploy on a real feeder. Utilities will not permit equipment that reconnects with a bump; the method in this patent is the kind that makes permission possible.
The systems point Marisol's beat keeps returning to: IP like this maps onto interconnection standards. Synchronization requirements are written into the rules that govern how distributed resources attach to the grid, and a resync method that coordinates multiple sources is the technical answer to a regulatory constraint. Patent and interconnection rule are two views of the same problem.
As always, the boundary: a method patent claims an approach, not a guarantee that any given microgrid resynchronizes flawlessly under all conditions. But the existence of a granted, participation-factor-based method shows the reconnection problem is solved in principle and protected in practice — which is precisely what turns islanding from a demo into a deployable feature.