One of the most underrated capabilites of classical, large-scale hydropower plants is their ability to "black start" a dead electric grid. I'm not going to go into details on the blackstart process here, or why hydroelectric plants are better at it than other types of power plants (whether fuel-burning or renewable), but suffice it to say that hydroelectricity provides a significant portion, if not the lion's share, of blackstart capacity in many, if not most, grids worldwide.
However, not all hydroelectric plants can provide blackstart power. Clearly, "micro-hydro" systems relying on DC generation and inversion to AC are incapable of blackstarting a grid, size aside, as the inverter simply won't try. Some micro-hydro systems, as well as most larger "mini-hydro" systems into the dozens of kilowatts, use induction AC generators and capacitor banks instead of DC generators and inverters. These systems can't blackstart either, as the capacitor banks used will not provide enough reactive power to the generator for it to come up mains frequency and voltage on its own no matter how much you spin it.
This raises a question, though, given that the emphasis in modern hydropower development is on smaller-scale systems (tens or hundreds of kW), often configured in a "run of the river" fashion, over massive, multi-MW to GW investments with equally massive impoundments. How small can one make a hydroelectric plant while retaining its capacity to supply blackstart power to the grid it's connected to? Is there some lower limit below where the use of synchronous AC generators is no longer feasible? Or could one theoretically build blackstart capability into small hydropower plants, whether they are based on impoundment or "run of the river" techniques?
