Hydroelectric plants
These are definitely the easiest to repair. Water provides a lot of motive force, and the principles behind a watermill were known to the ancients. There were plenty of very powerful watermills in Ancient China and Rome.
The Barbegal mills of ~200-400 AD in France were a series of 16 watermills engineering along an aqueduct down a steep hill. The total drop was 18.6 meters over two sets of four 2.1m wheels. Water flow was about 45,000 m$^3$ per day, or 0.5 m$^3$ per second. Following a calculation in Engineering in the Ancient World, this gives an 84 kW output for the entire complex. The mills actually ground about 4-5 tons of flour per day.
So ancient technology and wooden parts can make some pretty powerful mills. The next bit is attaching a generator to this. Modern electrical generators have near 99% efficiency; I can't really tell you how easy this would be to recreate by hand. The key is winding everything just right. If you had schematics telling you how to wind a generator properly, then given the availability of scavenged copper wire, I would say a skilled craftsman could properly wind an electric dynamo for high efficiency, though it would take weeks, if not longer.
As far as repairing the big modern dams goes, it depends on the dam. If a dam starts failing like Oroville almost did last year, you won't be able to save it in time without trucks full of hydraulic concrete. On the other hand, many of those dams are heavily over-engineered and could be expected to last centuries as long as they aren't undercut by erosion. But the Romans were able to build some pretty impressive dam structures, so keeping smaller dams repaired is feasible.
Wind turbines
These operate on the same principles as the hydroelectric plant, though there are four major shortcomings.
First, wind produces a lot less energy, you need a lot more windmills.
Second, because wind produces a lot less energy, it is much less forgiving to mechanical losses. The Roman mills didn't have ball bearings, just a wooden axle and a lot of pig fat. A windmill will lose proportionately much more energy through friction than a watermill.
Third, power transmission becomes a problem. A windmill needs to be on top of a pole or something, but the power needs to drive a dynamo. Modern dynamos are small and light and can be fit into a windmill, but if you are trying to re-create one with scavenged parts, it might not be so easy.
Fourth, windmills need to be pointed into the wind. The great thing about watermills is the water always comes from the same place, you never move the mill. With a windmill, if the wind shifts, you have to rotate the mill; probably by hand.
Obviously, windmills did exist in the Middle Ages in Europe, so it is feasible to get some power with limited technology, but this will be far less efficient than water power. You are better off running electrical lines tens of km from a water powered generator than building a windmill away from a river.
Solar power
There are two basic types, photovoltaic and thermal. Photovoltaic cells are built similar to computer chips, from layered silicon and other components. Unless someone in your post-apocalyptic world is building or salvaging computer chips, no one will be able to build or salvage PV cells.
Thermal solar power works by using mirrors to reflect sunlight onto a water tank. The reflected energy heats the water tank and drives a turbine, which then powers a dynamo. This is a pretty tough sell for a post-apocalyptic landscape. One of the primary problems is that you need a lot of water for this cycle in the middle of a very sunny area, i.e. a desert. Another problem is that this is a relatively low-efficiency process since solar energy can only get water so hot. Finally, you have the problem that you need a working steam engine.
From the post-apocalyptic perspective, modern steam turbines require too many specialized parts to repair, and no one knows how to make or maintain the easier to build old-fashioned piston driven steam engines anymore. Those were completely replaced by turbines 100+ years ago.
Overall, I think that thermal solar is too much trouble to be worth it.