I was just listening to a Planet Money episode about a consortium of tribes trying to build two windfarms. The killer is the grid interconnect costs. They had planned $1.2b to build the two windfarms, but the grid operator came back saying it would cost $875m just to connect those farms to the grid.
The standout fact from the episode is that the queue for projects waiting for approval to connect to the grid has a total capacity greater than the current total capacity of all power plants online. Now, that number is inflated because no doubt some investors have multiple projects and will, say, pull three the second the first project is approved. Still, it shows the scale of the problem.
Just the one grid authority in question has 600 projects in the queue waiting for engineering and cost estimate.
The current model is that if you want to connect and connecting your project would require upgrading lines and transformers, you pay all upgrades for as far out as needed, sometimes up to hundreds of miles. It basically makes individual projects untenable even if, in aggregate, if all projects were approved and all got built and all split the costs, it would perhaps be doable. Or if it were funded like highways where all users share the cost, it would be way easier to bring these projects online.