> If you want to make sure regeneration gets captured, in the absence of enough traffic to ensure other trains in a power block are using it when you're generating it, adding batteries and ultracapacitors to power substations is a well-known practice.

Odd that you mention batteries at the powerhouse, because this almost what the NWP did at the beginning of its electrified suburban service, when it was the North Shore: installed peaking batteries at the main powerhouse (the NWP suburban trains didn't use regenerative braking). These were maintenance intensive and done away with fairly early.

With regenerative braking, no load on the output = no braking. Saving AC regenerated power requires conversion equipment to change it to DC for storage in batteries of some sort because AC can't be stored. Then it requires more apparatus to convert it back to AC when needed. Each time you convert power, there are losses, so it becomes a matter of economics. Will the energy savings pay for the capital investment in all this apparatus?

> Much bigger batteries can be used there than on the trains themselves, so much more of the regeneration can be captured and delivered to other trains.

This did work for the VGN, N&W, GN, and Milw because they all used rotating electrical apparatus, which works very easily both ways (power and braking). Essentially all you have to do is drive the traction motors above slip and arrange for the locomotive's control apparatus to regulate the current generated. However, if there is no load, such as another train drawing power, there is no braking. When there wasn't that other train to draw on the power, the VGN and L&N dissipated it in liquid rheostats at the powerhouse; the MILW sold the power back to the power company. Not sure what the GN did.