Hacker News Reader

Comment by CydeWeys | Hacker News Reader

The crane idea appeals to me more because the infrastructure appears simpler. Rather than having to build a lot of rail and a whole switching infrastructure, you just install the one crane and a shitload of heavy concrete blocks for it to lift. Seems like it would be cheaper.

View on HN

Someone would need to run the costs per GWh stored and generated. To me it looks like once you built your rail loop you could run a lot of cars on it. Still, building instead quite a lot of cranes might be cheaper.

Rolling stock isn't cheap either, though. Not even remotely close to how cheap a concrete block is. Rolling stock typically costs in the millions per car. The rail might not even be the most expensive part.

The nice thing about the crane idea is you don't even need a lot of cranes, you just need a lot of stackable mass. The rail idea needs a lot of movable mass on wheels, and a lot of siding to store it on. It seems like the rail idea would need more land as well, plus it only works on hillsides.

You need enough cranes to smooth the dips in power generation as each crane finishes dropping one weight and has to go fetch another. The number needed would be proportional to the desired power output of the system (holding the size of the weights constant).

It looks like their prototype uses 6 cranes on a single central pillar, probably for this reason. It seems like with some basic algorithms you could get away with as few as 3 and still supply steady power as long as the cranes are tall enough.

So a few cranes?

Keep in mind these will be grid-tied too, so you only need the power to be smoothed across the entire grid; it doesn't have to be at a single location. So grid storage batteries elsewhere could play a role too.

You still want each individual power plant to provide as smooth a power output as possible because each dip puts some strain on the other plants in the network as they have to increase their output to accommodate.

You could combine these ideas, though, and have only a few cars that pull weights up/down the hill, with a crane to load it off. I'm not sure but it could probably allow for a far greater total storage and kWh/$.

Now you've got more moving parts though. You have both trains and cranes involved. I think it's probably simpler and cheaper just to have the one big crane than to do a mixed mode system.

I guess the only advantage the train approach has is if you can get a lot more height out of it -- imagine an elevation difference of 1 km from top to bottom, way more than you could get with a crane. This still uses a lot more land though and needs a lot more rail; just having several crane installations might still be simpler and use less land.

Reply on news.ycombinator.com