Is this not just pumped hydro but worse?

These were quite popular as a startup idea once. I remember Gravitricity (now defunct) being posted here: https://news.ycombinator.com/item?id=24414497

Publicly available data[1] on the pilot project in Nevada suggests a total of “50MW” generation capacity is planned across 10 rail lines, but the photos on the website seem to only show 1 set being built so far - and a claimed output of 5MW. The per-car mass of 720,000 lb (321 Tonnes) being lowered 229ft=70 Meters (510ft track length x sin(26.8) degrees) in Earth’s 9.81/ms^2 gravity field represents a maximum potential energy of only 220MJ, or 61 kWh per car. Reaching 5MW peak requires a car to be dispatched every 44 seconds. 10 cars would provide about 7.5 minutes of runtime - which matches the advertised 15-minute cycle length.

This all seems reasonable - but is a far cry from the performance of existing Pumped Hydrostorage plants which routinely exceed 1GW since the 1970s, and can run for several hours per cycle. They do require lots of Water and a mountain’s worth of elevation change, which limits the site selection, whereas this system seems to work with any open-pit mine.

It will be interesting to see if this technology can be made competitive with existing grid-stabilization techniques, and what challenges will be encountered along the way.

[1] https://www.sandia.gov/files/ess/uploads/2021/LDES/Russ_Weed...

Related:

Fortescue slashes electric train program but insists zero emissions 'on track'

September 04 2025 - https://www.boilingcold.com.au/fortescue-slashes-electric-tr...

  Three years after Andrew Forrest pressed go to develop an electric "Infinity Train," most of the experienced engineers who joined Fortescue's zero-emissions crusade are laid off as the miner goes back to the drawing board on how to have fossil-fuel-free locomotives by 2030.

  The engineers concluded that battery electric locomotives may be able to haul vast amounts of iron ore, eliminating 10 per cent of Fortescue's emissions, but the knock-on effects on its immense $21 billion a year integrated mine to rail to port iron ore business were unacceptable.

And: https://zero.fortescue.com/en/case-studies/infinity-train .. 404 Page not found.

Looked good for a while there: Fortescue rides the Infinity Train - https://www.electrive.com/2025/07/01/fortescue-rides-the-inf...

   engineering studies revealed that insufficient power was generated on the downhill leg to return the train to the mine, according to numerous engineers who have not been authorised to speak to the media and have informed this story.

  The team developed two solutions to the problem, but they both had unacceptable implications for Fortescue's core business of shipping vast quantities of iron ore to Asia.

For fun, let's say you have a 20-ton container, and you raise it up the rail by 100m. The stored gravitational energy is about 5.5 kWh.

According to Wikipedia, Tesla Powerwall 3 has 13.5 kWh of capacity: more than two 20-ton containers raised to 100m, assuming perfect efficiency. It costs $7,300, small enough to put in your house, and also (more or less) safe enough to put in your house, unlike a 20-ton container on a rail barreling down a slope, which probably needs professional hard-hat maintenance crew.

So consider me skeptical.

Where does the energy to lift the cars up in the first place come from? Another set of “freely available gravity” cars down the street?

Track maintenance is the single biggest cost sector of rail lines in good condition even on very low slopes. The ballast needs regular inspection and replacing even without adverse weather or underbed problems, and the track can develop cross fall or slew and/or creep in transverse and longitudinal directions over time, sometimes taking subsurface layers with it. Every other kind of gravity storage makes more sense on paper than this, even the tower crane rearranging concrete cubes proposal.

The main problem with this and many such ideas is that G (the universal gravitational constant) is just so damn small. Making this system store any serious amount of energy in a small footprint is limited by the density of materials you can raise and the height available. In a large footprint you need a lot of infra (eg: loooong rail) which has maintenance cost, and still, very small per-liter and per-m^2 efficiency.

Well, it's better than the scheme for lifting and lowering cement blocks with a construction crane. And the scheme for digging deep holes in the ground and lowering big rock blocks into them. And the scheme for using electric locomotives and a heavy train in much the same way as this scheme.

The animation hand-waves important details. How are those blocks moved on the flat part of the track? There's no backup braking system. No guards around the chain. No chain lubrication system.

Performance should be roughly comparable to pumped storage with the same height difference, so why bother? Pumped storage doesn't use up much water; it's the same water going up and down, with some evaporation loss.

I think this could be superior to batteries and pumped hydro in enough situations to matter.

You can build these in many more places (closer to generation/load), the capex is significantly lower, and you can probably build it a hell of a lot faster than a reservoir. This solution is also more incremental than pumped hydro and the equipment will likely last significantly longer than a lithium ion chemistry battery farm.

The biggest bottleneck for getting a big project on the grid right now is interconnection. If you can avoid having to deploy new transmission lines to a new site you can often chop 5+ years off a project's time table.

In terms of peak power wouldn't flywheel farms make sense ? Heavy duty, durable bearings might be a problem.