Hacker News Reader

Comment by hpb42 | Hacker News Reader

I always wondered if flywheels[0] would be a good alternative to batteries to store energy. Flywheels can absorb peak current, help stabilize grid frequency[1] and can have high efficiency.

But I haven't seen much on the topic, and a lot in chemical batteries.

What are the prices and downsides of flywheels? Does anyone have an up date comparison of commercial solutions with TeslaWall and similar?

- [0] https://en.wikipedia.org/wiki/Flywheel_energy_storage - [1] https://en.wikipedia.org/wiki/Flywheel_storage_power_system#...

View on HN

For transparency, I'm not very well versed on this topic. That said, one concern I've always had with flywheel systems is that it ends up being a crazy heavy thing spinning at a decent rate. I don't know how valid it is but many years of being around maker communities and stuff has given me a respect for heavy things spinning at decent speeds.

They do seem like they would be a simple effective way to get some brownout or power flicker protection though!

A flywheel can provide frequency stabilisation, or it can provide energy storage, but it can't do both at the same time.

To provide frequency stabilisation is must be synchronised to the mains (eg run at a constant speed), while to provide storage it must be non-sychronised, eg slow down when loaded, and speed up when charging.

Of course you can use non-synchronous flywheels with synchronised inverters, but then you can do that with batteries, wind-generators, etc.

> A flywheel can provide frequency stabilisation, or it can provide energy storage, but it can't do both at the same time.

Ah, that's a good point to know. I thought flywheels could to both at the same time. Do they also help with keeping the load resistive? I guess their high inertial mass would add some inductance, which is how they stabilize the frequency?

Do you have some resources about this topic to share to a layman like me?

https://www.torus.co/torus-flywheel

Cool! Thanks for the link.

A 10 kWh flywheel that can charge/discharge at 20 kW (lasts for 20~30 min at this power) costs ~11500 USD + installation and control system.

A Tesla Powerwall has 13.5 kWh capacity, delivers power at ~6 kW and costs ~11500 USD.

That doesn't seem a large difference to me.

That looks very cool. Mechanical energy storage. Any idea what the return efficiency of such a device is? And guesses about its cost?

No idea. But interestingly enough, it looks like someone built a grid scale version in PA with funds from the 2008 Recovery Act. There's a final tech report here https://www.smartgrid.gov/project/hazle_spindle_20mws_flywhe... in the Related Docs

well, if nothing else, that thing looks futuristic as hell

would love to point out in my yard and say "oh, that? that's our mechanical energy storage device"

flywheels solve a narrow problem and are worse than batteries at delivering long term energy. The examples linked in other comments are around 10 kWh of energy. Most north american houses consume around 30-80 kWh per day. The real problem with flywheels is scale and longevity. How many flywheels can you name that have been in service for 20 years? Lithium batteries have won the short duration energy storage race and most other solutions are for edge cases.

Sounds like they are -- thanks for the links, that was an interesting review!

Reply on news.ycombinator.com