Hacker News Reader

Comment by rsynnott | Hacker News Reader

I wouldn't bet on batteries, or at least conventional batteries, being the long term solution. There's interesting work being done with air pressure storage, flow batteries (the largest batteries currently under construction are mostly flow batteries in the 500MWh-1GWh range, though the technology is still in its infancy), and big scary vacuum flywheels.

All of these are far less developed than lithium ion cells, but have the potential to do a lot better on cost and longevity grounds.

View on HN

Planet Money covered this recently: https://www.npr.org/templates/transcript/transcript.php?stor...

Here's an entry on the storage facility: https://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Sta...

"The biggest battery in the world is a giant lake tucked away in the Appalachian Mountains on the border between Virginia and West Virginia"

One of the more promising things is actually water-based storage using back-to-back reservoirs. Water has a lot of mass per volume and is really low tech. And it's a proven strategy that has been operational for many decades.

Pumped storage is great where the geography makes sense, but there are a limited number of good sites which can be economically connected where needed to grids. Long-term, I'd have more hope for currently emergent tech (flow batteries etc).

it is true that pumped storage and great solar locations are often in different locations.

maybe there is a possibility for retrofitting some existing dams that are used primarily as resevoirs and not power plants, since pumped-storage are net-neutral on resevoir storage.

example: https://en.wikipedia.org/wiki/Parker_Dam

...but i am NOT an expert by any means, and it's probably way more complicated than that.

I think the easy way to do this is simply to build extra generation capability on existing hydro dams, then letting the dams fill while the sun shines and the wind blows and empty out faster when they don't

It also introduces interesting financial arrangements. e.g. rather than profiting on the electricity directly, they make more money by selling options to sell at a fixed price at peak times.

As a child (25-30 years ago) I visited an hydroelectric plant in Wales. It always impressed me that such a simple technique could work so effectively. They also had a pumping system for use whenever say the nuclear plant had excess production..

One weird thought I had the other day: high voltage DC transmission lines are getting more popular. I wonder if it would be economical for a high voltage DC line to form a loop around the entire planet, so that the load can follow the sun through a complete cycle.

This has terrifying geopolitical implications, so probably no-one will want to depend on it anytime soon.

It's no different than data links, which have mostly worked without political interference.

In general, your data does not have value to me to steal it. Not so with power.

Data has plenty of value, and is often copied without authorization, but putting that aside, nobody would be stealing anyone's power. It gets produced in one region, used locally, and excess sold to the next region, where the same thing repeats. You already paid for the power entering your region so you can't steal it and you'll sell it on if it makes sense to do so. This is already how gas pipelines work for example. It's shit with gas because the flow is one way and shutting off the gas is often done for political reasons but with solar you'd have your own production ramping up a few hours later if the neighboring region either shuts it off or uses a higher percentage of what they produced. Either way the metering happens at regional boundaries so there's no stealing involved - both sides of the meter need to be in agreement before any power flows.

> Either way the metering happens at regional boundaries so there's no stealing involved - both sides of the meter need to be in agreement before any power flows.

Play that out a bit. Suppose you have a 500MW or 1000MW transmission line. It normally has a 1% loss. Today, you notice that it has 1.25% loss. Maybe someone's stealing power. What do you do? Shut off the transmission line entirely?

At the demarcation points is where power is measured. If you have losses you can tell which side they are on, and then that side can decide whether they are worth investigating or not.

They really haven't.

If only a different species had reached our level of intelligence first... This would be a cool solution.

Loss is still an issue

If the loss is less than about 75% in the worst case (i.e. 100 watts of generated electricity supplying 25 watts to the consumer), the benefits of optimal panel location combined with the lower cost of solar makes it better than or equivalent to current baseload power supplies like nuclear.

In the US alone this would greatly aid a transition to wind and solar:

http://www.sciencemag.org/news/2016/01/better-power-lines-wo...

>In the US alone this would greatly aid a transition to wind and solar:

http://www.sciencemag.org/news/2016/01/better-power-lines-wo...

For all the talk about inventing new batteries, I wish improving the existing grid would be more prominent in the conversation. High-voltage longer-distance power markets are something that can be built right now.

Batteries are pretty convenient though. They don't have any moving parts and can supply electricity on demand.

Not all lithium batteries are created equal in terms of longevity. A traditional lithium manganese battery might have 500 cycles in it before hitting 80% capacity. A lithium titanate [1] battery might be closer to 5000 cycles. A battery could be designed for long term usage and thus lower the long term cost, if not the upfront cost.

[1] https://en.wikipedia.org/wiki/Lithium%E2%80%93titanate_batte...

Another benefit is that if you can reliably get off grid, it reduces infrastructure costs at the community level. Say, for rural properties.

I think the key is that we can start to develop infrastructure just fine on lithium cells, and it might even accelerate the development of more efficient alternatives develop because it proves the market to investors.

The market of grid-level energy storage is already proven to investors and anyone in the utilities industry.

See CAES storage (Texas) and Pumped Hydro energy storage (https://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Sta...)

Some sort of battery based tech that is fairly unconstrained by site requirements, as well as complex construction engineering requirements seems to have much more potential.

The Bath County Pumped Hydro plant can give 3GW of power with over 10-hours of storage. That's 30GW-hrs. How many lithium ion batteries do you need to compare against that? "Small" Pumped Hydro Plants are hundreds of MW-hrs, and large ones are dozens of GW-hrs.

CAES in Texas was constructed inside of abandoned mine-shafts. Line the mine-shafts with steel to contain compressed air, and bam, cheap energy storage.

------------

The USA has lots of lakes and mineshafts. And electrical storage in one area can serve multiple states. Indeed, the Bath County Pumped Hydro plant contributes to the electrical stability of 16 or so states.

A $5 billion project here and there will serve more people and in a wider area. If the USA can take advantage of natural resources, in a sustainable way (it doesn't seem like Bath County's Reservoir is damaging the environment), then it should happen. We know how to make hydro-plants that are safe for Salmon and other wildlife (we need to ensure that future Hydro plants need salmon ladders for example), but as long as environmental concerns are identified and addressed, leveraging our natural resources is the most obvious way forward.

I didn't intend to discount the projects you mentioned, there are lots of approaches being pursued. The more of them get a real crack at the market the better in my opinion.

I think what ARES is doing has the biggest potential: running heavy trains on a closed loop uphill to store energy and downhill to generate. They claim 80% charge/discharge efficiency which is quite good. The technology is very well understood and mature and the environmental effects are lower than building a reservoir.

Reply on news.ycombinator.com