Why energy storage sucks | Hacker News Reader

Why energy storage sucks

medium.com

91 pointsatilev64 comments

Open articleView on HN

Very limited view of energy storage tech frankly. Quick list of things I am aware of (some deployed and some in labs)

- Pumped hydro, by far the most widespread

- Flywheel storage, used at grid scales for frequency regulation

- Thermal storage used in conjunction with solar power using various materials - molten salt, HTF, hot dry rock

- Many different battery chemistries:

    - High temperature batteries (e.g. ZEBRA batteries specifically developed for cost effective grid scale storages https://en.wikipedia.org/wiki/Sodium%E2%80%93sulfur_battery)

    - Sodium-ion batteries

    - Metal hydride batteries

    - Redox flow batteries such as zinc-air systems

- Compressed air storage

- Phase change storage, including hybrid compressed/liquefied air storage

- Seasonal thermal storage https://en.wikipedia.org/wiki/Seasonal_thermal_energy_storag...

- Water electrolysis + hydrogen fuel cell cycle

Probably missed a bunch too. All to say there are a lot more solutions then lead-acids and lithium batteries- although in fairness those are the most accessible options in small scale systems

Some interesting products utilising PCMs in the domestic hot water markets... Replacing traditional hot water cylinders which suffer from very high heat loss even when insulated.

Had to look up PCM, which stands for phase-change materials.

https://en.wikipedia.org/wiki/Phase-change_material

Found a few vendors of various interesting products, here's just one:

http://www.pcmproducts.net

Found a few articles advocating water heaters using PCM. Sounds cool.

Aside: I've tried HotSnapZ brand hand warmers. They're reusable. Neat. I should have guessed this strategy could be scaled up for bigger applications.

http://hotsnapz.com/FAQ.html

The economics aren't great for replacing a hot water cylinder with something more efficient.

According to this [1] a typical hot water cylinder wastes less than 2000 kWh per year. That's a yearly expense of less than $400. This means a replacement system aiming to break even over five years must cost less than $2000.

Replacing a hot water cylinder with a bog standard new one costs at least $1000. Thus the PCM heater must be essentially as cheap as existing solutions, and so it faces the challenge of not being able to get enough initial volume to get the cost below break-even. Government subsidies are probably the only way to make it work.

[1] http://www.solarblogger.net/2012/11/heat-losses-from-hot-wat...

The title of the article is just a bit inaccurate. Perhaps 'Why domestic energy storage sucks' would be better. The full title is in fact 'Why energy storage sucks. (but there is hope)' but the last bit is left out above.

Pumped storage is great, but not in your house or for an off-grid hut in Africa. One interesting application of pumped storage is applying it to tidal power. You could build a tidal lagoon and not only get the super reliable free tidal energy but also use it as a store. Pump extra water out or in when renewable sources are producing. Then use the power when demand is high.

This is covered a bit in this excellent free book: http://WithoutHotAir.com

There is also Superconducting Magnetic Energy Storage(SMES), train energy storage, molecular spring energy storage.

Train energy storage is driving a trainload of concrete blocks uphill when power is cheap and doing regenerative braking on the way down[1]. It's just like pumped hydro, but without the need for water.

Molecular spring energy storage is essentially making a windup mechanism, except with advanced materials like carbon nanotubes. Unfortunately it is not very practical today.[2]

[0] https://en.wikipedia.org/wiki/Superconducting_magnetic_energ... [1] http://www.gizmag.com/ares-rail-energy-storage/28395/ [2] https://en.wikipedia.org/wiki/Carbon_nanotube_springs

Also, of note, Proton exchange membrane fuel cells can be used both for Hydrogen but also fuels liquid at room temperatures. For examples there are Direct-Alcohol Fuel Cells for Methanol and Ethanol.

There are broadly two types of energy storage. Grid scale and domestic. This article is about the latter as it focuses on certain batteries. It's actually pretty interesting but the scope is narrow.

Domestic is important as it can leapfrog other tech in places without a reliable grid. Much like mobile phones leapfrogged fixed lines in Africa. There are lots of options for the grid but only batteries really work at a small scale.

For more on this watch this recent video from Al Gore: https://www.ted.com/talks/al_gore_the_case_for_optimism_on_c...

At grid scale you can have exotic types of batteries, pumped storage and other things. For example, there are plans to double Cruachan: http://www.bbc.co.uk/news/business-35666993

Pumped hydro is much cheaper than batteries per KW (max storage rate) and KWh (storage capacity) is limited only by reservoir size. The roundtrip efficiency is about 80% and it's able to store for months at a time.

Heindl Energy in Germany is developing hydraulic storage, which can work even on flat ground (basically, raising a rock formation with pumped water). They've designed capacities up to 120GWh:

http://www.heindl-energy.com/hydraulic-rock-storage/idea-fun...

Of course. Despite the title the article is only about off-grid storage, basically batteries. Though even off-grid storage can be on-grid if it's sometimes connected, as some plans for the future of plug-in hybrids call foresee.

Sadly you cannot very well use a miniature hydroelectric dam for off-grid energy storage on-board an electric car or electronic mobile device.

Pumped hydro is interesting but the reservoir requirements are huuuge. This post makes a very convincing case that scaling it up is not that simple:

http://physics.ucsd.edu/do-the-math/2011/11/pump-up-the-stor...

There is some discussion about the German rock variation in the comments. The density doesn't seem to help THAT much compared to water only.

Wonder how this works in relation to the earthquakes we're getting here in Oklahoma due to injection wells.

Density isn't the only benefit of this system. At these scales, the water requirements alone are significant. This replaces many cubic meters of that water with rock.

The other big advantage is that you don't need a tall hill to put a reservoir on top of.

Sadly it seems the project has been at the "plan to get funding for a pilot" stage for at least 5 years now, and I still don't see how it'd seal.

"to get the amount of energy stored in a single AA battery, we would have to lift 100 kg (220 lb) 10 m (33 ft) to match it. To match the energy contained in a gallon of gasoline, we would have to lift 13 tons of water (3500 gallons) one kilometer high (3,280 feet). It is clear that the energy density of gravitational storage is severely disadvantaged."

http://physics.ucsd.edu/do-the-math/2011/11/pump-up-the-stor...

It's almost incredible that a AA battery can do that much work, but I checked the math:

  Alkaline AA Battery =~ 3 Wh =~ 10000 Nm
  Gravitational Potential Energy =~ kg * m * 10 N/kg
  100 kg * 10 m * 10 N/kg == 10000 Nm == AA Battery

It also puts into perspective how much solar energy is available. A single 200 W panel (~5 sq ft or ~.5 m^2) can charge hundreds of AA batteries per day, and thus would require lifting more than 10000 kg by 10 m to store a single day's output from just that one panel!

Surprised there's no mention of Vanadium Redox as an alternative that solves several of these problems even better than Lithium Ion.

They can be fully cycled over 100,000 times (compare to Li-ion's 400-1200, lead's 500-800) and are expected to last 20 to 30 years. They can stay in a single state for long periods of time without degradation and can be rapidly and efficiently charged or discharged. There's also no fire risk.

The biggest downside is size and weight. A refrigerator sized VRB in the attic of a house would work, but that is a dealbreaker for some/most people. These certainly won't go in cars, but as stationary appliances for those with the space, they would work extremely well for solar storage.

DOE factsheet: http://energy.gov/sites/prod/files/VRB.pdf

Wikipedia: https://en.wikipedia.org/wiki/Vanadium_redox_battery

Copper hexacyanoferrate batteries are supposed to last at least 40,000 times (compared to what the study on this I read awhile back quoted as Li-ion as 400), yet have the energy density of li-ion.

They just haven't produced a commercializable version of the technology yet.

> A refrigerator sized VRB in the attic of a house would work, but that is a dealbreaker for some/most people.

Even a refrigerator-sized VRB doesn't store that much energy, that's ~650 liters so about 40MJ, under 1.5L worth of fuel. And more importantly it weights half a tonne and (assuming it has the same shape as a standard fridge) has an area density of ~1.5t/m^2.

40MJ is a bit over 11kWh, which should be enough to run a typical house overnight. My house uses something like 500W average, not counting the car, so it could run almost a full day on this.

For a stationary application, half a ton doesn't seem like a big deal. Surely a normal house can hold up that much weight? That's equivalent to only six or seven people. If the weight is a problem, put it on the ground level or in the basement.

Personal VRBs make about as much sense as having your own gas turbine, but utility-scale ones for peak demand response make a lot of sense.

Speaking of having your own gas turbine, that was my existence proof a while back for how to handle events where lots of people want to charge there cars all at the same time: 30MW in a trailer http://www.pwps.com/gas-turbines/30-mw-mobile.html

Granted there are almost certainly better solutions, but anywhere we can deliver NG to, we can deliver stupendous amounts of electricity to, for a price.

Gas turbines have a loudness problem, but there are natural gas reciprocating engine co-generation plants available for your house:

http://world.honda.com/powerproducts-technology/cogeneration...

Better flow batteries under development: https://en.wikipedia.org/wiki/Flow_battery Some as good as li-ion, potentially great replacement for the cars

(someone actually built a car powered by flow batteries, take a look: http://www.gizmag.com/900-hp-supercar-flow-battery/31091/)

compare to Li-ion's 400-1200

This may be true for some kinds of li-ion batteries, but they can be engineered to last much longer than that. Tesla's daily cycle home battery is warrantied for 5,000 cycles, for example.

Assuming the battery doesn't die the day the warranty expires, this should give you at least 15 years of life. Not quite the 20 to 30 you're quoting, but current prices are much cheaper (and this seems likely to remain true, as vanadium in inherently expensive, and li-ion batteries have a head start in economies of scale). With massively reduced size and weight, shipping and installation should be much lower as well, and I suspect consumer uptake will likely be much higher for the foreseeable future.

I really want to read this article, but there are too many spelling and grammar mistakes or lack of punctuation. Please consider spellchecking it?

I absolutely agree. Already after the first sentence, I was questioning whether I was suffering from aphasia or having a stroke. It surprises me that someone would publish an article like that without having proofread it at all.

I didn't notice anything serious. Perhaps you should try to tone down your "superior grammar" dial?

Nah, he's right it's practically unreadable.

Unreadable? Sorry to tell you this, but your standards are unpractically high. Welcome to the real world, where not everything is how you want it :p

I'm sure that with such an attidtude, you'll be learning a lot from people with different backgrounds and cultures. /s

For future reference: unpractically is not a real word. Please consider using real words such as impractical.

Ah, my bad. Thanks for the heads up!

If your English skills are reasonable, it's fairly easy to parse the article and fill in the gaps yourself. Probably easier than a non-native writer wrestling with a spelling or grammar checker when they might not know if the suggestions presented are good or not.

I agree completely. It seems some HN readers think that if they have perfect English (they do not, sadly), everyone else should rise up to their unrealistically high standards. I guess us programmers are a bit more sheltered from the real world. Add to that the primarily US-based audience, and you have what is called a "hivemind".

I think it's more like Stockholm syndrome. Our compilers are berating us 100 times a day that we're making grammatical mistakes in the language we're using, and we begin to love them for it. It's no wonder we tend to demand a higher standard from the authors we read.

Spellchecking? What about innumeracy?

As you can see if you only drain your battery to 75% (meaning you only use 20% of your capacity) you get tremendously more cycle’s (more than double!) then say using 50%. So when systems are designed for energy storage you generally want your batteries to last a long time (since they’re expensive) so you want the maximum amount of cycles for them, which means that you should only discharge them 80%. As you can see this is getting quite frustrating, as now simply to power you’re off-grid system at night you need to purchase 80% more battery capacity than you actually need.

75 + 20 = 95. Okay, close enough to 100. But discharging to only 20% means you need to buy 500% the capacity, not 180%!

Yeah, I was hesitant to click on the article based on the title. Thanks for confirming my suspicions.

I did learn a few things from the article, though. The big one has to do with why the car batteries I use on a hydraulic and winch system have been failing fast. It has to do with the way I run them all the way down before charging them back up.

In my shed delivery business, I have a custom designed trailer with a hydraulic lift for the tilt bed and a winch for loading and unloading the buildings. They are 12 volt and I just run them with a car battery.

Plus, getting 2100 cycles at 75% vs. 1000 cycles at 50% is barely an improvement at all. With this particular example battery there is absolutely no reason not to discharge it completely. You get roughly the same number of watt hours before it dies, no matter what you do.

For medium.com would it make sense to include the username in the title, like this:

    Why energy storage sucks (medium.com/@atilev)

Energy storage not only consists of batteries… Pumped hydro, for example, works well and does not decays over time.

Pumped hydro is absolutely great, but it's an option for tremendous (grid-level) scales, and requires convenient locations (e.g. a hilltop you can "blow up" with a large stream nearby, or a standard dam with a lake or lake-able location close to its foot). It's way, way beyond the scales the article is talking about (note the picture from a health clinic in Rwanda).

How big of a tank on your roof do you need to replace a box of lead acid batteries?

Is your house 100 feet tall or taller, and I hope your whole house is made out of structural steel.

A 12V deep-discharge lead-acid battery is typically 120amp-hours. That's 1.4kWh. Suppose you only 50% discharge it for longevity, that's 700Wh. I'll assume your box of batteries has ten of them, so a usable capacity of 7kWh. 1kWh = 3.6MJ, so 7kWh = 25MJ

How high is your roof? Lets say 10 metres. Potential energy U = mgh. (m=mass in kg). h=10, g~=10, so U=100*m.

So, to equal those batteries, assuming 100% efficiency, the mass of water needs to be 250000 kg, or 250 tonnes. Probably going to need to reinforce the loft.

So 'roof' is clearly out of the question. Now let's look at basement. I install a 4x4x2 meter tank, then drill down 100 meters to create another such tank. 32 tonnes lifted 100 meters gives me 32MJ, perfect.

Now how expensive is it to build a water storage area underground? Is this a 50 thousand dollar project or a 50 million dollar project?

If the ground water is at the right depth, you would only need to drill a well. So this should be practical somewhere, at least.

Your 100 meter channel is going to get flooded.

That's your biggest objection to the idea? A 100 meter waterproof pipe is the easiest problem to solve here.

Plus you probably cannot beat the cost efficiency. I see challenges for mobile applications, though.

The comment I was writing turned into a blog post, so I'll just post the links. IMHO these are trending towards solving many of the problems faced by alternative energy generation and storage:

http://www.npr.org/sections/thetwo-way/2016/02/04/465568055/...

http://www.wsj.com/articles/energy-storage-startup-lightsail...

https://en.wikipedia.org/wiki/Nickel–iron_battery

https://en.wikipedia.org/wiki/Sodium–sulfur_battery

https://www.littleboxchallenge.com

https://www.littleboxchallenge.com/pdf/finalists/56568-Tech.... <- PDF

So what happened to the idea of using some modern version of flywheels to store power?

They were deployed in Formula 1: https://en.wikipedia.org/wiki/Kinetic_energy_recovery_system

I'm not aware of any building or utility-scale systems.

There are a few utility-scale flywheel systems deployed: http://beaconpower.com/stephentown-new-york/

Laboratories such as JET (fusion research) use them, because they cannot draw enough power from the grid: https://www.euro-fusion.org/fusion/jet-tech/jets-flywheels/

I've seen some in data-centers, mostly to filter out noise and provide continuation between power main failure and generator kick-in.

Fresno City has approved an Amber Kinetics project for flywheel-based large-scale solar storage: http://kvpr.org/post/fresno-project-store-solar-electricity-...

The company advertises 160kWh/TEU (as 40ft container units): http://amberkinetics.com/products/

Flywheels target a bit of a different problem than many other energy storage solutions. They are only able to store energy for very short periods. This is valuable to counter short term changes in the power generation, but it won't help you to store solar power for longer dark periods.

Wikipedia has some info about existing installations: https://en.wikipedia.org/wiki/Flywheel_storage_power_system tl;dr Only a few of them exist, they haven't been installed in large quantities yet.

> They are only able to store energy for very short periods.

Your own link quotes self-discharge rates of 5%/day (for high-vacuum wheels I assume), that's more than sufficient for daily solar storage.

I would consider that to be short-term storage. Long Dark periods could be weeks.

"Short term changes in power generation" are not day-scale, what you consider to be short-term storage has very little to do with the comment I replied to, or with the original article.

Even a long dark period would see very little discharge. The flywheel would take a trickle of power (enough that a human being could generate, if necessary) to maintain their energy.

https://www.kickstarter.com/projects/1340066560/velkess-ener... ... seemed to be going ahead until oil prices fell recently.

Market conditions aside, we remain absolutely confident in flexible flywheel technology and its ability to deliver safe, dependable, high performance energy storage at a small fraction of the cost of today's best storage solutions. We believe someday flexible flywheel technology can and will become a foundational technology in our clean energy future.

No mention of Ni-Fe batteries, they are significantly more robust than LA or Li and have 30+ year lives.

Great article from a practical standpoint of what is actually deployed and small scale.

Reply on news.ycombinator.com