Tesla – Lithium-ion storage is ready to power the grid | Hacker News Reader

Tesla – Lithium-ion storage is ready to power the grid

bloomberg.com

172 pointsdjrogers51 comments

There's a missing parameter here: the number of recharge cycles a battery can have without losing large capacity , because over the long term - decades, like done in this business, that's what determines the cost per kwh charge stored and supplied, i.e. what end users pay for.

Without those numbers, it's not possibe to judge Tesla against the competition, for example Alevo, which claims more the 50,000 charge cycles, vs ,i believe, 5000 charge cycles for Tesla.

For Alevo, the final cost per store/supplied should be ~3cents/kwh, over 20 earss lifespan, which should make grid storage very economical.

$0.03/kWh delivered is an order of magnitude less than current lithium ion systems which ranges $0.285/kWh - $0.581/kWh delivered to the grid after storage. [1].

Since the data going into Lazard's models are probably at least more than a year old, I wouldn't be surprised if its currently $0.15-$0.30/kWh to store and then deliver a kWh of electricity.

The largest price difference in Southern California Edison's time-of-use tiers is from $0.11/kWh to $0.47/kWh, so I expect that the systems that Southern California Edison is currently installing are fairly cost efficient for them.

If Alevo can actually deliver at $0.03/kWh at scale (and personally I'm skeptical of something that sounds so good), then that's game over for all fossil fuels and nuclear on the electrical grid.

[1] https://www.lazard.com/media/438042/lazard-levelized-cost-of...

I also wonder how it compares in cost to pumped storage hydroelectric.

You can use excess power to pump water uphill and then release it to generate hydro power when needed. Requires the right geography but it can essentially store huge amounts of power and it's apparently pretty cost effective.

California already has at least a couple that I can see:

- Helms Pumped Storage Plant[1] - 1.2GW capacity

- Castaic Power Plant[2] - 1.5GW capacity

That's like 60+ tesla plants each.

[1]https://en.wikipedia.org/wiki/Helms_Pumped_Storage_Plant

[2]https://en.wikipedia.org/wiki/Castaic_Power_Plant

Pumped storage hydro is great, but it is limited to places that have the right terrain.

I wonder what the cost would be of pumped storage in just like a massive water tank on a hill, or a big manually dug out lake, rather than in somewhere that could actually be a natural lake.

Taum Sauk, an artificial pumped storage facility which famously failed, was rebuilt to 3600 MW·h for $490 million. So that was like $136 per kwh capacity, though I think they didn't have to do much for the lower reservoir (which was built in the 60s). Seems to be a lot better than batteries, though still requires appropriate terrain.

https://en.wikipedia.org/wiki/Taum_Sauk_Hydroelectric_Power_...

Sure much better, this California push is a subsidy for battery storage. Not a economical choice.

Pumped storage has even the advantage of that it can be implemented during construction of hydroelectric power plants, you just need the interconnected grid, but you will need the grid either way with renewables.

In other words: you don't need to build a reservoir. You just need to pump back of the exit of a already existing hydroelectric plant and build the power line to connect both plants.

This is flat out wrong, the cost per MWh for stored electricity compares favorably to the natural gas peaker plants that it's replacing, at least when they're built on a usual schedule.

I would also like to see evidence for th claim about reusing existing hydroelectric plants, because that does not seem to be consistent with anything I've seen before. It would be very welcome news to me!

Every hydroelectric is a specific project and the waterfall size and volume usually determines the type of turbine (pelton, francis and kaplan are the most popular from high to low) .. the denomination of pumped storage systems is applied when a reversible turbine is used, so no standard hydroelectric with a reservoir will be called pumped system, those are called impoundment. If you abuse nomenclature could be a open-loop pumped system.

Discussion with example here: http://e360.yale.edu/features/for_storing_electricity_utilit...

Economic comments here: http://large.stanford.edu/courses/2014/ph240/galvan-lopez2/

To complete here are two examples in US of existing hydroelectric that had pumped systems added in latter date (decades after operation start) becoming the so-called open-loop system I referred:

[1]Richard B. Russell: https://en.wikipedia.org/wiki/Richard_B._Russell_Dam

[2]Grand Coulee: https://en.wikipedia.org/wiki/Grand_Coulee_Dam

So before saying something is "flat out wrong" be sure you are not ignorant on the subject.

Some of the reservoirs of the California Water Project are used for pumped storage. Water is pumped uphill at night, and some of it is let back down in the afternoon. This would never pay for the reservoir and plant, but it covers the power bill.

Early compiled reports by Tesla users suggest that there is far more cycles on those cells than many had previously anticipated. https://electrek.co/2016/11/01/tesla-battery-degradation/

If you limit the charge of the lithium battery to something like 4.0v versus 4.2v (70-75% capacity) you can double or triple the amount the useful cycles.

All of this is nice because electric motors are essentially infinitely rebuild-able provided they are not burned out or abused.

There is absolutely no question that electric motors are superior to gasoline motors in every way. More torque, simpler, less maintenance. The problem has always been the batteries.

Provided the batteries can be replaced without destroying a car and there isn't some sort of DRM protecting the car from third party batteries there isn't any reason why you would want to buy a new car except to waste money.

There seems to be some conflicting information in the article, I don't understand how these quotes line up:

"...prices for lithium-ion batteries have fallen fast — by almost half just since 2014." "But for the most part, according to a BNEF analysis, the costs of new projects would need to drop by half in order to be profitable on a wider scale in California, and that’s not likely to happen for another decade."

The prices for the batteries dropped by half in 2 years, but it will take a decade for them to drop by the same amount again? What am I missing here?

Also the chart seems to indicate the cost was $273/kWh to build a plant in 2016, but then they say "While Tesla declined to provide its pricing data, the similarly sized Altagas project was expected to cost at least $40 million, or $500 per kilowatt hour." I guess the costs unrelated to the battery cells themselves added another ~$230/kWh in costs?

It says the cost of the project needs to drop by half not the cost of the batteries. It could be that the cost of batteries is only a small amount of the overall cost. That is what you see with solar, most of the cost comes from installation.

the cost of the project needs to drop by half not the cost of the batteries

Manufacture larger self-contained units the size of a standard shipping container?

that would be awesome, but would be one heavy shipping container..

Well, given we ship around semi-trucks full (ok not completely full due to the exact weight you describe) of lead-acid car batteries for recycling, it wouldn't be unthinkable. Li-on, I believe, is roughly 200% lighter for the same battery.

IE: http://www.powertechsystems.eu/home/tech-corner/lithium-ion-...

Not sure how that carries over to larger batteries as I've only really dealt with them for ATV type use.

Is 200% lighter half as heavy, or is it buoyant? Cause 100% lighter is weightless. So 200% would be a negative weight. I am just confused.

You are correct, the link to powertechsystems.eu has those erroneous percentage calculations.

The back of my envelope says that a shipping container full of Li-on batteries would weight 192.5 tons at 2.5 kg/L compared to the maximum standard load of 30.5 tons. So it would have to be mostly empty. I don't think that makes it infeasible, though.

40kg for lead acid equals 13.6kg for lithium so that's 66% lighter.

Make it a 10ft shipping container. (1/4th size) That should be good for a full standard load of around 30 tons, with some empty space for equipment access. Maybe redox flow batteries will win for the grid scale use case, because it could become feasible to process the solution in-situ and save a bunch of shipping weight.

If it's just 50% away from profitability with the ridiculously cheap cost of electricity in California and the US in general, it will do just fine for vast swaths of the developed world.

Not to mention that the edge of profitability probably isn't so much determined by the consumer price of electricity, but the sources that feed a countries grid. Lots of renewables with huge output swings (Germany) versus full nuclear (France).

A big factor in profitability is normal swing in grid price which is going to tend to be pretty pretty big in California with it's high baseline energy prices and extensive use of renewables which come on and off line according to their own schedules.

Statistics aside, this is wonderful. It makes a teenager like me optimistic about our energy future. Love the fast roll out too. In 50 years we will be thanking Elon Musk for saving all our asses.

Can anyone explain the logic behind having lots of small batteries, rather than the power company buying a Big Battery and offering this as a service?

Two reasons occur to me:

1. Situations off the grid or with unreliable power (eg blackouts) 2. Transmission costs are high, so it's better to have the storage close to the source.

Even in the second case, it seems like the power company could distribute these. I wonder if they are just more regulated/slow moving/immune to marketing.

I am not any kind of battery engineer, but...

My understanding is a primary limitation for these batteries is cooling. Cooling ability is relative to the surface area of the battery. One giant battery has much lower surface area than a bunch of small ones.

Also having a large number of batteries lets you put different cells on different recharge cycles to maximize battery life. You can use a small number of worn out cells for brief minute-to-minute fluctuations while you keep fresher cells on longer cycle that optimizes lifetime. With one big battery you're just stuck with whatever wear characteristics the usage curve gives you.

2 is the main reason. Transmission is the biggest loss so by having the storage close to the consumer you need less storage for a given end user consumption.

Also availability of lithium cells isn't that great at the moment. It'll improve as factories come online.

Can we not use an inverter, convert to AC, step up and use the existing infrastructure for transmission.

Besides, I guess there should be an added advantage to monitoring and using (charge/discharge) the batteries better than letting the user do the maintenance like preventing discharging to zero, temperature etc.

I'm curious - if you take a normal Li-ion battery out of a phone (for example) and whack it with a hammer it'll put on a pretty decent and dangerous show. Can the same thing happen to a big battery?

I'd imagine they wouldn't be as 'flimsy' as the ones in a phone, and weight isn't really an issue so they could happily have some kind of extra insulation/shielding.

Why yes, there's engineering that goes into stopping a fire from spreading. Here's some info about a fire safety test done to a Tesla Powerpack:

https://electrek.co/2016/12/19/tesla-fire-powerpack-test-saf...

Nice, thanks for the link, I was sort of hoping it would explode violently but probably good that it didn't.

Yes, but fortunately it's a bit easier to deal with than natural gas.

I think (could be wrong) that these 'big batteries' are just banks of small batteries- a bit bigger than AA cells - the same that run cordless power tools and (older) laptops (and Tesla cars)

You mean the 18650 cells.

I'm not sure if Tesla still uses them (they certainly did in the beginning), or whether they've switched to some proprietary format by now.

EDIT: They still use the 18650, but plan to switch to a bigger format for the Model 3:

http://fortune.com/2016/07/27/tesla-bigger-battery-gigafacto...

Relevant: http://www.bbc.com/news/technology-38637357

This is slightly OT since it deals with home batteries, not gigawatts, but here is a simulation of the Tesla Powerwall including putting power back on the grid: https://forio.com/app/powersim/powerwall-simulation/

It seems like once they get their costs nailed down leasing them, and basically taking the first X% profit, giving 'green' thinking people new tech to show off, leverageable during power events at a profit if done in concert with energy companies.

for about $57,000 American you can get 370kwh centrifugal energy storage from Beacon Power excluding inverter, with a design life of 100 years bearing service every 20 years. that's about $154/kwh and 100 year life.

"battery plants take up a much smaller footprint than gas-powered plants"

Im pretty confused by this. The energy density of LPG is 10x that of lithium-ion, so why isn't the footprint smaller?

just guessing, maybe they mean the whole facility.

Kinda like nuclear core vs nuclear plant ?

Recent discussion: https://news.ycombinator.com/item?id=13467572

Says Tesla who bet the farm on lithium-based batteries.

There are better processes and chemistries hitting the market in the next year - they just need to reach volume.

Curious, what kind of stuff is coming out that's better than Tesla?

I'd imagine there's all types of companies trying to solve this in different ways.

High energy density matters a lot for electric vehicles but very little for stationary grid-tied storage, so many more chemistries are potentially viable for grid tied storage. The really interesting question is whether one or more of them will reach sufficient volume to overcome the head start of lithium ion.

A lot of thin film solar PV companies were quite sure that their costs would be lower than the incumbent, crystalline silicon, once the manufacturing volume got high enough. Most went bankrupt because their manufacturing volume never grew fast enough for their designs' advantages to overcome the handicap of smaller scale production. Tesla's chemistry uses cobalt, which is expensive, but I don't know if low raw material costs alone are enough for (e.g.) sodium ion batteries to win over something like lithium iron phosphate batteries.

It will be interesting to see the old v new businesses powerplay with technologies like these coming up so fast.

Is there such a thing as "conflict free lithium"?

Conflict free lithium is just called lithium. None of the top producers have active conflicts.

https://en.wikipedia.org/wiki/List_of_countries_by_lithium_p...

"But for the most part, according to a BNEF analysis, the costs of new projects would need to drop by half in order to be profitable on a wider scale in California, and that’s not likely to happen for another decade"

The plunder continues...

Reply on news.ycombinator.com