Falling lithium-ion battery prices to drive rapid storage uptake

https://pv-magazine-usa.com/2017/08/03/falling-lithium-ion-battery-prices-to-drive-rapid-storage-uptake/

by okket • 9 years ago

138 177 9 years ago

nabla92• 9 years ago

$200/kWh by 2019.

Assuming that households consume between 30 kWh and 15 kWh electricity daily on average and you want to have storage capacity for 1/3 of daily electricity consumption to get wind or solar energy work, you need 10-5 kWh per household. That's $2000 - $1000 per household. If that battery lasts 10 years, its $200 - $100 per year.

It's workable and scales if we are able to reuse lithium. When does that happen? (recycling is not reuse)

merpnderp• 9 years ago

That wouldn't work for much of the South where summer usage for a house can easily hit 100 kWh/day for air conditioning. The hottest part of the day usually starts at about the end of the prime solar hours. Plus you'll want massive overcapacity of your batteries as deep discharges drastically reduce their useful lifespan.

zAy0LfpBZLC8mAC2• 9 years ago

I would think that for that use case, using thermal storage should be more efficient? Don't store electricity to run AC later, but run AC to cool down water or the ground beneath your house or something, and then cool your house with that later.

EngineerBetter• 9 years ago

Wow - that's a crazy amount of energy. As a Brit who's never been the south of the US, is that figure really true? 100kWh is about 300 miles in a Tesla, or boiling a kettle constantly for about 30 hours.

If that's really accurate, then I'm staggered at how much energy folks must be using to keep cool, and makes my efforts at saving energy seem paltry in comparison.

sp3322• 9 years ago

It's only prime solar hours if you point your solar panels due south. That gets you a higher figure on paper, but if you need the power later in the day you could point them a bit more to the west and get a lot more usable power when you need it. With solar panel prices falling, it's a race to see if it's cheaper to over-provision solar panels or batteries.

skummetmaelk3• 9 years ago

100 kWh/day is just incredible. The average consumption per person in the nordics is about 1500 kWh per YEAR.

beefield• 9 years ago

deep discharges drastically reduce their useful lifespan.

That applies to lead acid batteries. Lithium ones are much more forgiving on that (as long as you stay within design limits.)

kilroy123• 9 years ago

There are far cheaper ways to keep your house cool, though they require an upfront investment.

http://energyblog.nationalgeographic.com/2013/09/17/10-myths...

baybal26• 9 years ago

This digit is unbelievable. A good aircon can pump 4 gigacalories per hour using 1kw/h while running at full power with +40C and 40% humidity outside.

5gkal per hour is more than enough to keep a 50m2 room with huge windows at 21-22C in scorching heat.

If your house is huge, like 200m2 McMansion, and your ventilation does no recuperation whatsoever, and you have near nil heat insulation (like most of housing in US,) you will barely use 50kw/h per day even if you live in the middle of Sahara

zip12343• 9 years ago

Even if the hottest part of the day is past prime solar hours, with a solar system you won't be discharging the battery as fast since the system will be powering the AC for most the day. Also, with that much usage, it might be time to investigate insulation.

Florin_Andrei8• 9 years ago

That wouldn't work for much of the South where summer usage for a house can easily hit 100 kWh/day for air conditioning.

It should be noted that peak energy expenditure for air conditioning happens more or less at exactly the moment when the influx of energy from the Sun is also at its peak. That's, like, a gigantic hint that mother nature is dropping on you, hoping you'll notice it. The disease, and its cure, are gushing forth from the same place at the same time.

Probably most of the energy for air conditioning in the future will not come out of storage, but it would be produced at the moment when it's consumed, by solar panels or some other solar technology.

usrusr• 9 years ago

A well insulated water tank that is electrically chilled during peak solar and used to cool down air later in the day/night could solve that storage problem without any cell degradation. Theoretically, with perfect insulation, this could even reduce the total electricity demand, when the conversion from electricity to a temperature gradient happens during a time when the base temperature is low.

toomuchtodo4• 9 years ago

When does that happen?

Soon.

https://electrek.co/2015/08/16/the-power-of-controllable-cha...

https://i.imgur.com/oLBNeW4.png

barrkel• 9 years ago

I couldn't find the information about reusing lithium on those links?

Someone• 9 years ago

Why wouldn't recycling be an option? Chances are that, if large batteries become commonplace, it will be cheaper to mine lithium from used batteries than to do so from the earth because they will have very high lithium content.

Also, covering 1/3 of daily use seems low to me. It wouldn't get one through a few windless winter days with dense cloud cover, for example.

dougmwne• 9 years ago

Is it possible that perfect is the enemy of good here? You could choose a lower ratio of battery storage to average use and just use non-carbon neutral peaker plants for the few days in winter where there's a shortfall of renewable energy. Still a huge boost to renewable energy, still a huge reduction in carbon emissions.

nabla92• 9 years ago

Extraction of lithium from old batteries is five times more expensive as mining lithium. At some point it may become profitable.

On the other hand, there is lots of lithium. If the environmental impact of lithium mining going off the roof is not a problem, there is no reason to reuse.

Beltiras• 9 years ago

There is another way of thinking about the economics of this. IIRC the typical price is 12 cents per kwhr. You would need to utilize the battery in sold energy no less than 2000 cycles just to break even. About 5 1/2 years of constant monetized operation (which is a ludicrous proposal). And this assumes no cost on the energy production side.

Zekio4• 9 years ago

12 cents is cheap! in denmark the average price is 37 cents per kW/h

markvdb• 9 years ago

Electricity price is more like .26€ here(~.31$), and this is certainly not the most expensive place in the EU. Mostly network charges and taxes.

Going off the electric grid should become fairly easy. Cheap solar and storage should make this feasible for an eco-conscious household...

marcosdumay• 9 years ago

recycling is not reuse

Reusing the lithium from the batteries means recycling the batteries.

hwillis10• 9 years ago

It's workable and scales if we are able to reuse lithium. When does that happen? (recycling is not reuse)

I assume you mean reuse is not recycling ie using old batteries as lower-capacity storage is not workable indefinitely.

First off the most pressing issue is that there is no practical replacement for cobalt in many li-ion applications. Chemistries without cobalt, like lithium titanate or iron phosphate, have a price and weight premium and reduced capacity. Cobalt is rarer than lithium, makes up more of the battery, and is harder to mine. The price might go up very quickly past a certain point, if supply from the DRC becomes tighter.

Lithium is very widespread and relatively abundant but most importantly its final impact on the price of a battery is only a few percent. If it becomes more expensive to reclaim or mine, it will still have a pretty small impact on the cost of batteries.

scythe• 9 years ago

Sulfur beats cobalt on energy density by a significant margin; its theoretical specific energy is 5x as large, although the chemistry is much more complex: https://www.researchgate.net/profile/Hong-Jie_Peng/publicati...

There is one company in Colorado making Li-S batteries at small scale: http://www.solidpowerbattery.com/

fenwick672• 9 years ago

By 2025, the world’s base of cumulative installed storage capacity will reach 52 GW, IHS Markit says, up from around 4 GW today. Last year, 1.3 GW of grid-connected storage was deployed globally, and this rate is poised to accelerate to 4.7 GW a year by 2020, and 8.8 GW annually by 2025.

GW is not a unity of energy storage capacity. It's a unit of power.

epistasis6• 9 years ago

GW is, however, a characteristic of an energy storage system, and just a valid and important measurement as the amount of energy it stores.

Gracana• 9 years ago

The maximum output is interesting to know, but if you asked how much my car's fuel tank held and I told you "100mph," you'd be understandably confused.

ptero2• 9 years ago

Maximum discharge speed (which is one of the very few storage characteristics I can think of to be measured in GW) is pretty much irrelevant for storage systems. Add a few quick discharge capacitors that cost almost nothing and boom -- your max discharge speed shoots through the roof (but speed goes down for a very short time).

Maybe sustained charge speed is a (little bit) more meaningful thing you can express in GB.

I am with the parent -- this is either a typo or a journalist quoting something he knows nothing about (what's that "h" in GWh? Strange capitalization; probably a typo; let me strike that).

Florin_Andrei8• 9 years ago

GW is, however, a characteristic of an energy storage system, and just a valid and important measurement as the amount of energy it stores.

Wrong.

The watt is a measure of power, not of stored energy. These are different physical notions. So, yes, you could say that some storage facility can output up to so many GW, but that's actually a measure of how much energy per second can exit the system when it's being used.

The proper unit for measuring an amount of energy that's stored in the system is the Wh (or its multiple the GWh), which has the same dimension as the joule, which is the standard unit of energy.

GW is like how many liters of water per second can exit the storage tank when it's being emptied. GWh is like how many liters of water total are stored in the full tank. And yes, both are important when designing a battery system. But saying "GW is [...] a valid and important measurement as the amount of energy it stores" would fail you high school physics.

GWh is static energy (or simply - energy). GW is energy moving from point A to point B, which is power.

Here's a brief introduction to the basic meaning of these terms:

https://cleantechnica.com/2015/02/02/power-vs-energy-explana...

hwillis10• 9 years ago

It's weird and not usually reported like that. Possibly it is a typo, but either way it's relatively closely coupled to storage capacity: li-ion is typically used around 1C, so ~1.3 GWh of capacity was deployed, within a factor of two or so.

mavhc2• 9 years ago

I paid $354/kWh last month for 24kWh of batteries (inc 20% tax), and I got a free car that goes around it

dangoor• 9 years ago

What kind of electric car costs $8500? Used Leaf?

wlesieutre• 9 years ago

I heard from a friend that the Leaf's battery pack had a major redesign of the cooling system in recent models, that might have driven down the value of used older versions with worse anticipated battery wear.

aphextron5• 9 years ago

I just bought a 2011 Leaf with 40k miles in mint condition for $5,000. They are insanely cheap right now, and amazing cars if the range works for you (I get about 60-70 miles on mine). Also with the free quick chargers everywhere in the Bay Area I literally drive for the cost of car insurance now. I think it's just a combination of fear from the consumers and lack of knowledge on used dealers part combined that has left these things dirt cheap.

ChuckMcM4• 9 years ago

I wonder how much of this is simply the Gigafactory coming online at full production + others trying to keep their factories running.

If you look at Tesla's plan to be making 10,000 model 3's a week, with a nominal 90 kWh battery in each that is 900 MWh a week of capacity or 46 GWh capacity a year. But it is all tied up in cars of course[1].

The power companies have been screwing around with rate plans to keep grid-tied solar from being as economic as it could be and so I do see a lot of people such as myself moving to a whole house storage system at some point.

[1] I fully expect a TV show to have a scene where they need more power and the 'genius' hero looks out the window and sees a row of electric cars charging and "hacks the system to feed the car's energy into the power grid for the extra energy needed to save the day" along with the gratuitous cars exploding as they are being drained too quickly but reaching the minimum level of power at just the right time to save the world.

baybal26• 9 years ago

I wonder how much of this is simply the Gigafactory coming online at full production + others trying to keep their factories running.

The "Gigafactory," is a drop in the ocean if you count noname Chinese factories, but it will be one of the biggest market players outside of China

ChuckMcM4• 9 years ago

We have different ideas about 'drops in the ocean' :-) See this article (http://fortune.com/2017/06/28/elon-musk-china-battery/) for some numbers. Projected Gigafactory production (as more come on line) could continue to be 25 to 35% of the global production of batteries.

My father-in-law, who is a big booster of South America, feels that if Argentina, Brazil, and Chile got together they could move up the 'value chain' and produce batteries for export. I do know that much of the raw lithium comes from Argentina and Chile so I expect there is some truth to that idea.

Florin_Andrei8• 9 years ago

And by "drop" you mean a blob of water the size of a small moon.

agumonkey3• 9 years ago

Will the GigaFactory operate on weekends ? 1400~ car per day .. I'd love to see that on video.

toomuchtodo4• 9 years ago

The Gigafactory operates seven days a week with two shifts.

k__• 9 years ago

What is this? 1 car per minute?

aphextron5• 9 years ago

with a nominal 90 kWh battery

Model 3 has a base 60kWh battery, with upgrade available to 75.

ChuckMcM4• 9 years ago

Ah thanks, for some reason I read the 60 as 90.

ClassyJacket• 9 years ago

I don't believe that's correct. The Model 3 has both a smaller range and lighter weight than a Model S did at 60kWh, indicating a smaller battery.

Tesla hasn't confirmed how large the pack is at all, but they have specifically confirmed it's less than 60.

I think it's either 40 or 50.

https://electrek.co/2016/04/26/tesla-model-3-battery-pack-co...

briffle• 9 years ago

Pretty sure they will also use the batteries in their powerwall's as well (I seem to remember that the grid storage in CA was produced with batteries from the gigafactory)

I would imagine they would quickly switch the powerwall to use the new battery size that the gigafactory is producing..

hwillis10• 9 years ago

The powerwalls and powerpacks use a different chemistry from the cars, and are still using the 18650 size going forward.

It's trivially easy to change the size of cylindrical batteries. Li-ion batteries are essentially a tape, and the second to last step of production is to roll the tape up and insert it into the battery. Different cells just require different cans and widths of tape.

the84722• 9 years ago

The TV plot sounds like an updated version of Operation Yashima.

HillaryBriss2• 9 years ago

Germany has plans and investment commitments for a 34 gigawatt-hour plant that will add even more supply to the global battery market.

https://www.bloomberg.com/news/articles/2017-08-03/germany-g...

thescriptkiddie• 9 years ago

Utilities are actually planning on using parked electric cars for storage.

https://en.wikipedia.org/wiki/Vehicle-to-grid

michaeloblak• 9 years ago

I don't know why, but this somehow reminds me an excellent paper[1] by lcamtuf about parasitic data storage (keeping data "on the cable").

[1] http://lcamtuf.coredump.cx/juggling_with_packets.txt

gwbas1c• 9 years ago

Hell no!

As soon as someone departs for a long trip and the battery isn't charged, there will be a huge class action lawsuit.

The lemon law attorneys will also have a field day!

This is a really dumb idea.

vanattab• 9 years ago

Wouldn't this wear out the battery life?

anovikov3• 9 years ago

Of course they aren't 90kWh. more like 55 and 75 in 2 options. Still that is 32GWh per year in battery production on a single factory alone. Shouldn't be hard to make it to 1TWh a year worldwide if electric vehicles to ever have anywhere like mentionable penetration.

In a way, electric car batteries have demonstrated to outlive their cars by 2x-3x (there are already several Teslas who were driven beyond median lifetime of a car and they lost just a few percent of capacity). So when i car is trashed, a battery is essentially still new. They can be bought out for say half the price and used for electricity storage. If there is a billion electric cars in the world instead of the current billion gas powered ones, and they are on average 55kWh of which 10% is lost while they are in a car, and another 20% is lost while they are used for grid storage, on average they will have 45kWh of capacity. Grid storage takes 6 hour capacity (pumped hydro plants usually have 6 hour storage, enough to compensate any daily cycles), so 7.5kW power per battery or 7.5TW overall... the world has 6.3TW of installed generation capacity.

Probably grid storage will be a side business of car makers because all storage plant will need to use same type of battery. Maybe they'll even offer like 10% discount on a new car on a condition that the battery is theirs and to be returned when a car is scrapped.

tjoff3• 9 years ago

Any (non-tesla?) source? Model S has only existed for nine years.

Also, the life-time of a car that needs a $20.000+ battery replacement will obviously prematurely make the car utterly worthless.

Usage and weather most certainly have a huge impact as well.

Don't get me wrong, I'm interested in a used model 3 when that time comes. But that hinges on that you can replace the battery for a reasonable cost, reasonable even for an older car.

comicjk• 9 years ago

Using lithium-ion batteries for stationary storage is interesting because their original advantage was all about light weight (lithium is the lightest metal element).

Now, it seems the development effort devoted to lithium-ion has paid off so well that they are the best OVERALL battery solution. The single-minded effort we put into cell phones and cars is so strong that it now reaches all the way to homes and the grid.

To me, this suggests that all sorts of cool technologies might be possible if we as a society could concentrate on making them work.

wongarsu3• 9 years ago

In the last three decades battery technology was mostly about powering small, lightweight gadgets. Now that there's massive amounts of money to be made with car batteries and grid-level storage, I certainly expect some great battery technologies that focus on that in the next decade or two.

That's also why I don't care about all these fears about us not having enough lithium: nobody is going to power the world on lithium, it's just the cheapest method right now because of the massive amounts of research money already dumped into it, and great economies of scale. Other technologies will catch up as their use-cases become more valuable.

StillBored2• 9 years ago

Yah, lithium's power/weight advantages had driven so much investment that its now has awesome power/volume advantages too. The result is that even things which don't care about the weight (http://www.storagereview.com/apc_backups_pro_500_lithium_ion...) are starting to switch from lead acid because its no longer the super cheap solution.

That is something I've found really curious because if you tear into just about any lead acid battery it still looks like technology from the 1950's. Just about everything else has shrunk, become lighter, used less material, etc. But lead acid's last major "breakthrough" was in the 1970's with gel cells. I would have expected lead acid to look more like a capacitor at this point (thin layers of material rolled/folded up) but its still these fairly thick plates sitting in electrolyte with non-existent charge control. Surely some creative person sitting in a lab could come up with way to minimize sulfation or to decompose the lead sulfate within the battery without basically melting them down and building new ones.

keenerd• 9 years ago

That's also why I don't care about all these fears about us not having enough lithium

Who is afraid of that? Take a look at this: https://en.wikipedia.org/wiki/Abundance_of_elements_in_Earth...

Lithium is #33, estimated at 20 ppm of the crust. We presently mine 64000 tons per year. It is more common than lead, which is #37 and only 14 ppm. Yet we mine 4.8 million tons of lead per year. Now examine boron. #41, 10 ppm and 9.4 million tons per year. Are these unsustainable? Are we on the verge of peak boron and peak lead? Probably not. We have twice as much lithium so we probably aren't going to run out of lithium even if demand increases by a hundred-fold.

I am not a geologist, so maybe much of this lithium is more inaccessible than I'm assuming. But everyone who says we are going to run out of lithium have been only looking at the known deposits and are ignoring the possibility that more mines will open to meet demand.

aphextron5• 9 years ago

There is no possibility of ever running out of lithium for battery production. It is an extremely common element, and the actual lithium makes up about 3% of a LiPo cell by mass.

leggomylibro2• 9 years ago

Don't lithium-ion batteries degrade extremely quickly, though? Won't these all be down to half capacity or so within a decade?

Is it really worth investing in gargantuan quantities of energy storage if they need to be replaced every few years? I'd be eyeing EDLC technologies or something like molten salts over giant stationary phone/laptop/car/etc batteries.

toomuchtodo4• 9 years ago

No, they don't degrade that fast. They'll be good for at least 15 years based on charge/discharge cycle data Tesla has collected from vehicles.

wongarsu3• 9 years ago

They degrade extremely quickly in the usage pattern of a typical consumer device (alternating between full charge and full discharge). They last fairly long if you keep them between 30% and 80% charge.

It's not the most long-lasting technology, but we replace most infrastructure every 20-years already. A 10 year lifetime wouldn't be that bad.

hwillis10• 9 years ago

Li-ion batteries have the second slowest self-discharge of any battery and the longest cycle life/easiest upkeep.

An li-ion battery will do about 500 discharge cycles from 0-100. They'll do several thousand from 20-80, and tens of thousands cycles from 30-70%. If you have a few days of storage, the battery as a whole will last an incredibly long time even when cycled every day.

As for why: When the battery is kept at a high or low voltage, it puts an overpotential on one side or the other of the battery. There's more or less lithium on one side of the battery and it tries to diffuse across the battery very slowly. It never makes it to the other side and reacts with the electrolyte and is lost, reducing capacity.

The closer it is to 50%, the less lithium is pulled/pushed out into the electrolyte.

aphextron5• 9 years ago

Using lithium-ion batteries for stationary storage is interesting because their original advantage was all about light weight (lithium is the lightest metal element).

Not entirely. The real killer feature of LiPo which has allowed things like Roadster and Model S to exist is their high discharge, or "C" rating. The power density of LiPo is far beyond any previous chemistry available for batteries at scale. Energy density and power density are two completely different things [0].

https://www.quora.com/What-is-the-difference-between-power-d... [0]

Florin_Andrei8• 9 years ago

Li-ion batteries have a high energy density in terms of both weight and volume, when compared to other types of batteries. And yes, it's all relative to which tech you're comparing them to, but it's worth remembering that they are both lightweight and physically small, for the amount of energy they can store.

the84722• 9 years ago

In principle redox flow batteries should offer better scaling properties for stationary storage, but Li batteries have a process advantage of several years. So the picture might change down the road as long as there is still enough money flowing into alternative research.

gozur88• 9 years ago

I wonder. One of the reasons internal combustion has lasted so long is great boatloads of cash get showered on the technology trying to eke out that last 0.2% of efficiency, and when that's the situation eventually someone has a bright idea that makes a bigger difference.

If lithium ends up with the same advantage, other technologies that ought to be better may end up lagging even in the long term.

sliverstorm• 9 years ago

We do this a lot. CCD image sensors are inherently more suited to astrophotography, but so much development has been put into CMOS that it is catching up to (and may surpass) CCD. A single widespread technology, with more applications & development effort, surpasses a better but more niche technology.

It's an interesting phenomenon. In the case of batteries, I wonder if we fully account for all the costs though. Lead-acid, for example, is a very sustainable (long-lasting, highly recyclable) battery technology that is inherently suitable, but if the full mining/recycling/disposal costs of lithium-ion are not priced in, the market won't account for this.

cwkoss• 9 years ago

I've been watching Joel Salatin's talks on his farming methods. Been surprised to hear that there is a huge amount of 'low tech' innovation occurring in farming within the past 50ish years because long-used materials and parts that were historically expensive to produce are now order(s) of magnitude cheaper.

- His mobile chicken coops need to be lightweight. Before thin-bladed sawmills, milling lumber to be smaller than ~2" was cost prohibitive as each cut had a 0.25" kerf. Making 1"x1" planks would mean nearly half of the wood goes into the waste pile!

- Cheap electric fences allow pastured hog farming to be mobile. Pigs will wreck strong fences, but a lightweight deterrent allows quick placement and removal.

Hasz• 9 years ago

LiS will be even more popular with grid utilities. Super high specific energy, but low overall discharge weight + low cost of Sulfur means a popular, stationary battery perfect for grid tie in. Ironically, it's even less dense than li-ion.

The current issue is commercialization (i.e, doesn't exist) and low conductivity of the sulfur cathode (meaning a low discharge rate to avoid self heating) -- it could be a decade + away.

derefr2• 9 years ago

Hah, I misinterpreted "storage" here, and thought this was going to be about Li-ion cells getting cheap enough to replace the capacitors in the Power Loss Protection circuits in hard drives/SSDs/NVMe boards.

Which doesn't make sense, after giving it a second's thought—batteries wear out after just a few years, while capacitors (very nearly) don't, so it'd put a hard lifetime on drives. (Though, for server use-cases with constant workloads, they have hard lifetimes anyway...)

But it's an interesting question! If—rather than UPSes as specialized enterprise-level hardware—we instead had Li-ion cells on computer motherboards (picture a big brother to the CR-2032 CMOS battery) that kept the disks alive for up to, say, 30 minutes after power loss—could we architect our storage subsystems differently?

(Sorry if that's a complete tangent from the topic at hand, but I'm not sure where I'd post this otherwise!)

wongarsu3• 9 years ago

Keeping the disks alive is certainly doable. At a rough estimate of 10W per drive, even storage servers would only need very moderate batteries. But what would you gain by that?

More interesting (in my eyes) would be to power down spinning disks and processors, but keep caches and ram alive. Then once power is restored, you could basically supply power and continue where you left off, without any reboots etc.

Then again, power loss should be fairly infrequent, and you have to architect against connection cuts or physical disruptions anyway.

StillBored2• 9 years ago

Congratulations, you just invented the ACPI S3 power state. Also known as standby or suspend to ram. I've been pretty happy with it since ~1998 when I started returning hardware that didn't support it properly. Although these days even linux on a laptop tends to work properly with S3 and S4 (hibernate, or suspend to disk). That said the concept was also part of the earlier APM standard which was a lot more hit/miss.

Anyway, its great, I used it today to ride out a nearly hour long power outage with a couple of fairly trivial UPSs connected to my NAS, desktop, etc. The desktop was already in standby because that is its default state after being idle for 30 mins. But the NAS went into that mode when the UPS indicated 70% battery life.

The problem of course is that UPSs/etc have miserable power conversion efficiencies when supplying just a couple watts of power, like 10-30% efficiency, so they tend to have max run-times less than 2 hours even when completely unloaded.

More on point, for servers, if your server is lightly loaded, or loaded for only parts of the day, using WOL to wake it and going into standby after a few minutes can save a fair amount of power. With SSDs, the resume times are generally less than a second unless you have bad cards/drivers that take forever to reinit.

maxerickson2• 9 years ago

Portable computers sort of do that.

I had a vague idea that I had heard something similar before:

http://www.zdnet.com/article/google-reveals-secret-server-ha...

aarongough• 9 years ago

If you have a look through some of the Google datacenter tours you'll see they actually implement this idea on a massive scale, though I believe they use lead-acid batteries... Every motherboard in the datacenter has a small battery capable of powering it for a long enough period of time for the backup generators to kick on and take over.

It's definitely a clever idea! Makes maintenance much less risky as it's distributed, and also reduces the chance of having a single large UPS fail when it takes over the load!

derefr2• 9 years ago

I guess the hardware does exist, but I was thinking of a different use-case for it: not for fault-tolerance per se (i.e. waiting for the power to come back on or the generator to kick in), but rather for giving the disk cache up to 30 minutes (after the rest of the machine is well-and-truly cut from power) to flush to the disk itself before finally shutting off, rather than a few seconds to empty its pending writes and park the disk head.

You could, therefore, have a really, really big disk cache, and a relatively slow-to-flush disk. With such an architecture, you could build systems that use disk+cache the way Optane NVMe is being used, without having even needed to invent Flash memory to get fast, highly-parallel writes first.

pjc502• 9 years ago

kept the disks alive for up to, say, 30 minutes after power loss

Not relevant to disks, but you could have battery-backed DRAM? Commonly used in RAID clusters as a cache.

The main problem with storage architecture is too much hardware that lies about whether it's actually committed a write.

baybal26• 9 years ago

IHS Markit expects li-ion battery prices to fall below $200/kWh by 2019

In that article they claim the digit of $200, while you can buy an assembled 1kw/h battery pack in China in retail quantities for around $120 today.

epistasis6• 9 years ago

Is that NCA or NMC?

Typically grid storage batteries are rated to 10+ years or ~5000 cycles or something on that order of magnitude.

One way to get your cells to last that long is to lessen the discharge depth; that may account for some of the price difference.

baybal26• 9 years ago

Most likely consumer grade LiCo with carbon anode. LiFePo modules are twice as expensive, demand for them certainly exceeds supply

burntrelish1273• 9 years ago

Any medium-to-large commercial campus with time-of-use-billing and large electrical bills may likely benefit from building a 500 kWhr - 3MWhr li-ion battery bank, 480VAC inverter, transfer and switchgear to buy electricity off-peak to use at-peak. The costs for lightly-used inverters and LFP module packs make it compelling to evaluate. For example, Jehu Garcia (@jag35) is involved in a 1 MWhr bank for a manufacturing customer in SoCal that looks really P&L sensible.

Falling lithium-ion battery prices to drive rapid storage uptake

Related Stories