UK’s largest flow battery energised | Hacker News Reader

UK’s largest flow battery energised

energysuperhuboxford.org

66 pointsvanburen33 comments

Vanadium flow batteries have gotten lots of interest because vanadium has 4 oxidation states, so the positive and negative electrolyte tanks have distinct charged and discharged chemistries. This adds a ton of flexibility to usage from multiple cycles per day to seasonal storage while the battery is still operating daily.

Manufacturers are advertising 20k cycles vs 6-7k for Li-ion, but in theory, flow batteries can last forever with maintenance.

Since TFA didn't really explain what a "flow" battery is: https://en.wikipedia.org/wiki/Vanadium_redox_battery

> In the ESO hybrid energy storage system, Invinity’s vanadium flow batteries are used to ‘front-end’ the energy asset, acting as a first line of response when the system is called into service; only after the required response exceeds the capacity of the Invinity battery does the lithium-ion battery get called into service. Since Invinity’s flow battery does not degrade with use and can cycle indefinitely, it performs much of the ‘heavy-lifting’ required from the system while reducing wear on the lithium-ion battery.

From TFA.

The mix of lithium ion and flow batteries here (50MWh of the former vs. 5MWh of the latter) kind of makes this feel like a proof of concept to me on the flow battery side -- I'd sort of expect the reverse, and to want to mostly lean on the flow batteries because of their suitability for long-duration storage and their tolerance of lots and lots of cycles, and mostly only use the lithium ion when you needed lots of power or to handle rapid changes in demand. But maybe I misunderstand the dynamics here?

If you think of the flow battery as a fast cache and the Lithium Ion as a slower one then that might help you form a better picture.

I think they covered this question pretty well here:

> Invinity’s vanadium flow batteries are used to ‘front-end’ the energy asset, acting as a first line of response when the system is called into service; only after the required response exceeds the capacity of the Invinity battery does the lithium-ion battery get called into service. Since Invinity’s flow battery does not degrade with use and can cycle indefinitely, it performs much of the ‘heavy-lifting’ required from the system while reducing wear on the lithium-ion battery.

What I think this means, is the the 5Mhw of flow batter covers demand spikes and the 50Mhw allows for larger longer draws during off peak renewable times.

I am just an enthusiastic auteur in this field and could be very wrong.

I think you're basically right. I know some people that looked in to building a flow battery (it didn't pan out for other reasons). The utility wanted it for frequency regulation - in other words, extremely short timescale response times.

But why? I thought one of the chief advantages of flow batteries was a lower cost per kWh at scale, as the expensive electrodes can be sized for peak power, while the cheaper electrolyte can be sized for capacity. Perhaps the economies of scale in lithium-ion batteries have had a bigger impact on cost than this clever split, so lion is actually cheaper per kwh?

Yes, exactly, this was my question as well. I understand their explanation for how they're using it, but it seems like the economics would be much better if configured differently.

Kind of battery cache?

FYI I think you are using “auteur” incorrectly.

This is a very interesting development. 55MWh is a drop in the bucket at grid scale but it definitely is a step forward if the installation performs as designed over a longer period of time without reliability issues.

For comparison purposes, also in the UK, Cruachan Dam has 7+ GWh worth of (pumped) storage, but this one even in its present state should be able to spin up even faster (though Cruachan Dam is very impressive in that respect).

At this size, it is likely acting as a grid stabilizer rather than any kind of backup. Even 'drop in the bucket' size batteries can have a very positive impact on a grid when used in this way that bring down costs for network operators, and increase allowable renewable generation on the network. They can do this because of their response times being in the millisecond range, rather than seconds/minutes for hydro(15+ seconds) and other generation types.

Eg, Tesla battery in South Australia mainly saves the operator money this way [0].

[0] https://www.pv-magazine.com/2018/12/05/south-australias-tesl...

Yes, precisely. The alternative is to have a much larger installation running idle 'just in case', this thing can react in the time it takes to spin up the larger installation if the load surge persists. They are super useful, the effect is the same as if you hook up a large flywheel to the grid.

Why is this news? It's a 5MWh vanadium flow battery. Another country put online a 800MWh vanadium flow battery (160x larger) this year [1].

> Not only will this be the largest directly-transmission-connected battery installed in the UK to date, it will be the largest vanadium flow + lithium-ion hybrid battery ever deployed.

This is like boasting you have the world first, largest ever potato+nuclear reactor energy source by throwing a potato in an existing nuclear reactor. Good PR for securing the next contract from dimwitted bureaucrats, I guess, but hardly a genuine exposition of cutting-edge technological development.

[1] - https://www.power-technology.com/marketdata/dalian-uet-rongk...

"Why is this news? Another country put online a 800MWh vanadium flow battery"

Did they actually finish building it and get it working and achieve a reasonable cost? Or is it just a big announcement and PR release so far? But yes if they do manage it that is amazing. Otherwise just peacocking.

Yes, the project started 5 years ago IIRC and completed this year.

It was scheduled to complete this year (in fact, it was originally scheduled to be completed in 2019, then 2020, and then it was supposed to be 2021), but it has as far as I know not been completed yet and has not been taken into general service. Proof to the contrary would be gladly accepted, if they get it to completion and it functions as advertised it will instantly be the largest of its kind in the world.

From what I know about it it got delayed twice by more than a year and isn't operational.

Spec sheet of a VS3-022 container (note website slow response - hug of death?): https://invinity.com/wp-content/uploads/2021/10/Invinity-VS3...

  Non-flammable  Safe by design  25 year lifespan  Unlimited cycles  Lowest LCOS  High recyclability

  CUMULATIVE ENERGY DELIVERED OVER TIME
  INVINITY VS3-022 3783000kWh versus LITHIUM 973kWh
  (Assumptions: 220 kWh DC capacity installed, 2 cycles per day, 100% DoD per cycle, 365 days a year).
  [Note corrected for typo in PDF: 3,783.000kWh should be 3,783,000kWh]

  PERFORMANCE SPECIFICATIONS
  Ambient Operating Temperature 25°F to 110°F (-5°C to 45°C)
  Nameplate Rating DC Voltage 1000 VDC
  Operating Voltage, Nominal 850 VDC
  Operating Voltage Range, Full Power 750 to 950 VDC
  Max. Continuous DC Current ±104 A
  Max. Continuous DC Power 78 kW
  Energy Storage Capacity 220 kWh
  Energy Storage Duration 2.5 hours @ 78 kW 4 hours @ 56 kW 8 hours @ 28 kW
  Max. Recommended Depth of Discharge 100%
  Cycle Life > 20,000 cycles
  Lifetime Throughput 3,783 MWh
  Annual Capacity Degradation < 0.5% per year
  Max. DC Round Trip Efficiency (RTE) > 78%
  Annual DC RTE Degradation < 0.1% per year

3.8 GWh lifetime? 78 kW DC, so about 70 kW delivered? Hardly seems a big battery.

Site had 27 containers installed, so maximum ~2MW delivered (really depends on configuration of inverters).

Implied by the installation is that the price of the batteries is a bit more expensive than Lithium batteries (risk adjusted price given the technology is less proven so perceived as higher risk). Vanadium batteries can do many more cycles before they lose efficiency, and presumably the liquid is easy to replace and recycle, so lifetime costs should be lower for Vanadium batteries in high duty cycle applications (as the article talks about).

About their cost: example 1: “The 0.5 MWh system would comprise two Invinity VS3 flow batteries, and was expected to contribute revenue [to Invinity] of approximately £0.48m, relating to the Invinity battery system itself, ancillary components and associated services.”; example 2: “Last week the company announced the world’s largest solar-plus flow battery project so far. The Yadlamalka project in South Australia will pair a 6-megawatt solar project with an 8-megawatt-hour vanadium flow battery. The project, worth 20 million Australian dollars (USD $15 million)”.

It is hard to know what else is in those systems (solar panels? inverters? support contracts?) so you can only infer the batteries cost less than those implied prices.

I am guessing another major advantage is reduced fire risk, albeit at 1/8th the density of LiIon. Nobody wants a Lithium battery fire like this 2MWh installation: https://spectrum.ieee.org/dispute-erupts-over-what-sparked-a... “[The fire] was initiated by an internal cell failure within one battery cell, identified as cell 7-2 on Rack 15.” “thermal runaway cascaded from cell 7-2 through every other cell and module in Rack 15 via heat transfer.”

They don't say why it's combined with traditional lithium ion - wasn't one of the promises of Flow that they can scale it up for longer storage just by adding more liquid?

It's not a matter of being able to scale it up, it is mostly a cost optimization.

Guessing blindly here, but maybe the liquid is pretty expensive still?

27x 20ft shipping containers combine to form a 5mw vanadium flow battery buffering a 50mw lithium battery. Which is interesting but the fact it's not 100% flow battery suggest lithium must still be sufficiently cheaper to overcome its shorter life.

> 27x 20ft shipping containers combine to form a 5mw [it's 5MWh, not 5MW] vanadium flow battery

This seems like a LOT of space required for such a low capacity. For a 5MWh flow battery, I wonder what the peak power output actually is from the system? for LiIon, you would see about ~1MW peak power from such a small reserve of energy. Flow batteries are longer lasting, so I assume this battery is less peak output. I am pretty sure that the LiIon battery will be doing the heavy lifting here, sized as it is. This isn't a huge leap forward.

Why do you say the enrrgy desnsity is low? Comoared to hudro this is very dense storage, an ordinary warehouse would back up multiple Gwh

5 MWh = 370 tesla powerwalls - for a grid energy storage this is a small POC

That is true. But the Tesla powerwalls would be using Lithium Ion tech, and that is used here too but 50 MWh worth of it. So the whole installation is considerably larger than what you write.

Or two megapacks at about $2.5M which is product targeting this commercial / utility space.

https://www.tesla.com/megapack/design

Be curious what this cost. Sure, li-ion has poor endurance but we can't really compare without a cost number.

Its like no one has heard of Stocking Pelham from 4 years ago:

https://stateraenergy.co.uk/Energy%20Storage%20Facility/pelh...

But that's just Lithium Ion cells.

Reply on news.ycombinator.com