Tesla’s Megapack Battery Burned For Days In Grid Storage Fire

Lithium rechargeable batteries have been heralded for their high-density energy storage, enabling all manner of technologies to come to fruition. From drones to practical electric cars to large-scale grid storage, the applications are endless.

The fire as seen from a drone overhead. Source: Twitter/@FireRescueVic

However, the lithium rechargeable battery has always had one major flaw–flammability. Pushed outside their operating range or otherwise tipped into thermal runaway, and they can burn ferociously as a result.

This came to pass in late July, at the Victorian Big Battery in Geelong, Australia, and it took significant effort to extinguish the blaze. Let’s take a look at the project and see how this came to occur.

Grid-Scale Storage

The Victorian Big Battery is a grid storage project similar in construction to the Hornsdale Power Reserve in neighboring South Australia. However, where the Hornsdale facility fields 194 MWh of capacity and 150MW peak power delivery, the new project aims to go much further. The Victorian project aims to install 450 MWh of capacity and deliver a peak power output of 300 MW.

Concerns were raised about the smoke from the blaze, though local authorities gave downwind areas the all-clear soon after. Credit: CFA

Operated by Neoen, the facility is built using Tesla Megapacks, large battery installations designed for grid storage purposes. Each Megapack contains batteries, inverters, and thermal management systems inside for a turnkey, plug-and-play solution to grid storage.

Initial testing of the battery was undertaken on July 30, with fire breaking out at approximately 10:15am according to official reports. The site was quickly disconnected from the grid with no interruption to the local electricity supply.

The fire burned for days, with firefighters announcing the blaze had been brought under control by 3PM on August 2. Lithium batteries tend to burn quite fiercely, and will often reignite after a time, so crews were left on site to monitor the battery for some time afterwards. Temperature readings were taken every two hours so that any heating or reignition could quickly be subdued.

Water Was Used, But Not Directly on the Fire

The fire was fought by the members of the Country Fire Authority as well as Fire Rescue Victoria. CFA incident controller Ian Beswicke spoke on the blaze, noting the difficulty of tackling such fires. “They are difficult to fight because you can’t put water on the Megapacks… all that does is extend the length of time that the fire burns for.” Acting on advice from Tesla, Beswicke noted that “…the recommend process is you cool everything around it so the fire can’t spread and you let it burn out.”

Visible are the individual Tesla Megapack units, each with its own thermal management system built-in. Credit: CFA

The fire began less than 24 hours after the battery began operating on the grid, according to sources quoted by the Sydney Morning Herald. Quick action by firefighting crews kept the entire facility from burning down, limiting the flames from spreading beyond a second battery pack.

Out of an abundance of concern for people in surrounding areas, a warning was given to those downwind of the incident. Two mobile units were deployed to the area by the Environment Protection Authority of Victoria. Despite early concerns and a warning for residents to shut windows and remain indoors, official reports soon gave the local atmosphere the all clear.

Obviously, a large fire lasting multiple days in a grid storage facility is an outcome that nobody wants. However, the events that followed serve as an indicator that authorities were well prepared to deal with the situation. No injuries were reported throughout the incident, and the fire was contained to a limited area of the facility. Only two battery packs caught fire out of the many on site (210 are planned in total), and electrical risks were properly managed to avoid disaster.

Questions remain as to how the fire started in the first place. Whether it comes down to an installation error or faulty equipment or batteries will likely be revealed by investigators in due time. For now, it’s a black mark against Victoria’s new battery project. However, in time, it may serve as an example of how through proper emergency management, lithium battery fires can be managed safely. The future of the electrical grid, and indeed, personal transport, may depend on it.

173 thoughts on “Tesla’s Megapack Battery Burned For Days In Grid Storage Fire

          1. Exactly this, though after this event they may spread out more in future installs when land is almost free, so its only the cost of a little extra foundation.

            Somewhat surprised its so densely packed in Aus, I mean the place is renowned for being big, very empty, for horrible hot weather and catching fire on its own! I always assumed the ‘big battery’ they had built were much more distributed arrays, if only to spread out the heat generation and allow more airflow, but also because it wouldn’t cost much to do so…

          2. It’s the logistics of it. You can see they’re using a crane to access the modules, so if you space them out too far you have to drive the crane all the way around to reach all the modules in one “block”.

          3. I was thinking more the fact that it seems to be a very large x-y plane with little space – I had expected greater gaps between row – both to allow service access and just keep the modules apart enough that if anything goes wrong only that column is at much risk.. Using a mobile crane more like the ones used in dockyards that straddle the loads, rather than just a giant single spot of contact with the ground style less mobile one.

            Just not knowing what logistical model they had chosen I assumed they would keep each of their battery crates in rows and the rows further apart using something like a forklift or dockyard crane… And that method allows you to put the rows however far apart you feel like, even over multiple satellite locations sharing the crane(s), the only limitation is the rows can’t get too close together that the service vehicle can’t get access – which with a system all about vast scale at the outset seems like a more scaleable model to further increase scale as/when needed.

          4. Australian land is extremely expensive, you would pay less for land in Texas than almost anywhere within cooee ( Oz Slang for Short distance ) of any constructed environment in Australia.

            I know, it makes absolutely no sense but it is true….

          5. @Kanniget

            Without knowing land prices in Texas I’ll lean on your knowledge/experience for the comparison, but saying “it makes absolutely no sense” is a just silly. Instead, you could say “I don’t understand why that’s the case”.

          6. I said it doesn’t make any sense, because It doesn’t make any sense. Real estate should have a value somewhat relative to the general demographics of the area. i.e. the ability of the land to pay for itself, through available local income for the owner or through direct earned returns.

            https://www.loopnet.com/Listing/1051-S-Highway-67-Cedar-Hill-TX/23071478/

            population approx 45000. median income $72K USD, 70% employment rate, on/near highway 6 Acres…$799K USD

            vs

            https://www.commercialrealestate.com.au/property/111-lillkar-road-goulburn-nsw-2580-2015927188
            population approx 22000, median income $32K AUD, 85% employment ( 30% part time ) and on/near highway, 3.3 acres $2.2M AUD ( roughly $1.6M USD ).

            and some of the reasons for this are :
            1) There are taxation loopholes that are extremely favourable to real estate land banking,
            2) Banks have no interest in loaning for business.
            3) A general population with an obsession that real estate is the only investment game in town and that thinks value is what someone else will pay for it in 3 years time.
            4) Essentially very low or almost non existent holding costs.
            5) Politicians at all levels who have large property holdings and use them to build wealth. Hence policies designed to prop up the market when it is already very strong.

          7. I didn’t see where anyone took into consideration the part about reaching down from a crane with some sort of lifting device to attach to a FLAMING HOT MEGA watt battery pack ! Now who is supposed to go in there to disconnect all that power while flames are wafting around then attach lifting clamps to the battery pack so it can be raised out of there on fire and then where to go with it? Lithium batteries may be nice for the power they can store but I don’t think enough engineering has been put into them for the reason they are so dangerous. We need a better solution and it’s coming, just not fast enough.

          8. @Kanniget Hang on, that listing is in another state..

            That said, I tried to search for similar in VIC, and every price said “Contact agent”. Makes me think it’s one of the “if you have to ask the price, you cannot afford it” situations.

          9. @SkitCoin I never said Victoria in my original comment, so when I produced evidence of the ridiculousness of Australian real estate “values” I used Commercial land in an are of NSW I am familiar with.

            There are a number of reasons Real Estate agents mark the listings as “Contact Agent”, mainly to avoid sticker shock and to give them someone to contact repeatedly to tell about the really great offer they have had on the place, and how you really should counter…

          1. I do live in Victoria, and own rural land. Three square km of poor farming land (mostly forest) lying 250 km from the capital is worth about the same as a middle of the range house in town.The cost of farms is currently surging, though, – across the state. The price of land suitable for a grid-scale battery may be a bit higher, because it has to be adjacent to a major power feeder, ideally near a solar farm. The small area required for a bunch of batteries, compared to the square kilometers for a big solar farm, can’t break the bank though. In extremis, just plonk the batteries in a spare corner of the solar farm – its land has to be cheap enough for the PV generation to be cheaper than coal, and make a profit, after maintenance and interest.

      1. So does the labor in fighting the fires, environmental fallout should the blaze not be contained, etc. At the very least there should be proper fire breaks between the packs to contain individual fires. It’s not if, it’s when. If this kind of thing catch on, there’s eventually going to be a Texas City-like incident.

        1. For grid scale transformers, there’s usually a high concrete wall and a bit of space. More or less designed to be able to buen out without taking out their neighbours. I’d assume that grid transformers are probably less likely to fail and cach in the first place too – just because theyre simplier.

          So perhaps design the concrete wall structure such that dumpung a truck full of sand on it will smother a fire? Or invent the lithium-asbestos battery :)

          1. Grid scale transformers are typically filled with a highly toxic liquid for heat management and to act as a dielectric. The concrete walls are designed to act as a tub to capture those chemicals in case of a failure. It helps contain the chemicals and prevent large scale environmental contamination.

            You will normally see crushed rock under the transformers but below that is another layer of concrete that makes the bottom of the containment area.

            It has a secondary effect of preventing problems from spreading to other transformers but that isn’t their primary function.

      2. well that is the same reason we started building tilt panel houses and modern ghettos (townhouse complexes) and allowed 400m2 blocks…. to save on infrastructure so the council officials could shuffle more money elsewhere…. and now we have lots of people in small places, easier to control them if needed and they are really dependent on handouts, high demand for services and supplies. ..

    1. There are hundreds of engineers looking into this but as far as I know there is no evidence that the fire spread even inside the 1 damaged Megapack. For the fire to spread, or propagate, means that a cell in a neighboring section or pack has to ignite. While the many pictures show damage it doesn’t seem to show the fire spreading beyond on of the 13-14 half cabinet sections in the first Megapack. If it did propagate it would loose it’s UL listing.

        1. Regardless of the potential Ameri-centric discussion, UL is probably still applicable in this case in terms of the design. Tesla is based in the US and undoubtedly subjects their designs for UL testing as they want to push their products onto American soil as well.

          1. An end mill isn’t electrical. Yes even and especially in industrial environments all electrical components have a UL OR equivalent rating. Literally the law. Yes putting in that chinese smart relay as a switch replacement could screw you over bad if it is the cause of an electrical fire.

            To over-simplify it electronics (but not low voltage or extra low voltage) is different, that’s covered by different code and I can’t comment on whether something needs UL rating.

            I say this as a Red Seal Industrial Electrician.

          2. In reply to Murray:
            “In former UK colonial parts of the world CE tends to be more relevant than UL. Now that we have Brexit, we will have to see what happens.”

            There’s a new thing already, it’s called UKCA.

    2. would be better to literally bury them individually with a wooden roof structure, so when they caught fire, they could self extinguish, or just keep all that carbon local into said pile of dirt or sand, or other cheap/free extinguishing material.

      Part of the reason they have these ‘cells’ is modularity…. while they technically shouldn’t have to shut down the entire park for this one cell, i don’t doubt that they did for safety and investigation as to why they set fire.

      while lead acid batteries are not as efficient, etc etc… they do not have as much of a tendency to spontaneously combust under load. If they don’t figure out this burning issue, they’re never going to reach the goals set fourth.

      They have to figure out better monitoring and safety mechanisms.

      however, it’d be better time and money spent to stop the military from literally scorching the earth with all of it’s backward policies including burn piles toxic waste, and using depleted uranium for munitions.

      1. Maybe buried installation would be a wise choice or underground compartments. The danger is heat from the intense fire overheating other packs. Lithium batteries are inversely stable with temperature. The hotter they get the more conductive they become increasing current flow. They will have to come up with some kind of suppression system. “Just let it burn itself out” is not going to fly in a lot of jurisdictions.

    3. Didn’t this event show they’ve pretty much picked the optimal spacing? At most they lit one adjacent pack (may have been part of the fault) under a full burn to the ground event for 2 units. Given rarity of the event and effectiveness of containment, it’s not really worth spacing them further. I am slightly surprised they don’t have sprinklers to remove dependency on fire crew arrival times. That might not go so well next time.

      1. It’s and and or situation. Suppliers globally can design and certify their modules/assemblies to not propagate, case of Tesla, OR they can build in sprinklers, typically actually it’s a water supply with a temperature controlled valve that floods the battery module to control the temp so the fire does not spread.

      1. Unlikely, there’s a lot more power in this unit than any car.

        (Also gasoline/other petroleum product fires are definitely a thing. Any sufficiently energy-dense thing has its risks.)

      1. I came here to something similar- some fire resistant shroud you drop over and leave it.
        But having read around a bit, ‘some fire resistant shroud’ probably turns your (not good) fire into a (really not good) bomb.
        One issue is that oxygen (maybe) is being produced by on-fire battery to fuel fire.

        https://www.e3s-conferences.org/articles/e3sconf/pdf/2021/14/e3sconf_100res2021_09001.pdf
        https://nobel-fire-systems.com/industry-sectors/renewable-energy/bess/

        google ‘bess lithium fire’ for more.

        The acronym/term BESS (Battery Energy Storage Systems) could maybe included in the original article?

    4. Anything that burns for three days and can’t be extinguished with water is a problem. They and emergency response for car crashes will need to come up with something better. In this case they were lucky. Maybe they could Design a type of snuffer, similar to a candle snuffer that can be placed over the burning unit and let the fire burn the oxygen out, thus removing one portion of the fire triangle.

      1. Isn’t that the case with the peaker plants these batteries are replacing? They are in their way out but there’s still tons of diesel and fuel oil peaker plants around the world and there are several MJ of energy in those huge million gallon fuel tanks. They do catch fire and they take a long time to put out, the plume of smoke is toxic.

        1. This is very true, while Lithium battery techs do pose a different fire fighting challenge it is not like the various fuels, which are fuels because they burn reasonably easy do not pose significant risks themselves.

          All you can really say in favour of the old systems is everyone is more familiar with how to deal with, so it seems old hat. The ol’ familiar breeds a degree of complacency and blindness to the problems – as that is just the way life is.

          Its not just the end user of fossil fuels own stockpile you have to consider either – as they need to actively receive shipments of whichever fuel from the refinery, so there is a risk of a big fire on the road/rail/ship/pipeline. And to actually create fuel the refinery needs still pretty flammable crude oil, which has to be shipped in from somewhere – so even if these batteries prove a little more unstable than ideal all the various fire risks – and the scale of the devastation (see the tanker the front fell off for instance) when a step in that line does go wrong means they are probably worth it!

          Plus one small well managed fire with huge potential energy here and a few toasty wrecks is far to early to cry wolf over the whole lithium battery technology – though I would like to see safer chemistries and other energy storage methods in use more, especially when there is not a real need for the highest possible electrical potential density, but still lets see what the final analysis of this grid battery event, and just how common runaway EV fires become in the real world actually turns out to be before even thinking that…

          So far it seems to me as nasty as lithium batteries can be they are not inherently worse, if anything they are safer on many fronts including fire risk – as all the many steps of generally much worse fires if there is an accident they are largely replacing with just one step – the electric transmission (though perhaps electric generation should also be counted, at least for now, as there is enough fossil fuel in use to make it, and that won’t disappear overnight).

          1. > they are not inherently worse

            A lithium battery contains the entire fire triangle: it has combustible materials, it produces oxygen, and it spontaneously generates heat when damaged. Fuels generally have only one part: the combustible material. To me that spells “inherently worse”.

          2. So what if they contain it all – we live on a planet with an atmosphere that pretty much makes certain that anything that wants to burn has the oxygen. A car of any sort is full of electronics and/or hot surfaces to potentially ignite whatever form of fuel is in use… In effect they all have all points of the fire triangle at all times! Most ICE cars have a really hot exhaust basically right next to the fuel tanks…

            So what matters is how often such fires happen, how significant and manageable such fires are – and so far Lithium fires have been reasonably rare on the roads, lots of crashed EV without fires, to the point I’m not sure they are really more likely to burn than ICE – but its still very early days, with so few EV accidents its statistically not really relevant yet – and till older EV perhaps prove more combustible as well its not a fair comparison to all the rust buckets that really shouldn’t be on the road getting into accidents…

            The management of an EV on fire is definitely worse than the ICE, as you can’t actually properly win in a reasonable length of time on the road – but you can move ’em to somewhere safe and let them burn out easy enough.

            But if/when a tanker/refinery/fuel lorry accident happens its orders of magnitude worse, and the more of them working you require because you keep burning fuel the more likely such an accident. And that for me is where Lithium battery become no worse, as even though management of a lithium fire is more challenging, it is not from all the evidence so far that much more challenging, but the scale of disaster (environmental in general) that is at all likely in a highly lithium driven system is smaller than any of the fuel delivery and refinement accidents – so much less potential energy in any one space to go wrong with lithium. When the ICE supply chain goes wrong even on a small single fuel truck scale its many, many EV’s worth of burning, and a significant environmental impact dumping all that oil based stuff even if the fire is rapidly controlled or it miraculously didn’t catch fire at all..

  1. Actually, the site layout could be designed differently, in a way, that when a container unit would have a thermal run-away event it could be safely disconnected from grid and be ejected away from the other containers to an empty row safe area.

    Perhaps even with moveable brick walls made from aerated concrete (light, non-flamable and good insulation) that could be shifted to heat/water shield the other containers.

    This way the firefighting/cooling would be easier and the danger from igniting other containers would be significantly reduced and the site would not need to be shut down and can continue to operate.

        1. I suppose I see the heat problem, but those Tesla Megapacks don’t look like solid individual batteries, but battery systems, I’d think you could enter the packs (from the top if buried) for mnaintenance.

      1. ah well, it’s what they do here with a burning EV, ditch it in a large container of water, to cool it, then wait when it is done, it requires a lot of water, and well, this may be a tad too much energy indeed, this is not your average car battery pack.

        1. The water also reacts with the battery chemicals to consume them without sustaining a flame. It doesn’t actually stop “burning” under the water – it only slows it down to a more controlled rate.

    1. Pfff, no self-respecting terrorist targets these things, especially since the grid is distributed – you might make a mess at the site (which is isolated anyway) but the grid will chug on regardless, and it might rate a news article or two. Not worth the effort.

      1. Are you really sure about that?

        Distributed means “spread out”, it does not necessarily equate to redundancy. The fact that a grid might be relying on secondary storage (remember these grid pack do not generate any new power, merely store excess) could result in a relatively easier way to destabilise the grid because generating capacity has been taken out (i.e. coal, gas etc.).

    2. Australia has its own redox battery, using cheaper zinc-bromine chemistry. It’s made by Redflow. Also marketed to households, it’s now getting a lot more traction with telcos around the world, and beginning to be put into big container-sized units for grid-scale storage.

      As each quarter-ton 10 kWh unit contains 190 L (that’s 200 quarts in Texas) of fire-retardant electrolyte, they are claimed to be more than a little resistant to burning. Redflow is working on a 2 MWh installation stateside now, I hear.

      For my off-grid build, it’s a choice between a pair of them or a bunch of Lithium Titanate batteries, which are also claimed to be long-life and less self-immolating than other lithium chemistries. Now that we have choices, all we can ask is that prices decline. Yeah, the early adopters have to drive up the quantities for that to happen.

      While most flow batteries are humungous, the plan here is to drive economies of scale by spitting out thousands of the smaller units from an automated factory. If one goes bung, it switches off-line, and the installation keeps on trucking.
      A massive reflow battery can’t do that.

    1. The facility was still in the process of being commissioned, it was not operational and the only batteries that would have been effected are those in the single unit itself. There has been lots of misreporting on this from Australian media ( backed by anti renewables ‘news media’ ) so I am not surprised.

        1. This.

          So despite what Kanniget would have you believe, it seems like none of the mainstream media outlets are prepared to go against the renewables juggernaught by reporting this.

  2. The solar farm they built on my friends grandfather’s old sod farm had quite a few growing pains. The units that take all the solar energy and send it to the distribution point, think there’s inverters and transformers in them would just spontaneously explode! He still lives on the property surrounded by 800 acres filled with panels and he heard them going off and said it put off a bad smell, so probably transformers magic smoke. They never told him why almost every single one had to be replaced. Now it’s a lot rarer but still occasionally happen during heavy rains.

    1. Well, with a Lith-poly R/C battery that I was ‘disposing’ of by driving a nail through it (usually get a nice little fire going) , I tried putting water on it…. Seems to ‘increase’ the chemical reaction that is going on :) .

    2. Pouring water on an alkali metal fire will produce heat and hydrogen gas, so in effect you are adding more fuel into the fire. It would be like filling a traditional sprinkler system with petrol, it is not going to help.

    3. I think this is a good idea for use on neighboring battery packs to help cool them and prevent the fire from spreading. Obviously on the battery pack that has the fire should not have sprinklers running on it. Some automation on this could help prevent an expensive nuisance turn into a full catastrophe.

  3. was there no option to lower an upside down box over the battery with a crane and flood the battery in argon starving the fire of oxygen and any external sources of moisture ?

      1. In contrast, a data center generally has twenty or more large cylinders of CO2 or nitrogen piped to the server rooms. Fire alarm goes off, everyone scarpers, and the air is flushed out. If the battery units were sealed, with heat exchangers in lieu of simple ventilation, then fire suppression could be fully automatic.

        Zinc bromine flow batteries would be an alternative, I figure.

      2. Toxic gases produced by the Lithium-ion battery fires might be why.
        https://www.nature.com/articles/s41598-017-09784-z
        hydrogen fluoride (HF) may be generated, ranging between 20 and 200 mg/Wh
        I do not think that anyone who reads here does not know what HF does – totally scary stuff.
        phosphoryl fluoride (POF3) was measured in some of the fire tests 15–22 mg/Wh
        (Poison, corrosive, can form HF on contact with H2O)

        1. Yup. The electrolyte in lithium batteries is lithium hexafluorophosphate. When it comes in to contact with water, it splits and combines with the water to make lithium phosphate and hydrofluoric acid. Nasty stuff.

  4. in 10 years from now, car accidents will kill people not by the shock itself, but by the resulting fire.
    We start seeing those in the news, where the driver could not escape and died by the fire, and the firemen could not do anything either.
    There aren´t much electric cars with big batteries on the roads yet, so it´s still anecdotal… for now.

    1. It certainly could happen, but I don’t think it is all that likely to happen, any more so than stuck in a burning ICE. Both can happen, and there isn’t a great deal to do once the fire is really going to save the poor trapped sod – its not like petrol or diesel is easily put out safely, or you can just walk up to something that may decide it does want to explode, being on fire and full of fuel vapours to cut your way in…

      The thing that is bound to happen is the serious EV accidents taking longer to clear up safely because the battery is cascading and hard to get too with the right sort of fire suppression to kill that chain reaction.

      But many if any extra deaths from being stuck inside the burning car assuming crash safety cells continue to be a thing required in new cars are very unlikely – a crash bad enough to get you properly stuck and actually rupture the battery probably caused you enough injury you are dead and just don’t know it yet anyway…

      That said I’d rather not check out of this life that little bit early burning to death, even if I was going to die of other injury soon anyway. It seems like one of the nastiest ways to die, and more burnt dead in the accident I do see as more likely with the current battery chemistry being so volatile…

        1. I would not call that stuck myself – as anybody coming to help can move you. To me Stuck meant car is a crumpled mess that you have to cut into to free the person.

          It is a good point though, just not how I had considered ‘not able to escape’ – as first responders should with all the crash safety stuff be able to get you out rapidly if there is a need to in almost all crashes.

          1. Also, if you have a neck or spinal injury, not just anybody should try to move you.

            A regular fuel fire can usually be put out with a handheld extinguisher before it spreads, but a battery fire can’t be put out once it starts, so the rescuers have to risk paralyzing the person as they rush to pull them out.

          2. In a heavy frontal accident not through my own fault, and ended up in a ditch and then head on into a telegraph pole located the other bank of it here. In a ex-military 4×4 truck (in civilian use, its a collector vehicle, running on gasoline), and the whole front of the vehicle had distorted by the two impacts around us and I think we even ruptured a fuel tank as I could smell gas. I managed to get out with steam burns to a lot of my body, and my daughter was trapped inside with a bleeding leg but no steam burns thankfully. WIth one hand working I isolated the electrical systems and shut off the fuel at source to try and reduce the chance of fire & wait for the first responders to extract her, and kept her calm while trying to remove hatches and vehicle sides/tilt to help access because I didn’t want to make her injuries worse trying to drag her out alone. 45 minutes they took to arrive and get her extracted… She’s fine now (and I was ugly to start with), she had a small fracture on her kneecap from the fire extinguisher mount bending into her and now just a tiny scar. You can be stuck in a crumpled vehicle without being next to death… And we’re both still alive because of how it was constructed from decades of design improvements incorporated into its layout and fuel supplies etc. I like EV’s also and own one (a VW e-up..) , but we need to think about adding standard emergency ports for first responder units to pump water INTO the packs to cool them if they’re hot and compromised inside to stop thermal runaway while extraction can take place, not abandoning them to their death because oh well it happens. It takes a LOT less water to directly cool a pack in situ to stop a fire than if its sprayed onto the outsides.

          3. The battery fire can be suppressed, and slowed significantly, and unlike a vapour full ICE car it shouldn’t ever be able to really explode..

            Its all swings and roundabouts, and getting to a fire early in either case is hugely important, don’t get there early and managing it in a way to save any occupants is basically impossible.

            You are obviously quite correct on not moving people with suspected spinal damage on a whim, but if you have to you have to, and that decision is made often enough that its not like an EV is unique in potentially forcing that decision.

          4. @”I only post when I have something to add”

            I really like that idea – a fuel filler cap for flushing directly to the battery, not sure it would actually work very well though – the batteries are not full of tubes to run the coolant through etc so actually getting the water to the damaged cells in sufficient quantity without redesigning the battery such that its structural integrity drops, making fires much more likely, and probably making them spread through the pack very much faster before help arrives…

            Sounds like you did everything right (other than being unlucky enough to be in bad accident), but unfortunately that doesn’t stop a fire being possible and had it started before help had arrived you would have had to move your daughter yourself, even if you could access the fire extinguisher (assuming the mount was populated) actually truly stopping such a fire with a single extinguisher is going to be tricky in such an event. I am definitely not suggesting abandoning folk to their fate, but it is the way of the world that sometimes nothing can be done, shit happens all the time to some poor sod…

      1. Having heard my father’s tale of when he was witness and first responder to a single vehicle accident where the drive shaft to the rear wheels snapped, kicked out sideways, then simultaneously flipped the vehicle and tore through the fuel tank into the cabin area suddenly filling the whole car with fuel mist, which ignited in a massive fireball engulfing the occupants who were trapped alive on fire breathing nothing but flames, I can tell you I would prefer an EV fire any day.

        1. Panel Van back in the 1960’s. Things weren’t as well designed then as they were now. I have seen the news article clipping from the paper years ago so I know it’s a legit story, but pre-dating the internet and most computers trying to find it online now will be next to impossible.

          1. Hydrogen in a balloon goes up. Hydrogen mixed with air lingers around because it loses buyoancy, especially since it’s cold after spraying out of a high pressure tank.

      2. firefighters can easily put out a fire in an ice powered car. They actually practice this all the time, and I have first hand experience seeing them do such a thing.

        1. That invites so many jokes. Are they ice wagons? Are tongs used to refuel the car? Does the engine consist of uniformed Immigration and Customs Enforcement officials pushing the vehicle?

    2. Electric car fires are rare per capita. petrol car fires are not. Once ecars catch fire they are hard to put out, petrol fires, on the other hand, usually put themselves out more quickly when they explode.

      1. I like your thinking. Clearly this thread leads to suggesting nitroglycerin powered cars as beng the most fire safe as they would have she shortest burn time.

        Over on the Museum Of Retro Technology, there’s various articles about steam engines using things other than water for the steam. Boiling petrol and ether are mentioned and the resulting fires commented on.

        Seems we humans have been attempting flammable transport and energy solutions for a while. When will we lears?

    3. Chemistry and packaging has a large say. 500,000 Nissan Leafs made and and 0 confirmed battery fires. Leafs have a very robust double armored battery compartments and a different physical format designed for safety, additionally until 2017 no cobalt (manganese instead); so deliberately more stable battery chemistry than say Tesla.

  5. I guess the foam they use to fight aluminum fires does not work. Too bad. Seems like the initial testing after install needs to be broken up and stretched out with maybe more sensors (temporary?) to track current, temperature and such. And temperature checks every 2 hours seems very risky too.

    1. Not sure if it’s the same stuff, but military aircraft firefighting foam is made of some nasty, persistent chemicals that really enjoy getting into groundwater.

      Given that I have temperature sensors around my house reporting every few seconds, I assume they didn’t go 2 hours between internal temp checks. Something probably got mangled in reporting. Maybe they took a thermal camera around every 2 hours?

    1. That it proved to be enough this time doesn’t mean it really is enough – the environmental conditions can play a really big part in how easy a fire spreads, and until this is properly studied we just won’t have enough information to draw on to know if this spacing is enough to prevent issues in almost all weathers, with any delay in the containment activities for whatever reason, etc.

      I have to say it seems it was handled well, and sounds like the spacing might be enough with the safety systems as designed. But its too early to really tell if the design is behaving as well as expected and should therefore really be safe – and pushing them further away definitely has advantages to preventing a larger cascade – so when its cheap and easy to do so perhaps it should be done…

  6. How much more efficient is Lithium over lead-acid? I know there are size & weight savings considerations, but in a application like this that should not matter.

    Plus lead acids are cheap, and easily recycled.

        1. IIRC the charge/discharge efficiency of a lead acid battery is around 60%, while lithium gets to 90% or so.
          Aside from that, lithium cells are hermetically sealed and maintenance free. The type of lead acid battery suitable for grid storage (Yuasa has 1 series of grid storage suitable, deep cycle, long life batteries) has to be checked regularly to see if the acid needs to be topped up.
          The efficiency is really quite important, for as long as we don’t have a true energy excess.

          Finally, when overcharged or short circuited – both very realistic situations – they will generate hydrogen and the explosions can be intense. That doesn’t typically cause a fire, but you can imagine what the acid does to the whole installation.

          1. wikipedia says lead-acid batteries have a cycle efficiency of 50% to 95%. I presume the high and of the range can be achieved if the conditions are ideal: don’t overcharge, don’t deplete, keep the temperature comfortable, and charge/discharge slowly.

      1. The national highway traffic safety administration used to argue against drunk driving on the basis of the statistic that “half of all highway fatalities involve the use of alcohol.”

        Ironically, on the basis of that statistic alone, an uninformed individual could logically argue that an equal number of lives could be saved by banning *sober* driving.

        The point is that statistics don’t always mean what they appear to mean and logical deductions based upon them can result in incorrect conclusions.

        Where do you get your metrics, and what is being presumed about the energy cost of lead acid battery manufacture? Are you including the cost to mine and refine the lead? If so, since lead is one of the most recyclable metals there is, the systemic energy cost of lead acid batteries goes down the more often used up batteries are recycled.

        Now how much of a lithium cell is recyclable? It may have higher energy density, but if the only thing you can do with spent lithium batteries is to landfill them, thats not really a sustainable solution, either. Thus, the lead acid solution may well be better, both from a cost and environmental impact standpoint.

        Mind you, I’m not arguing a position here one way or the other, just saying that figures don’t always mean what they appear to mean at first glance.

        By the way, if you measured every joule that came out of your average solar panel over its usefull life, I strongly suspect that it would not offset the total energy expended to refine silicon, grow purified photovoltaic crystals, convert petro chemicals into insulation, adhesives and coatings, and mine and refine the aluminum, glass, copper, and other materials needed to produce that panel.

        This is a more damning indictment of the solar panel than a lead acid battery, for example, because the lead acid battery is just a storage device, while the solar panel is alleged to be a viable primary power source.

        1. It’s called the ESOEI – energy stored per energy invested.

          https://en.wikipedia.org/wiki/Energy_return_on_investment#ESOEI

          In the ideal case, probably requiring more optimized designs, a lead-acid storage battery can store about five times the energy spent in constructing it. Previous studies I’ve read said 2-3.

          The number represents the optimal use of the battery. For example, in the source article, the figure for 32 lithium-ion comes from assuming a 6,000 cycle lifespan. If you only every use the battery for say 500 cycles then the ESOEI will be proportionally worse.

          The ESOEI also attempts to count in the round-trip efficiency of the battery. Lead acid batteries are terrible in this sense because they self-discharge and they are trickle-charged etc. for maintenance which represents additional energy losses, which counts against their ESOEI but this is harder to quantify. In the article the PbA battery is assumed to have 90% efficiency for 700 cycles, but in practice you may get 50% efficiency for 300-400 cycles (or full cycle equivalent partial discharges).

        2. You can definitely count every joule no matter how remotely tied to the making of a solar cell and get way less than it should produce over its lifespan!

          Just how much less is going to be very variable – for instance if you set up a fixed install pointing at the mid-day sun you will get a greater amount of power out of it than pointing at the morning or evening sun – but evening sun is actually a good idea sometimes as that is when in many households the peak of consumption is. Then you have the environmental conditions where you set up – if its a really dusty environment and the panels are never/rarely cleaned that will be worse than somewhere less dusty but otherwise comparable etc etc.

          But the biggest variable is what you call the Solar cells lifespan – modern ones really don’t degrade the way early ones did (and those ones still work at very reduced output if they avoided physical damage). So saying the lifespan is 20 years is low, looked after the useful lifespan is probably going to approach infinite in human terms -for longer than you will live it is likely to output a very useful amount of power (though the electronics behind it will probably need replacement a few times in its life), and averaged out for all the ones that do get damaged still well into the multiple decades…

          1. The EROEI of various crystalline silicon panels is around 10-20 in an ideal location, like equatorial Africa. Going up north drops it to around 3-6 in places like Scandinavia.

            CdTe panels were supposed to be around 30-40 but then the Chinese dumped prices and competed all the western thin film manufacturers out of the market.

            Just because a solar panel can theoretically make ends meet with a good margin, doesn’t mean the panels you buy actually do. Cheap price comes with tradeoffs, and the usual tradeoff is that the cheap panel is made somewhere that gets power from an ancient coal plant without emissions controls at a terrible efficiency.

          2. Oh I agree @Dude where you get your panel, and how its made changes its green credentials. But in terms of Energy in to out it really doesn’t change much.

            But as the panel lifespan is near as we can tell going to be basically infinite, just at reduced output with current generation panels the EROEI numbers become very woolly.. As with near infinite time producing power they have to pick a lifespan that makes sense more from an economics return on investment point of view than would be the case with most techs were the lifespan or cycle count are very much more finite. I think those numbers you quoted were for 5 years, possibly even less…

            So when the panel has much much much more life than that even the dirtiest produced panel should vastly pay back its carbon cost over even a shortened lifespan, and they will every single time pay back the energy cost in very little time – as that is independent of how dirty the power source for the factory is.

          3. There’s no such thing as infinite lifespan solar panels. Even a simple electric cable has a limited lifespan because the insulation deteriorates over time – the cabling in a house should be rebuilt about every 50-70 years to avoid risk of fire.

            Solar panels have a number of non-ideal properties that thwart the theoretical calculations. For example, if a bird poops on a panel and covers one cell in a series, that will develop a hot spot because the shaded cell turns into a resistor. In the worst case this can crack the glass. In the best case it just ages that cell faster.

          4. Coatings on modern panels make bird dirt much less of an issue – the panel surface during the heat of the day ‘self-cleans’ pretty well. Its far from perfect of course – and nothing is truly infinite, but in terms of power generation a solar cell is by an enormous margin the most long lived without replacement parts…

            With just a little looking after a solar panel should keep working for a very very very very long time… Infact there are heaps of them around of much more primitive construction that are still just working at reduced output after decades…

    1. Size, factor if 2. Weight, factor if 2. Cycles, factor if 20. Discharge ability, factor of 2. Flat voltage during discharge, factor of 20%? 2x2x10x2x1.2… factor of 96? Now adjust for price ratio. Maybe overall factor if 25-50?(Disregard that LA self discharges a lot more in standby mode.) Some studies say we haven’t enough Li in the world to satisfy potential car demand… was that just fir US demand? Anyway, there’s not enough, even if none goes to a landfill.

  7. These installations are exquisitely vulnerable to extortion and terrorism as a single cheap drone and a thermite payload is all it takes to cause this sort of chaos. They should put them in tunnels under the ground and have an inert atmosphere.

  8. Recent comment by a friend: “a few months ago, a pizzeria in town had a major fire – some electric scooter standing outside had a runaway battery meltdown, and took all the nearby scooters with it – firemen had a very hard time, ended up letting them burn out – the pizzeria + apartments above had massive smoke and water damage”. Sounds like we will see more of this…

  9. Never understood why Lithium is used for grid level storage. We have enough space in Australia – why not good old serviceable and non flammable lead acid? Need to vent excess hydrogen but much more robust I would think.

    1. The facility would be an order of magnitude larger and still not deliver anywhere near the same instantaneous power.

      Australia is large, but land is still expensive anywhere near cities. You can go further out but then you have logistical difficulties in install and maintenance, and often agricultural land isn’t cheap either.

    2. Lots of reasons, including low cycle life for resonable depth of discharge, not getting as much energh back out as you can with lithium chemistry’s, and bigness. Rather limited operating temperature range which might be a problem in some places.

      But they are inexpensive and easy to recycle so perfect for small home solar setups. They still seem quite viable for UPS solutions too, where cycle life is less of an issue.

      However, they too can catch fire too. Anything that can supply a high short circuit current can. Seen a few charred bits of wall in old battery rooms. But at leastthe rooms are mostly still there.

  10. still better than the coal mine that caught fire and burned for 45 days

    burning battery pack -> send the EPA out to warn people

    burning coal mine -> don’t warn people until they start dropping like flies

    https://www.abc.net.au/news/2020-05-19/hazelwood-power-station-latrobe-valley-2014-mine-fire-fine/12261858

    https://www.abc.net.au/news/2014-09-02/hazelwood-coal-mine-fire-morwell/5711564

    and now we are getting a nuclear waste dump!!

    joy…

      1. There are a few coal mines that caught fire, the one fire in South Wales seems to burn for thousands of years already, and there are some newer fires that were started in the industrialization period that still burn. In the end, we are all kinda floating on a bed of molten rock, and i hope that our core fire does not stop any time soon…

  11. In reply to Murray:
    “In former UK colonial parts of the world CE tends to be more relevant than UL. Now that we have Brexit, we will have to see what happens.”

    There’s a new thing already, it’s called UKCA.

  12. I just watched a demonstration of a fire blanket that can be draped over a burning EV.
    (On the “Guy Martin” program about electric cars)
    Seems like a larger version would be ideal to contain this problem, you just pull it over the battery pack, and it excludes the O2. Then you leave it for many days for the pack to cool down.

    If done early enough it would be possible to save some of the cells in the pack, to boot. Probably.

    1. Don’t know if I’d want to save the cells in a partially burnt pack – at least not to reuse in an “industrial” scenario. Ebike batteries would be a good destination for them.
      Mind you, there’d be nothing wrong with most of them, but there could be latent issues caused by being overstressed that will come back to bite.

  13. For anyone interested, I track battery fire and explosion incidents as a part of my job doing fire prevention research. Here is a link to a web page for viewing some of the data.

    http://tools.utfireresearch.com/apps/incident_map

    The things about lithium-ion battery fires is that there is a lot of disinformation. In general, assertions that lithium reactivity with water means that it shouldn’t be used are incorrect. A lot of this comes from confusion with lithium batteries (think watch batteries) for which this is true. Lithium-ion batteries have comparatively little lithium and it is not in its elemental state.

    The concern with lithium-ion systems is typically that water could cause electrical faults and shorts in undamaged cells. For example, a fire aboard a Norwegian hybrid electric ferry several years ago was probably made worse because of a seawater sprinkler system installed in the battery room. For cases in which you aren’t trying to save some portion of the system, this goes away. Water is frequently used as a tool in fighting these fires. However, the goal is usually not extinguishing, but cooling to keep the thermal runaway from propagating between cells. This requires copious amounts of water though and is often only marginally effective.

    As others have mentioned, a common strategy for EVs is to fully submerge them in a large container of water. This also has the added benefit of providing a discharge pathway for any remaining energy in the system. Reactivity and fire potential drop rapidly with state-of-charge.

  14. I told my township’s Fire Marshall about using a battery tender to charge my neighbor’s car and he told me they had at least two fires and lost a fire truck or more that burned down. I was told to be careful or not do it and unless you can be there and watch it, there isn’t always something you can do about it.

    I may not want a battery backup in my house for this reason. I’m extremely concerned about lithium fires. If you want the risk for your house, I have nothing to do with that but I can’t afford to have my house burn down because that would be bad and its not something people can recover from.

    A Tesla Model S Burned Down Someone’s House

    https://www.motorbiscuit.com/tesla-model-s-burned-down-house/

    1. Nice vid!
      Looks like the problem was not caused by stabbing (and withdraw), but caused be stabbing and leaving the knife in there long enough (not long!) for it to short out, get hot, and catch – in a most spectacular runaway fashion!

  15. It would have been great if the whole lot went up and we stop this nonsense and go for a proper energy source, like nuclear, which can supply basal demand without storage.

      1. There’s some newer designs which have a reactor cooled by molten salt (not a molten salt reactor) with a large pool of the salt to act as a heat reservoir, so the actual power turbine can spool between 0-100% while the reactor is providing the daily average power.

        But even regular Gen III reactors can follow load and run up and down between 50-100% on a daily basis. They’re just less economical to run that way because a lower average load factor means less energy produced over the permitted lifespan of the plant.

        1. In fact, the ability to do load following is one of the design criteria of new nuclear reactors:

          “the NPP must be capable of a minimum daily load cycling operation between 50% and 100% Pr, with a rate of change of electric output of 3-5% Pr/minute.”

          Even Gen II reactors currently in use in France and Germany are capable of at least that, and many are actively used in load following mode, especially in France. The 50-100% range matches approximately the daily variation in power demand, but EDF sometimes runs the reactors all the way down to 10-30% in times of strong negative prices, usually caused by renewable power.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.