Battery-swap networks are preventing emergency blackouts


136 points | by belter 13 days ago


  • neallindsay 13 days ago
    There's lots of talk about grid-scale batteries to shift cheap electricity from solar panels into the night time. But batteries at the periphery can also help a lot because they can just stop charging (like the example in this article) or even feed energy back into the grid. Electric cars mostly don't have the circuitry to do this yet, but they're becoming a huge resource that will be tappable this way. Most of them are massively over-provisioned for the way that they're used 350 days out of the year.
    • mauvehaus 13 days ago
      Green Mountain Power does exactly this in Vermont: they have a program that'll install a pair of powerwall batteries at customer homes and charge them off peak to draw on them on peak, particularly to avoid GMP having to buy peak-rate power from their upstream suppliers. In return, they leave the batteries charges when the weather is turning bad and the customer has battery backup.

      The deal is structured as a 10 year lease, and the total cost is vastly less than that of owning the batteries outright. Their installation allowance is pretty good too; it didn't cover splitting our panel, but it covered everything else.

      Not affiliated other than being a very happy GMP customer.

      • hedora 13 days ago
        I wonder how the lease is funded. Usually these sorts of incentive programs have a big catch.

        That happened a lot with solar panel leases, so now people have solar panels on their roofs that are effectively owned by a bankruptcy courts, which can make it hard to sell the house. (There's no one to contact to transfer the lease to the new owner.) Since GMP is a power company, that's probably less of a risk.

        Anyway, the other features you mentioned are pretty standard. Enphase does all those things, and I'm pretty sure powerwalls, lg systems, generac, etc do too.

        This page has info on the GMP program:

        So, $55 / month for a lease of two powerwalls (26KWh total) which is $6600 over 10 years. They'll also provide a $10,500 incentive if you purchase your own battery. A powerwall costs $9300, so it'd be $8100 out of pocket to buy them instead of leasing.

        (The above doesn't include installation.)

        I wonder what happens at the end of the lease. Either way, the lease looks like a better deal, even if you assume a zero percent interest rate.

        • pchristensen 13 days ago
          Utilities can have huge capital budgets, and programs like this are weighed against the cost of building new power plants or contracting with external power sources. It's money that was already going to be spent - programs like this redirect funds.
      • onionisafruit 13 days ago
        You say they leave the batteries charged when the power is turning bad. Do you have a benefit of hosting the batteries at other times? For instance if the power goes out when the weather isn’t bad, do you typically have enough charge to make it through the outage?

        I’m also curious if your neighbors benefit from your battery during outages.

        • mauvehaus 13 days ago
          De jure, I think they're pretty free to do with them what they want. De facto they leave the batteries pretty fully charged most of the time.

          AIUI, they'd have to have a disconnect somewhere upstream of us to let us back feed the grid during an outage, otherwise they couldn't safely work on the lines. And they'd need a way to enable and disable back feeding remotely[0] to allow them to work on anything between that hypothetical disconnect and us.

          At the moment, none of that exists, so only we benefit from the powerwalls. It's also 34kwh in the pair of batteries. That wouldn't add up to too many household-days if you started adding households. You'd have to do some fancy switching to not overwhelm the relatively few powerwalls in an outage.

          For context, based on our usage, we're at about three days of backup just for us [1]. It doesn't take much imagination to come up with a scenario where that happens where we are. We are the very last pole.

          [0] Which in turn requires connectivity. The cell network here is sparse, to put it mildly. We just got fiber, but that's hanging on the same poles as the power. The fail-safe mode for no grid and no connectivity is to not backfeed, but at that point, you're back where you are today.

          [1] It's the fridge, mostly.

          • mprovost 13 days ago
            You don't need internet for everything... by definition these systems are connected to the grid by wires which can transmit a limited signal. Utilities in countries such as NZ have managed load for decades by turning hot water heaters in customers' homes off and on remotely with ripple control units which transmit a signal over the power line itself. [0]


            • bobbob1921 13 days ago
              This is correct, an added benefit is if the power poles are actually down then the communication over the powerlines would be down, but that’s fine as power couldn’t be sent over the lines anyway in that case. What I wonder about/have concerns in regards to, is as someone else mentioned the relays required to turn on and off backfeeding in terms of electrical Utility workers having to de-energize a line to perform work. I’m not too sure that these workers can entrust their lives to the reliability of various (multiple) relays to not sporadically switch back on once the line has been confirmed as de-energized (in other words depending on the relay not to malfunction and re-energized while the workers are working upon the line )

              I assume there will be some types of tools to monitor / alarm the workers for exactly this scenario (ie something they can attach to the line while it is being worked on and if it becomes re-energized an alarm will sound).

              • Shog9 13 days ago
                Generally grid backfeed capability for these systems is predicated on the presence of a stable voltage and frequency already on the grid - this is both a safety measure and a practical necessity, as the inverter needs to synchronize with the grid frequency.
                • crznp 13 days ago
                  Just to add the terms: a "grid following" matches the grid's voltage/frequency, so it only works when connected to the grid. A "grid forming" system can set its voltage/frequency independently and can only work off-grid

                  It sounds like mauvehaus is talking about a hybrid system that has both grid following and grid forming capabilities: switching to grid-forming as it disconnects. But that means that you can't power your neighbors.

                • screcth 13 days ago
                  I wonder why it is that real world systems don't use a GPS receiver to obtain a 1 Hz signal that's synchronized to the GPS clock and then generate a local 50/60 Hz clock for grid synchronization.

                  I'm just an EE with no experience in real power systems, so this may be absurd to someone who knows how this stuff works.

                  • bsder 13 days ago
                    It doesn't help under normal operating conditions.

                    It takes roughly 16ms at the speed of light to cross the US. 16ms is also coincidentally one entire cycle at 60Hz. So, who should sync to GPS and at what phase?

                    In addition, when the grid begins to get overdrawn, the phase begins to slide due to the physical nature of generators.

                    Global sync doesn't really matter; local sync is what's important.

                  • burnished 13 days ago
                    I didn't take power systems either (EE that went CE), but even if your clock was accurate and the system was tuned wouldn't you still likely be out of phase? I don't remember enough about the whys (like, does the type of load modify the frequency? gone down the memory hole) but I remember a guest lecturer mentioning that the actual frequency frequently had small deviations from nominal, too
                    • throwup238 13 days ago
                      Yeah that's what the synchronization is for - the precise frequency doesn't matter except for consumer devices that use it as a clock signal. If the hot line is at 120V and you're a quarter cycle out of phase at 0V, all of that energy is just going to flow directly to your circuit and destroy something. Since the frequency represents the spinning of large turbines, a power plant going online out of phase is essentially slamming their turbines into a massive external force (things go boom).
                      • screcth 13 days ago
                        If you use GPS-based synchronization you have a reference of when zero crossings should occur. This does not mean that you have to output exactly that waveform.

                        Once you have a reference you can use a control algorithm. like a PID, to adjust the phase of the grid that you are generating locally.

                        If you sense that the grid is running out of phase from what your GPS clock says is the true reference you can increase/decrease your power output a bit to increase/decrease its frequency and catch up with the phase error.

                        It would be analogous to a type 2 PLL. The 1 Hz GPS clock would be the reference clock, the 60 Hz grid would be the "PLL output" and the VCO would be the power turbine, or AC inverter.

                        • burnished 13 days ago
                          Ok, potentially stupid question, but if you're already using the grid as a control signal what do you need the external clock for?
                          • screcth 13 days ago
                            From the comment I replied to:

                            > Generally grid backfeed capability for these systems is predicated on the presence of a stable voltage and frequency already on the grid.

                            You are using the reference clock to know precisely how misaligned the grid is and you use the reference to make small corrections to the generator power or inverter phase to restore phase alignment.

                            If you have an external stable reference clock (GPS) you know what the grid *phase* should be and it should be easier for network of multiple small inverters to keep the grid stable and to bring it up by themselves if there's a large blackout.

                            • burnished 12 days ago
                              Bless, thank you for helping me take that one full circle.
            • abdullahkhalids 13 days ago
              Back in the 90s, I interacted with a home intercom system that ran purely on the electric wires. Just plug in the devices into any electric wall sockets, and they just worked. Voice wasn't too bad either.
          • foobarian 13 days ago
            Question: when an actual power outage happens, does the powerwall backup take over immediately without loss of power to the household? Or if not how long does it take to switch the grid interlock? Just wondering if this would be suitable as a UPS for my computers.
            • mauvehaus 13 days ago
              It's entirely transparent. The app will notify after the power has been off for some amount of time to avoid nuisance notifications for flickers or brief outages.
    • cogman10 13 days ago
      > Electric cars mostly don't have the circuitry to do this yet

      Electric cars do have the circuitry to do this. It's built into the CCS protocol which nearly every EV speaks (US and EU).

      And it's not hard to imagine how this would work. Two big connectors on EVs can ultimately be hooked pretty much directly to the battery. When you plug in, you'll hear a giant "cachunck" which is the relay switching to DC charging.

      If you want to turn that into a vehicle providing power, the action is exactly the same, "cachunk" the relay which completes the circuit and have the charging station convert that DC power to AC.

      That's where the actual equipment gap exists. AFAIK, it's basically only ford that's producing a V2H charger. Enphase has one announced since last year, but I've not really seen anything about it since the announcement.

      • shiftpgdn 13 days ago
        Tesla has a V2H/V2G (Powershare is the name Tesla uses) available for Cybertruck owners included as part of the vehicle purchase. I am still waiting on mine, despite receiving my Cybertruck almost 6 months ago.

        I also have the Ford equipment from when I got my Lightning, however finding a competent installer has been impossible. Right now it's collecting spiders in the corner of my garage.

        • cogman10 13 days ago
          Do you have anyone that does home battery installations? It should be pretty much the exact same process to install one of these. It may be somewhat new tech, but ultimately it should look identical to battery controller install.
        • mensetmanusman 13 days ago
          Don’t forget about airplanes, you can fly in competent installers, some may enjoy the trip :)
        • tnel77 13 days ago
          A Lightning and a CyberTruck? Have either touched dirt? :)
          • shiftpgdn 13 days ago
            The Lightning has plenty of times since I am confident I could get it repaired or replaced easily. The Cybertruck... not so much.
            • tnel77 12 days ago
              Good! I had to ask. I’ve seen multiple Cybertrucks driving around my neighborhood and I’m pretty certain none of them will ever touch dirt. I hope you enjoy both vehicles!
      • soco 13 days ago
        In Switzerland it's also forbidden (as of now) to even think using your EV as external battery. The energy companies are fighting tooth and nail against people becoming independent from the(ir) grid.
      • neallindsay 13 days ago
        Is it in the CCS protocol? My understanding was that the F150 used a custom extension in order to do its thing where it will power your house. And even they don't feed back into the grid, which would require synchronizing the A/C.

        If there's already a protocol built into CCS for vehicle-to-grid then that's great news, but manufacturers would still need to enable it.

        There have been some test deployments of vehicle-to-grid in the US from Leafs since they use CHAdeMO.

        • cogman10 13 days ago
          > Is it in the CCS protocol?

          Yup [1]. It was an early proposal before CCS was ratified and has been there since the start.

          > My understanding was that the F150 used a custom extension in order to do its thing where it will power your house. And even they don't feed back into the grid, which would require synchronizing the A/C.

          Both the F150 and CT have onboard DC->AC inverters. This, IMO, is a bad move. It makes way more sense for the inverter to be a part of the house. However, the onboard inverter does make it easier to integrate with generator switches in panels.

          If I were to guess, the reason we've not seen a charger/generator combo has more to do with the complexity of battery voltage platforms. Unfortunately, CCS does not specify pack voltage which means any inverter would need to handle anywhere from 600->900V from the car.


          • toast0 13 days ago
            > Both the F150 and CT have onboard DC->AC inverters. This, IMO, is a bad move. It makes way more sense for the inverter to be a part of the house. However, the onboard inverter does make it easier to integrate with generator switches in panels.

            Having the inverter in the truck makes it useful for construction sites where you need electrical equipment and utility power isn't connected for whatever reason (a good one is construction hasn't gotten there yet; but also sometimes 240v tools are needed and there's no convenient access to 240v), and makes the truck a drop in replacement for a portable generator that people may already have.

            • cogman10 13 days ago
              Agreed, I should be clear that I think it's a bad idea for a general V2H solution.

              The problem with both of these is their capacity will be determined by the manufacturer. CT has a max output of like 3kW, which is unfortunately quiet low.

              These batteries on pretty much any EV sold (including phevs) should be capable of spitting out 15->50kW.

    • jeffbee 13 days ago
      I think it's appealing to nerds to believe that a long-tail disaggregated phenomenon like EV V2G is going to save the grid, but the apparent truth is that the grid will be saved by central planning and top-down command and control. We see this in California. We have a million EVs registered in the state, but our grid-scale, CPUC-regulated battery stations are still larger in terms of energy storage, and unlike the EV fleet the availability and responsiveness of battery energy storage stations is assured.

      Load shedding like that in the article is a vanilla feature of electric grids everywhere. Large consumers get discounted power in exchange for agreeing to disconnect during grid emergencies. Because the number of such customers is small, that feature is easy to regulate and you can ignore the game theoretic shenanigans that would emerge from widespread V2G with millions of individual participants.

      • hedora 13 days ago
        California is a bad example of "working power grids".

        Thanks to decades of graft, PG&E keeps blowing up / burning down neighborhoods / cities, and then passes the liability cost to consumers.

        At this point, we're paying $0.40 / kWh for grid power. National average is $0.08. Our solar + battery system is saving us something like $600 / month for a small, energy efficient house + one small, energy efficient EV.

        With the new rates, there's barely any financial reason to connect to the grid at all. You have to get solar + battery anyway, and getting a backup propane generator to keep the batteries from running empty is only an additional $10-20K.

        At that point, your house would be off-grid capable. Assuming a 10 year depreciation cycle on the generator, that's an additional $166 / month over the equipment's lifespan.

        Finally, assuming people move to battery + solar to avoid paying $500-600 per month ($60K over ten years, which is more than battery + solar cost, so everyone should do that right now), they'll have to raise the rates so that people that are almost never importing power are still paying enough to cover liability + grid maintenance.

        $166 / month / house isn't going to cover that. PG&E's screwed.

        Personally, I want to buy a direct air capture device that I can plug into my house. Whenever the batteries charge above about 80%, I want it to convert all surplus electricity into little solid carbon bricks, and dump the "waste" air (with low CO2 content) into my house. That + propane generator + close my PG&E account seems ideal.

        • jeffbee 13 days ago
          > At this point, we're paying $0.40 / kWh for grid power. National average is $0.08.

          It doesn't help to exaggerate, that only undermines whatever legitimate point is buried in here. California weighted residential cost was 33¢ according to latest EIA data and the national average was 17¢.

          > You have to get solar + battery anyway

          Since more than 99% of California homes lack such a system, it is obviously not necessary to own one.

        • burnished 13 days ago
          From a quick peak it looks like most of the gas to solid processes require a non-trivial amount of heat (ignoring turning it into dry ice cause it'd just boil away anyway), but I suppose your capture can be separated from your condensation and cooled separately.

          You ever look at what the throughout might be? I was just curious if a homelab setup of what you described was possible but that was as far as I got

      • applied_heat 13 days ago
        Tesla will come to the utility and say we can offer you 3000 MW of demand response in any granularity you want for $x/hr.

        $x is usually not very big since the modus operandi for the bulk electric system in North America has been everyone can use as much power as they want whenever they want, so we have had lots of supply.

        But a couple of times per year $x could be thousands of dollars per MW per hour, for a few hours. So not worth the effort really.

        Maybe Tesla will offer the demand response more geographically so distribution utilities can also use it for peak shaving on their substations. But that is pretty brave area to be operating in for a utility, depending on Tesla to avoid overloading transformers or risk a localized black out.

        There would have to be a serious penalty for bidding in the demand response but then not performing as well.

        I agree it does seem far fetched at this time.

        Edit: price signals to consumers is probably a better mechanism

      • szvsw 13 days ago
        > appealing to nerds to believe that a long-tail disaggregated phenomenon like EV V2G is going to save the grid, but the apparent truth is that the grid will be saved by central planning and top-down command and control.

        I think this sentiment applies to a lot of facets of the energy transition/preventing climate collapse - perhaps sadly?

        A lot of the cooler distributed solutions and demand-side solutions could have been our saviors had we started on them in earnest 10-15 years ago. They should still be pursued as part of a robust and resilient system (everything from re-insulating homes to V2G to cutting out dairy and meat from diets to dispatchable home heating/cooling-appliances etc etc) but the truth is that we are spiraling towards climate collapse and those sorts of solutions are simply put more difficult to implement and maintain, especially because they often involve massive social buy in; at the same time, those solutions don’t tend to address the source energy question in any meaningful way (besides rooftop PV which is not enough even for low energy homes in many climates), so supply-side solutions will always still be needed, and like you said, there are economies of scale in both thermodynamic/physics efficiencies and in management and planning. It would be awesome if everyone stopped eating steak, but the realist in me knows that expecting that to happen at a scale which moves the needle just does not seem realistic.

        It’s sad to me in that the “distributed” solutions tend to actually address root causes of climate catastrophe - addiction to energy consumption, “capitalism” writ large, being good denizens of the planet, etc etc in one way or another - and would entail a real improvement to how we understand the fabric of society and modern civilization (I’m speaking broadly here of course, but I think it holds water), whereas the centralized solutions feel in some sense like bandaids that obfuscate our collective responsibility and engagement with the issue (besides tax dollars and utility rates going towards the solutions I guess), but at the timescale of the crisis we are dealing with the size of that lever and its effectiveness is kind of unarguable I think.

        Build big solar and wind farms, lots of batteries, start using cleaner steel and cement manufacturing processes, hope that super deep geothermal or micro-reactors progress, hope that we find a better way to sequester carbon coming from either direct air capture or post-combustion carbon capture.

        Everything else is just (important) gravy. And I say this as someone mostly focused on the demand side, not supply side! (I’m a building scientist)

    • pfdietz 13 days ago
      Even lower hanging fruit would be increasing the dispatchability of demand in homes. Simply allowing heating and cooling to be dispatched would be very helpful. In the past this was done with bespoke add-ons to furnaces and AC compressors, but now it's done with smart thermostats.
      • smileysteve 13 days ago
        Dispatched versus random dither are similar;

        For most smart thermostats, one of the lesser known quality improvements is the 5 minute + dither time for startup. This is designed for compressor health - but prevents repeat cascading brownouts - where programmable thermostats or manual thermostats would have either a strict 5 minute wait time or none at all.

        Other features like time of use, humidity correction, and fan circulation also drastically reduce energy usage.

        • hedora 13 days ago
          Our thermostat seems to apply the delay even when you manually turn it on.

          It drives me nuts. (After a power failure or during scheduled transitions make sense; after I push the button: not so much.)

      • szvsw 13 days ago
        Heating and cooling would be massive. Another one is dishwashers (eh) but also laundry machines (washers and dryers/hot water heaters). Pushing those to random overnight spacing or preferably midday can help at least a bit.
      • immibis 13 days ago
        ISTR they did this somewhere (getting a discount on power in return for letting them decrease your cooling remotely at peak times) and the political right threw a national hissyfit.
        • pfdietz 13 days ago
          I signed up for this in Illinois and never heard a single peep of complaint about it (nor did I ever notice it when it kicked in.)
      • Workaccount2 13 days ago
        If you don't have a smart thermostat, check whoever your utility is for discounts. I got a 50% discount buying it through them.
        • op00to 13 days ago
          My utility keeps sending me google thermostats for free despite having ecobee. I keep giving them away but they keep sending them!
    • t_tsonev 13 days ago
      On-demand charging is very much happening today. All chargers communicate the available charging current with the vehicle and some can adjust it dynamically.

      As for control, there are proprietary and OSS solutions (shout out to that allow you to control the charging process based on price, grid load, local PV generation, etc.

      • ckdarby 13 days ago
        I see the basic version of this already with the Tesla vehicles where if you plug in another appliance on the same breaker the car will decrease its draw to prevent triggering the breaker.
        • cesarb 13 days ago
          > where if you plug in another appliance on the same breaker the car will decrease its draw to prevent triggering the breaker.

          That's alarming. There's no way the car can know that there's another appliance on the same breaker. The only way it can know it should decrease its power draw, is by detecting that the voltage decreased on the circuit, but that voltage decrease will only be noticeable if the wiring is undersized for the load (in fact, that feature is probably designed to protect against overheating undersized wiring on the circuit feeding the car). That circuit is probably severely overloaded, which can be a fire risk.

          • jeffbee 13 days ago
            This feature doesn't exist. The Tesla home charger can share current with other Tesla chargers on the same branch. It can't share current with an electric water heater or some other non-Tesla load
            • Retric 13 days ago
              I think the OP was just confusing a feature used by someone with a Tesla from a Tesla specific feature.

              They might be using one of these:

              There’s also smart breaker boxes that again have nothing to do with Tesla specifically. The breaker box will preferentially shed EV charging first when someone would otherwise trip their whole homes’s connection to the grid. It’s not a perfect solution, but sometimes your local utility doesn’t want to upgrade your connection to the grid because their substation is fully provisioned or whatnot.

    • rthomas6 13 days ago
      I wish it were standard practice to have some kind of house battery (what Tesla calls "Powerwall") for new construction homes. They could smooth out peak usage in this exact way, in addition to making individual homes more resilient to occasional power outages.
      • izacus 13 days ago
        So you want to make one of the most unaffordable assets a person can buy even more unaffordable by mandating a big extra investment to prop up underinvested infrastructure?

        Perhaps for building blocks (if US ever densifies away from single family homes) would make sense.

        • rthomas6 13 days ago
          The building materials aren't what's causing housing to be unaffordable. It's single family zoning and land speculation. The best thing to do to lower housing costs is to drastically reduce single family zoning (some cities have like 90% land area zoned for detached single family houses only!) to allow for infill development of SFHs into denser housing, and to implement an unimproved land value tax to discourage profiteering on land.
          • izacus 12 days ago
            Sure, but slapping a 20-30.000$ on top doesn't help at all.
        • burnished 13 days ago
          Its a drop in the bucket relatively speaking, more like insulation in terms of upfront cost to provide a benefit and reduce long term costs.
          • izacus 12 days ago
            It's... really not if you tried to actually build it.
      • adgjlsfhk1 13 days ago
        IMO this isn't the direction we're going. I'd bet that in 20 years, there will be some form of grid-scale specific battery that is 2-3x cheaper than lithium iron (per joule), but with much worse energy density. Many of the batteries of this type are unsuitable for home installation (because they get more efficient with scale). Some of the potential options for this are things like iron sulfur batteries which require pretty high temps, or some designs for batteries that basically involve physical pumps to pump liquid anodes and cathodes around.
        • szvsw 13 days ago
          > some designs for batteries that basically involve physical pumps to pump liquid anodes and cathodes around

          Yeah in addition to flow batteries there are also grid scale thermal batteries, though these are targeted more towards long duration energy storage: run electricity through a large carbon block in a shipping container form factor to heat it up, then use infrared photovoltaics to convert back to electricity when desired.

      • practicemaths 13 days ago
        It might happen! There's a fair bit of active research into repurposing batteries from cars into such a purpose or similar.

        This both extends the usefulness of the manufactured good (the battery) while also hopefully lowering the upfront costs of new electric cars.

      • ethagnawl 13 days ago
        This is the ideal. I think a happy medium -- which some companies are already working on -- is adding batteries to appliances. This allows them to charge when electricity is cheaper or provided locally (i.e. solar) and act as a buffer for outages.
      • dpkirchner 13 days ago
        I think I'd prefer having new large-ish developments (20+ homes?) install a single house-sized building filled with batteries so individual homeowners don't have to sacrifice space or deal with maintenance, replacement, insurance, etc.
        • shagie 13 days ago
          What if Your Town Doubled as a Private Power Grid? Around the country, developers are building microgrids — energy-resilient communities that act as their own energy source.

          > ...

          > The couple liked that each house came with a three-kilowatt rooftop solar system, which would reduce their carbon footprint and cut their utility bill by a third. But as they toured the 31-home community, they discovered that they were looking at North Carolina’s first residential “microgrid” development.

          > A microgrid is a network of buildings that essentially acts as a miniature power grid. It can operate outside of the larger municipal electrical system by ensuring backup power for the entire development, which can be produced by a solar array system and stored in a battery.

          > ...

          > At Heron’s Nest, the 62-kilowatt solar system and 255-kilowatt-hour battery are maintained by the local utility, Brunswick Electric Membership Corporation. Homeowners get a monthly energy credit on their utility bills by signing an interconnection agreement with the utility, which controls their hot-water heaters and thermostats when there’s high power demand. If the local power grid is stressed and the utility believes a power outage is necessary, residents here will be able to keep their lights on.

        • skybrian 13 days ago
          It depends on the situation. That approach doesn’t seem great for rural areas where outages are most likely.
    • rtkwe 13 days ago
      There are a number of industries that also do this where they'll be voluntarily shut off or pulled back during times where the grid nears brownout/blackout because their processes are interruptible with minimal cost which for battery charging stations is pretty true unless they're entire stock is at a very low charge.
    • adrianmonk 13 days ago
      I wouldn't want to participate during emergencies, though.

      If a hurricane or a winter storm is coming, I want my car at 100% in case I need to evacuate, drive around to get food, etc. On top of that, who knows when I'll be able to charge the car again.

      If we rely on cars to help stabilize the grid, it's not ideal that people will opt out just when it's needed the most.

    • jogjayr 13 days ago
      Wouldn't you want to reduce the number of charge/discharge cycles on your EV battery? Using it for household needs outside of an emergency seems wasteful.
      • moooo99 13 days ago
        A modern battery far outlives the rest of the vehicle when you look at the charging cycles. So you can afford to use some of it for home electricity storage. AFAIK some cells types easily last 3,000 cycles at >80% capacity. Assume 400km/charge you’d be looking at 1.2mil kilometers of driving just based on the battery.

        Also, there is a a natural limit to how much you’re using, considering you also want to use your vehicle for driving as well.

      • neallindsay 13 days ago
        Not every charge/discharge wears on the battery the same. Staying at 0% or 100% is worse than staying around the middle. Charging and discharging extremely quickly wears more than slow charges/discharges. And high temperatures are bad for battery health in general.

        Electric car batteries are tuned much more conservatively than phone batteries and are massively over-provisioned because US consumers are nervous about exactly the thing you're talking about. That's not too surprising because to a large degree the car IS the battery, but battery wear is generally very well mitigated (with the exception of the Nissan Leaf).

        • hedora 13 days ago
          Early Nissan Leafs. At this point, the newer used ones include battery health management, and should be similar to other used EVs.

          Note that getting an old one with a dead battery makes even less sense than you'd guess: It'll burn out any replacement battery just as fast, assuming you can actually get one.

      • MostlyStable 13 days ago
        If I was able to use my EV as an external battery for my home, I could potentially save ~$100 per month on my power bill, and that would represent <1/2 additional charge cycle per day. I'm pretty sure that would be worth it. This is going to depend on the power rates in ones particular area though.
  • robocop_legacy 13 days ago
    “ According to numbers provided by the company, 590 Gogoro battery-swap locations (some of which have more than one swap station) stopped drawing electricity from the grid, lowering local demand by a total six megawatts—enough to power thousands of homes.”

    From the title I thought the batteries would stabilize the grid by discharging back into it but apparently that capability is not implemented yet. Still interesting that the stations increase demand enough that grid stabilization can occur by simply disabling battery charging.

    • cesarb 13 days ago
      > Still interesting that the stations increase demand enough that grid stabilization can occur by simply disabling battery charging.

      That trick is not limited to battery stations. Another thing which is done in some places (at least my country does that), is to have some circuits from substations configured so that they cut power when the frequency drops too low and/or too fast, preventing a generalized blackout at the cost of temporarily losing power to parts of the system (the circuits to be powered down in each step are chosen so that important loads like hospitals do not lose power). In the country-wide blackout last year (, that mechanism stabilized power in the southern part of the country, allowing it to recover quickly (the northern part, where the disturbance originated, unfortunately lost too much transmission and generation, and needed a black start).

    • redleader55 13 days ago
      I'm bivalent on this topic. On one hand 6MW is very little, compared to 3,200 MW lost power during the earthquake, especially when spread over 580 stations. It's likely the feature to disconnect when frequency drops is designed to protect the station, rather than the grid. In general, this sounds like a feel-good story rather than a something real.

      That being said, I see this as a huge opportunity to stabilize the grid by having large-ish battery stations discharging their stored power when frequency drops. It's likely this will need to be grid controlled rather than individual and decentralized completely. It's hard (for me) to say what needs to be built first: an automatic way to reduce power draw from the grid by large consumers that are non-essential (eg. industry) or battery stations that can keep the frequency stable in case of brownouts.

      • p_l 13 days ago
        A huge part of previously installed grid scale battery power (like famous pilot Tesla plants in Australia) were precisely for frequency stability. Considerable portion of automatic cutoff or otherwise curtailment support systems in smart meters and the like is also related to grid stability contracts. Sometimes it's as little as "giving more time for steam generator to up the pressure" while wind/solar wanes, but it means there's less load on turbines which means there's less desynchronization which means less possibility of catastrophic failure due to unexpected flows in the grid.

        So, 6MW might be a little, and the comparison suggests it enabled more power for houses, but such little details can mean considerable differences in keeping grid frequency in sync and preventing further outage

      • ljf 13 days ago
        I spent a short while going down the rabbit whole of the meaning of bivalent, and how that applies to this scenario/phrasing - before I realised it was likely a typo for ambivalent.
      • lanthade 13 days ago
        I agree, more of a feel good/marketing piece. Bad electric coming into a system can cause damage so protection circuits are built in to prevent that damage from happening. I wonder how many UPSs kicked in and also lowered demand. What about poorly designed devices that stopped working?
    • datadrivenangel 13 days ago
      I wrote a book on the topic of virtual power plants, and it's amazing how supply and demand have to match exactly in an electrical system, so reducing power use is the same thing as increasing power use!
  • jillesvangurp 13 days ago
    There are multiple terawatt hours per year of committed battery factory investments for the next few years. Current yearly production is only about one twh. With production outstripping demand, we can look forward to prices coming down further. Committed of course does not mean these factories will get built. Quite a few of them will be a combination of redundant and obsolete before they are built.

    Bloomberg NEF is tracking close to 8 twh worth of battery manufacturing investments for 2025 with demand being estimated at only 1.6 twh. There's a battery bubble happening and this is good news.

    Also, most of these batteries are just sitting there are relatively high charge states (70-80%) not being utilized at all. The point here is that we won't have a battery shortage and a growing amount of potential energy locked up in charged batteries. Keeping these batteries charged is also not that big of a challenge because we're mostly not even discharging them. Cars on average only drive around 14K miles per year. That's only about 3.5 mwh of energy needed per year (at 4 mile/kwh). Or about 59 charging cycles for a 60kwh battery.

    Batteries also last quite long. So we have this builtup of tens and soon hundreds of twh of battery just sitting there in cars, trucks, grid storage, etc.

    So with all that happening in the background, swappable batteries are a bit redundant. Also when considering their potential to grid stabilization.

    • burnished 13 days ago
      Do you have an estimate for their stable life? I don't have an intuition for it and am curious
      • jillesvangurp 13 days ago
        15-20 years depending on usage. Mostly, EV manufacturers are very comfortable offering 8 years or 100K miles on their batteries (whichever comes first). Meaning they don't anticipate failures to be a regular thing before that; or long after that. Most batteries should last a lot longer.

        Most EVs produced ever (i.e. in the last 15 years) are still driving with their original batteries. So that should build some confidence. Some early cars with relatively primitive battery management obviously are getting their batteries replaced. But even then, it's not going to break the bank and extends the lifetime of the vehicle by a decade+. A 4K battery replacement for a 12 year old Nissan Leaf to double its range doesn't sound like a horrible deal, for example. Replacing a 20kwh battery with a 40kwh one for that money is a good deal and extends the life of the vehicle by about 10+ years.

        Of course it all depends on battery chemistries and battery management. There has been a lot of improvement on both dimensions. Mostly I don't think there will be a lot of surprises on this front as this is a well studied thing.

        • nkingsy 13 days ago
          4k for a leaf battery? Last I checked it was way more than that.
  • ggm 13 days ago
    Demand management is fantastic, batteries are just amazing and can sink power, provide power, manage reactive power and frequency and do it in milliseconds.
    • gorbypark 13 days ago
      Once upon a time (~15 years ago!), I worked at a ski resort in New Zealand that had a fancy new "automated" snowmaking system. When all 200 (give or take) snow guns were running, along with the required water pumps/air compressors, the system could draw megawatts of power. Generally it didn't get below zero very often in the closest city to the resort (Queenstown), but when it did, people would turn on their electric heaters, and demand would go through the roof. The way it was explained to me is that the power company actually sent a signal through the power supply which would trigger our snowmaking system to shut itself off. It was pretty cool to see and also infuriating, because of course you can make much more snow when it's cold.

      Being a giant nerd I of course asked if this was really happening via some network connection, but I was told they actually modulated the frequency of the entire grid in a very small, but predictable way. Some sort of device sat there and when it picked up the pattern in the frequency changes it would then trigger the shutdown.

      I've spent a while doing some searches about such a system but couldn't find any information. Anyways, your comment about demand management made me think of this and how it's been possible in one way or another for a long time (albeit not milliseconds!).

      • flakeoil 13 days ago
        Are you sure they modulated the frequency of the actual AC power (typically 50 or 60 Hz) and didn't overlay it with a high frequency signal? It's very common with power line communication or power line modems. You modulate a data signal on top.
      • shagie 13 days ago
        > The way it was explained to me is that the power company actually sent a signal through the power supply which would trigger our snowmaking system to shut itself off.

        > Modern alternating-current grids use precise frequency control as an out-of-band signal to coordinate generators connected the network. The practice arose because the frequency of a mechanical generator varies with the input force and output load experienced. Excess load withdraws rotational energy from the generator shaft, reducing the frequency of the generated current; excess force deposits rotational energy, increasing frequency. Automatic generation control (AGC) maintains scheduled frequency and interchange power flows by adjusting the generator governor to counteract frequency changes, typically within several decaseconds.

        If there's more demand on the grid, the generators have to work harder (slower) to generate the power and this shows up as a drop in the frequency of the grid. In New Zealand, that is 50 Hz.

        If the frequency drops a bit, it is then an indication for generators to kick on or load to be shed. If the frequency goes above the 50 Hz frequency, it becomes a sign that the price that power is being bought at is lower and so more expensive / dirty power generation should be throttled back or shut down (and batteries should recharge rather than discharge and pumped storage reservoirs should pump rather than generate )

        On the other side of it, you can see the frequency changes back with the Texas blackout -

        Those changes are less than 1% of the frequency, but that's all that's needed to signal information between the generators and demand.

  • oliwarner 13 days ago
    Not the story I thought it was going to be. They turned off the chargers, releasing enough capacity for the grid to stabilise. Big woop. This is basic grid planning. Non critical circuits should cut out to stabilise the grid by design.

    With this sort of praise, I thought they'd returned power to the grid when it slumped.

    • amelius 13 days ago
      They could have momentarily raised the energy prices and a similar thing would have happened.
      • oliwarner 11 days ago
        You can't guarantee industrial customers won't just increase their sales prices with supply. Consider normal EV chargers: They just spike the price for the user. And depending on the market, wholesale prices might be locked in for 24h or more so industry knows its day-to-day costs before it incurs them. Smashing up the prices for tomorrow can generate usage spikes today. Good controls for weather-based supply but not a feature in a national emergency.

        Above all of this you still need to be able to prioritise physical circuits for healthcare, security, groceries and homes, and be able to cut off others.

        That said, Gogoro seem to be more about overselling contracts so spiking price probably would make shut down themselves.

  • EGreg 13 days ago
    Can't we just charge our electric cars during the DAY, when they are PARKED, and therefore THOSE batteries can be used? Seems like a good use of electricity. Sell it back to the grid during the night so electric cars can charge during the day. Arbitrage the fact that people with panels sell cheaply during the day.

    It seems to me that anyone could start an electric car parking garage, that

      1. receives excess electricity from solar panels around the city and suburbs during the day
      2. cheaply charges the electric cars during the day
      3. subsidizes powering the homes at night, for those same credits
    Or probably a better entity to do that would be the public subway system

    Storing electronic credits instead of energy hehe

    • thesis 13 days ago
      Sell it back at night when there's already a surplus of electricity? This is exactly why they want EV's charging at night, because there's less demand.
      • toast0 13 days ago
        In a lot of places, peak supply is during the day now. Overnight conditions are more about a lack of demand than a surplus of supply (although they're sort of equivalent), but really the most effective use would be charge while the sun shines, dump the battery onto the grid when you get home, and then charge slowly overnight after the end of peak. Assuming there's no downsides from cycling your battery and you and the system have perfect knowledge of your evening plans.
      • rtkwe 13 days ago
        Depending on the renewables mix the grid settles on, selling some back during the night might make sense. Both solar (obviously) and wind (less obviously, at least to me) decrease during the night so we may find we need to supplement the generation with battery reserves during the night for demand peaks.

        Nuclear and traditional dammed hydro prefer to be setup to provide extremely consistent power too so there's a gap in the power generation if we completely eliminate fossil fuels. Even if we don't completely eliminate them, peaker plants are extremely expensive to operate because their base costs and ideally low usage means their cost per megawatt is pretty high.

      • EGreg 13 days ago
        I there is a surplus of electricity at night, then why have batteries? There seems to be no problem at all with just spot market demand in that case!

        I thought the issue was that solar panels are generating electricity during the day only. So as someone with solar, you want to sell during the day and buy at night. No?

        My point is, rather than using batteries to store huge amounts of energy and lose much of it in the process, find efficiencies and use credits instead.

      • jandrese 13 days ago
        Typically the sell back time is in the evening, after solar production is done but before people go to bed. Also some in the early morning, albeit to a smaller degree.
  • ck2 13 days ago
    Vaguely related reminder for the USA to buy your large lifepo4 battery backups and solar panels sometime soon in June before the tariffs kick in and double prices next month.
  • kkfx 13 days ago
    Well... the long story short is that a smart-grid works better than a classic one. This example is a very limited kind of smartness, just "stop what you can stop on a strained grid to keep it up", for certain kind of loads in certain countries is a mandatory things: in France IVRE norms for domestic EV charging stations demand a device connected to the utility meter data port (2-wires system) to allow the grid stop (and potentially in the future just regulating the maximum allowed amps) a charge if needed, in some other countries I've read it's common to have connected thermostats the grid can tweak to lower heating/condition when there is too much energy demand.

    The issue of a smart grid though it's the fact that we need software from various actors, talking together, to regulate a DELICATE and CRITICAL infra. Just imaging if in a nation-wide smart-grid a bug or an attack tell all connected cars/chargers "please charge at maximum amps, we have a renewable production spike, we need load to lower the frequency" while the grid is definitively NOT in such situation or simply the potential cascading effects of some smart-grid demand that can't be immediately applied by all smart-connected customers, creating something we can see also in domestic p.v. with ac-coupled classic string inverters, where the battery inverter demand more energy to the p.v. inverter, while a small cloud reduce the p.v. output, the cloud go, the p.v. suddenly inject much more and the battery inverter flip "reduce, reduce, reduce", another cloud, another spike.

    IMVHO there is so far no tech safe enough for a PUBLIC smart-grid, we can at maximum have smart-grid-homes meaning local batteries (perhaps small + the EVs in V2H bi-directional direct DC-to-DC mode, since most cars use 400V batteries as most p.v. hybrid inverters on sale, there is no damn reason for a double DC-AC-DC conversion like Tesla PowerWall and others do) and local p.v. keeping a separate home grid, with the national grid directly powering only the main inverter or islandic device. The local system can just keep watching grid frequency shifting regulating the eventual p.v. surplus injection only if the grid frequency is a bit low, and using batteries to a certain SOC level if there is no p.v. and grid frequency is low and tend to trend lower in a certain timeframe.

    It's not smart, but at least it can't destabilize the grid much, an eventual attacker must been able to poison frequency shifting signals witch typically means having last mile physical access on the grid, something doable locally, but hard to do at a nation-wide or at least wide-enough areas to destabilize the grid.