Battery-swap networks are preventing emergency blackouts
technologyreview.com
technologyreview.com
As for control, there are proprietary and OSS solutions (shout out to https://evcc.io/en/) that allow you to control the charging process based on price, grid load, local PV generation, etc.
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.
They might be using one of these: https://www.chargedfuture.com/240v-outlet-smart-splitter/
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.
This both extends the usefulness of the manufactured good (the battery) while also hopefully lowering the upfront costs of new electric cars.
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.
I'm also curious how it compares in cost effectiveness to this battery that's based on sand.
https://tambaynews.blogspot.com/2022/07/worlds-first-commerc...
In addition to a few hundred mil in funding, they have support from NREL/DoE, eg https://www.nrel.gov/news/features/2024/how-nrel-helps-start... I also was at a conference recently that had some cool feasibility studies done by people at NREL about using TPV for LDTES with Antora product specs used for boundary conditions, but I don’t think that’s published yet.
I’ve also seen FourthPower mentioned a few times but I don’t know much about them or how feasible their tech actually is.
https://gofourth.com/our-technology/
Energy storage is not my area of expertise (but very much adjacent to it).
https://www.nytimes.com/2023/08/07/realestate/microgrid-sola...
> ...
> 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.
Perhaps for building blocks (if US ever densifies away from single family homes) would make sense.
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.
I’m also curious if your neighbors benefit from your battery during outages.
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.
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).
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.
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.
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.
> 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.
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.
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:
https://greenmountainpower.com/news/gmps-request-to-expand-c...
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.
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.
It drives me nuts. (After a power failure or during scheduled transitions make sense; after I push the button: not so much.)
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.
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.
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.
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.
[1] https://www.greencarcongress.com/2009/09/emobility-harmonize...
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.
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.
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.
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).
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.
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.
$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
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)
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.
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
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.
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.
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.
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 (https://en.wikipedia.org/wiki/2023_Brazil_blackout), 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).
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.
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
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. https://about.bnef.com/blog/china-already-makes-as-many-batt...
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.
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.
These guys cite between 3 and 5K for the 24kwh battery. The 40kwh battery is a bit more expensive.
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!).
https://en.wikipedia.org/wiki/Zellweger_off-peak
In use for 7 decades or so.
https://en.wikipedia.org/wiki/Utility_frequency#Frequency_an...
> 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. https://www.emi.ea.govt.nz/Wholesale/Datasets/FrequencyKeepi...
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 https://www.energy-storage.news/new-zealand-considers-invest... )
On the other side of it, you can see the frequency changes back with the Texas blackout - https://www.houstonpublicmedia.org/articles/news/energy-envi...
Those changes are less than 1% of the frequency, but that's all that's needed to signal information between the generators and demand.
With this sort of praise, I thought they'd returned power to the grid when it slumped.
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.
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
Storing electronic credits instead of energy hehe
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.
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.
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.