Hacker News new | past | comments | ask | show | jobs | submit login
The Bidirectional Battery Edition (whyisthisinteresting.substack.com)
53 points by RickJWagner 6 days ago | hide | past | favorite | 55 comments

Fully charged recently did a video on the city of Utrecht in the Netherlands where I used to live: https://www.youtube.com/watch?v=L_BYDKz3_Jg. They are implementing a plan there with vehicle to grid that basically boils down to having enough plugged in capacity at any time to be able to run the entire region for a while. The city of Utrecht has a few hundred thousand people and the wider region has closer to a million people.

Many people have the wrong ideas about the intentions of batteries on the grid. They are mostly used to deal with short term dips and peaks in demand and supply of power. A few thousand cars can provide a lot of power. Long term, we are talking millions of cars.

For reference, the Netherlands has about 8M cars on the road and a production capacity of about 35GW. If millions of those cars were plugged in they could easily keep the Netherlands powered for a few days even if all that production capacity was wiped out. That is of course not the goal and it would not be a good plan. But it's interesting that it could work.

A more likely scenario would be that those millions of cars would be used to deal with dips and peaks to the extent of a few hundred MW. Instead of switching on a gas peaker plant of e.g. 1GW for an hour or so, you'd instead end up draining less than a KWH per car over the course of an hour. The point of having that much capacity is not using all of it but being able to spread the load such that the load on each car is minimal. That minimizes the wear and tear on the battery and makes it easier for power companies to deal with even substantial demand peaks.

Additionally you can match local demand with local supply and not deal with efficiency losses associated with getting power from elsewhere; or worse importing it from abroad over long distance cables.

> Ford's new F-150 Lightning electric pickup truck and Volkswagen's ID.4 are examples arriving with this feature

This isn't actually the case. What both manufacturers have announced is extra hardware to install at home that connects to the car battery directly over DC and has a grid-tied inverter. There's nothing special in the cars themselves, the same solution can work in any car with the same charging connector. The F150 allows providing power directly from the car but not to the grid, only to use directly while camping or something like that.

Actual bidirectional chargers in cars would be a much bigger deal as it would allow feeding energy back into the grid in many more situations. Without that this will be a niche solution that won't have grid scale impacts.

The F150 and cars like Ioniq 5 and EV6 do have built-in inverters (as I understand it's just designing the OBC inverter used to charge the battery from AC so it can run both ways). It's only capable of around 1.8kW. But that's enough to facilitate what's mentioned in the article. What's lacking then is some communication protocols to enable this. Might need some additional hardware too since there's extra challenges when feeding the grid I assume (has to match the frequency and phase I guess)?

I covered those inverters. You can get power directly from the car to run appliances but to connect it to the grid they just bypass that entirely and have you install an offboard inverter in the house.

Ford points out that, as part of the company’s forthcoming Intelligent Backup Power system, the feature is enough to power a home for up to three days in a power outage. Taking full advantage of that will require an 80-amp Ford Charge Station Pro, a home energy management system, and an inverter, which Ford plans to bundle together. But to output that power for EVs, appliances, or camping, you only need Pro Power Onboard.


It would be nice if they allowed their onboard inverters to do vehicle to grid. Even if their power is not enough to power a whole house it's more than enough for grid services and for emergency use at home. But so far no one has done that.

Sounds like you know the area.

I've been dreaming about the day we can have solar on the roof, charging and storing that energy in the car, and releasing that energy to the home via V2H when needed.

I looked into this about 6 months ago and options seemed very scant due to technical limitations or manufacturer hesitation.

Are you aware of anything new or actually working in this arena?

Question open to all, obviously. Thanks.

I only follow this out of curiosity and because I do have an EV and solar and am wondering when/if a battery will make sense. I have yet to see any solution to power the house from the car unfortunately. These announcements by Ford and VW are the most recent ones I've seen. One option that is out there and covers a part of the use case is a generic EV home charger that can be configured to modulate it's load based on the availability of solar:


The same company also has some smart electric water heaters that also allow optimizing how much of the solar gets used. How each solution helps depends massively on what your situation is (e.g., on or off grid) and the specifics in your geography (e.g., year long vs 15 minute net metering). Here we have 15-minute only net metering so I have a script that runs every 15 seconds, figures out if I'm giving away power to or drawing from the power company and turns on and off loads based on that (water heater and radiant flooring). If you have a battery somewhere in the system it becomes an easier tradeoff as you just need to react to the state of charge which changes much more slowly. All this is still pretty much in flux. With LFP batteries becoming very cheap and ideal for home solutions I expect Enphase or some other company to offer a price competitive full solution soon.

Interesting, thanks. That's still a step forward from where I thought we were. We'll get there.

I am reading The Windup Girl by Paolo Bacigalupi. In it Joules are used as currency, and they can easily be traded between such energy storage devices, which can themselves be used to power machinery. The nutritional value of food can also be so-measured.

Oo sounds fun. Thanks for the rec, I'll try it.

The "distributed battery for the grid" seems potentially significant. The "cars charging other cars" seems unneeded except in rare emergency circumstances.

I think they're one and the same. Imagine every parking lot having what is effectively a bus, switch, or router connecting all the cars together, and to the grid, at the same time. Keep a net account on the vehicle level, maybe with some smarts about planned trip distances.

That's a fun idea, but it has extreme complexity (both electrical and the human elements like accounting and permissions) and I don't see much benefit to anyone.

Maybe I'm missing something. Why would I want to plug in my car at the parking lot to have it discharged into somebody else's? At the parking lot, I just want a charger.

> Maybe I'm missing something. Why would I want to plug in my car at the parking lot to have it discharged into somebody else's? At the parking lot, I just want a charger.

If I've got easily enough range to get back, why wouldn't I want to sell some of that electricity at a higher price?

Being connected to the grid is more important though.

And on the flip side, I may want that so that my car is able to get the charge it needs in more places.

So to begin with let me be clear that I think using car batteries to time-shift load on the power grid makes some sense (mostly for cars parked at homes). However, it sounds like you're talking about using car batteries for solving distribution over geographical space rather than time, which I don't think makes sense at all.

> Why wouldn't I want to sell some of that electricity at a higher price

1) Charging, driving, and charging is never going to be economically competitive with transferring power through transmission lines.

2) Even if you could somehow drive from an area that has chargers connected to power plants to an area that ... has a decentralized off grid charging thing (?), and there's enough density of customers there in this electricity-less place, you can at most sell something like $2 per hour and a reasonable profit is going to be in the tens of cents per hour, max. Then you have to cover expenses: wear and tear on your $20,000 battery, on the rest of your car, opportunity cost of sitting there giving away power, and most importantly the planning and risk involved in driving afterwards with less in the tank.

Well, you were explicitly asking about that scenario so that's why I addressed it.

> 1) Charging, driving, and charging is never going to be economically competitive with transferring power through transmission lines.

I can charge at 5p/kWh during the night but during the day would be buying at ~25p/kWh as a regular customer at home - buying from a charger on the road would be more expensive (if nothing else, the VAT situation is different I think).

Is it not reasonable to think that it may be more economic to transfer power between two cars near each other than across the country? That's what would be happening if I'm putting back on the grid while someone else is charging their car somewhere.

> . Then you have to cover expenses: wear and tear on your $20,000 battery, on the rest of your car, opportunity cost of sitting there giving away power, and most importantly the planning and risk involved in driving afterwards with less in the tank.

I think you're picturing explicitly driving to a car park to sell excess power. That's not what I'm talking about. I'm saying that going to the shops/work and plugging in my car, it may easily make sense for my car to sell some of that electricity to someone else as I'm not going to need it today and can charge up at night cheaply. There's no wear & tear on my car, and no opportunity cost about sitting there.

Wear & tear on the battery is identical to any other load shifting, home battery or car, so if it doesn't make sense to do in the car park it wouldn't make sense to do at any other time.

> and most importantly the planning and risk involved in driving afterwards with less in the tank.

That's what the commenter was talking about with some smarts on trip planning. Frankly for most people long trips are rare enough that a default "leave me with X in the tank unless told otherwise" would deal with almost everything.

I think you're thinking too large scale. Think of replacing a peaker plant that might only be in operation for an hour by having a city full of cars each supply 100w for that hour. It's 1/800th of a tesla battery, but it replaces $500 worth of peaking plant usage.

Yep that makes sense. That's one of the situations I was referring to when I said:

> let me be clear that I think using car batteries to time-shift load on the power grid makes some sense

Note of course that you could probably achieve something similar by just having teslas be smart enough to stop charging during peak times if the owner OKs it instead of actually discharging full ones.

If all car manufacturers can offer bi-directional charging as default, without an increase in vehicle price, the potential offered by vehicle-to-grid charging is massive. The grid can be more easily balanced and consumers will have the potential to generate revenue from the EVs.

A UK trial involving 330 Nissan Leafs (https://www.ofgem.gov.uk/publications/case-study-uk-electric...) saw some participants earn up to £750/year, with grid savings potentially being £3.5 billion. The downside of the trial was the need for a dedicated, expensive bi-directional charger, highlighting the need for the functionality to be integrated into the car.

The downside of the trial was the need for a dedicated, expensive bi-directional charger, highlighting the need for the functionality to be integrated into the car.

I have been heavily involved in that project (writing firmware for the charger). I also have one on the side of my house connected to my car. I don't personally see any particular need for it to be integrated into the car. The main issue is cost, but that cost doesn't go away if it is in the car instead of on the wall. Either way, the hardware is there and the house wiring will have to support it. It is also important to keep in mind that the charger is essentially a prototype. The main cost is the bidirectional power module. Future versions will likely be considerably cheaper as manufacturing matures and economies of scale kick in.

EDIT: From a regulatory point of view, it is probably easier to have the technology in the charger as the grid operators (Distribution Network Operator - DNO) do not like having unknown generators being connected to their grid. You have to notify the DNO (and in some circumstances, get advanced permission) when you install solar or indeed, a V2G.

And if it's on the wall it will not be automatically scrapped when the car is scrapped as well as allowing you to easily switch vehicles and maintain that functionality.

That's interesting about the DNO. When notifying the DNO about the V2G installations does the property then become a generator?

Re. economies of scale I'd argue that an auto OEM can achieve better economies than a charger manufacturer

Well, the V2G is considered to be like a generator. You are generally allowed to install one device of up to 3.68kW on a install and notify basis. This is the same rules that would apply for a solar installation or whatever. If you want more than one device or you want to exceed 3.68kW, you will need permission. There are a whole series of increasingly complex rules after that.

The size of batteries in cars is huge too.

I was looking at a home battery, and the tesla powerwall is one of the better ones for a relatively high power output (assuming you can get one somewhere). That's around £10k+ for 13.5kWh, while a volkswagen ID3 has a 58kWh battery and costs £33k. That's less per kWh, and it's also a brand new car. It'd power my entire house for days, and is capable of charging drastically faster (so my home power supply would easily be the limiting factor) and I assume capable of discharging faster too (?).

There's an opportunity for a company that signs contracts with EV owners to use their car batteries when they are parked for distributed grid balancing and in exchange they replace these batteries after they drop to let's say 80% capacity.

That company can then recycle the used batteries or use them in stationary battery banks for a few more years before recycling them. I wonder how it pans out financially, but it seems it could work.

This is what we're doing with Jedlix (I'm the Technical Director there). Currently we're doing grid balancing and load optimizations so you as EV driver make money while you charge (!) and at the same time you help to keep the electricity grid stable.

Of course that works well with times that there is over capacity on the grid and there is a lot of flexibility in the car (time plugged in vs charge demand).

When there is a lot of production (lots of solar for example) and not a lot of demand (during the daytime) that energy can go nowhere. At night there is usually a lot of demand but less renewable power production (it's dark after all). This is one of the areas where we will leverage vehicle-to-grid to power homes using renewable energy and also allows to more efficiently use the solar power you yourself produce (if you have a PV set up).

Have a look at https://jedlix.com if you have an EV.

Nice, but AFAIU you only shift the charging time, you don't discharge car batteries to balance the grid, right?

That's still great, but it won't let us use solar power at night.

Currently we shift only indeed on the public part of the platform. Discharging is in development with our partners (EV manufacturers)

This actually sounds like a brilliant idea.

A shame everyone is busy.

> a dedicated, expensive bi-directional charger

The electronics to do this typically cost no more than a regular charger. In fact, with a firmware upgrade, many cars and chargers could do bidirectionality already.

The main cost so far is that few car manufacturers are prepared to enable bidirectional functionality unless they get a substantial cut of the profits of doing so. It's a business decision not to enable it.

You appear to be talking about alternating current bidirectional charging. There is no standard for that (and none in development that I am aware of). Also, I am not aware of any vehicle that has the onboard inverter that would be necessary (except possibly, the F150).

Existing bidirectional chargers are direct current chargers. Basically, they connect directly to the car's battery and communicate with the car's battery management system. This is the same way that rapid chargers work. The main cost is the bi-directional inverter/rectifier, but that cost will come down.

Correct - there is no standard for AC bidirectional charging. But someone could write one, and lots of existing cars could support it with a firmware update. Cars onboard chargers tend to use synchronous rectification[1], which means the power can flow the other way if software wants it to.

Most DC rapid chargers could likewise support it, although again it would require new firmware to be written for the car (to not abort when it sees current flowing outwards), and new firmware for the charger.

If you disassemble a car or charger, the thickest traces are the power path. If there are no diodes in the power path, then its likely the charger uses synchronous rectification and could support vehicle to grid with just firmware changes.

[1]: https://en.wikipedia.org/wiki/Active_rectification

I read somewhere that the inverters in most newly-produced Teslas are capable of doing this already, from a hardware perspective.

Might have been one of Munroe’s teardowns. No official word on why it isn’t supported yet.

It is my understanding that for "sane" charger topologies (ie. without tricks that share power components between charger and drivetrain) of the power level required for EV charger the SMPS comes out capable of bidirectional operation somewhat by default because active rectifier and inverter is (topology-wise) the same thing (and the symmetry is quite obvious in typical implementation).

Yesss, yesssss. We’ve largely “solved” capture. When batteries become efficient enough, storage and transmission will begin to collapse into the same. Let’s hope we make it to a world where self driving trucks are on a constant loop, delivering shed-sized batteries from massive solar farms to power neighborhood-sized grids, and only the last mile is traditional transmission.

Sounds like you you are envisioning dedicated trucks that just shuttle around batteries night and day. Why not just leave the batteries on either end (on parked trucks, for all I care) and put up a traditional high voltage transmission line?

Much like the "bandwidth of a container full of DLT tapes", it would be entertaining to calculate the amperage of a truck full of 18650 cells driving down the highway. It will probably turn out to be comically small compared to a wire.

>Researchers have found continual bidirectional use of a battery wears it down faster than normal, too.

Why? It seems to me it could be a few causes

1) Difficulty keeping the battery cool while stationary.

2) Higher loads than driving (really, though, how?!?)

3) Just using the battery for more full discharge cycles.

I strongly suspect durability (3) is the issue. It'll always be a tradeoff, and the car makers can adjust it.

I would like to see a citation for that. The Scurius trial[1] that another commenter mentioned found no evidence for increased battery degradation. Of course, it is going to depend on a lot of details on how the battery is managed, both for driving and for V2G. Generally, the worst thing you can do is fully charge the battery and then leave it sitting for long periods of time. V2G generally won't do that, but then neither will smart charging, generally. If you take the mentality of always topping up as soon as you get home, however, you are probably going to wear the battery out considerably faster than you need to.

[1] https://www.cenex.co.uk/projects-case-studies/sciurus/

It is 3). If you consider how many total miles the vehicle will drive before the battery is unusable, you can figure out how many fewer miles you'll be able to drive for each dollar earned from vehicle to grid charging.

Today that number varies widely, especially in grids with lots of renewables and highly variable prices.

What would be the liability implications if one person was charging their car from another car and for whatever reason broke their car (faulty/damaged charging port? Bad wiring? Something else?)

I’m guessing insurance would need to cover it. Would they want to cover it, or just make it another exclusion?

Today, using a Jerry can would be similar for emergency fuel but very unusual/rare. Plus a lot less risk of damage (still some risk of wrong fuel or sediments in the can).

I can definitely see the usefulness in smart homes with onsite batteries, but not quite convinced about the car aspect. Especially if it means having to replace your battery sooner due to more wear and tear.

> Plus a lot less risk of damage (still some risk of wrong fuel or sediments in the can).

Risk of electrical damage through the cars charging port is substantially lower IMO than risk of damage from misfuelling or dirt in fuel.

Electrical charging ports are very robust.

Probably similar to jump-starting?

The vast majority of the benefits for the grid can be achieved in a one way fashion (just delay or bring forward charging in response to grid signals and/or predictions) and that works for things without electrical batteries too (anything that heats, cools or pumps uphill).

But still, super handy to have mobile electricity generators that we often use ICE vehicles or custom built items for, move to a cleaner, better, cheaper, multipurpose solution.

Is anyone actually genuinely worried about charging? It's hard to separate it out from all the people worrying about it in the same way they worried about wind turbines affecting birds a.k.a. concern trolling.

I think people are genuinely worried about charging simply because it's unfamiliar, or they're familiar enough with their phone running out. And lots of people live in inconvenient home charging situations.

Car-to-car charging is a new idea to me. Of course, that depletes the range of the charging car, so where does that car get its charge from? I'm imagining a chain of charges like https://en.wikipedia.org/wiki/Operation_Black_Buck

More realistically it would be Uber Recharge: if you're stranded you pay some money to someone who comes along and gives you 50km of charge to get you un-stranded.

This is a slightly different use case (charging for people who don't have off-street parking), but I suspect that it will be a good model for out-of-charge rescues as well. It is essentially a mobile rapid charger:


I think there was a HN backed startup that did this with gas for cars parked at workplaces.

The win was abitraging the price of petrol and the drivers time. If these mobile batteries charge at non-peak periods they can probably replicate that.

A Tesla with the largest battery pack is 100kWh. UK electricity prices seem to be about 15p/kWh, so that's £15 of electricity maximum. Not a huge amount of room for margin, but I suppose it's in the same range as Uber trips or food deliveries, and small compared to the £50-£100 of a full tank of petrol.

Price from https://powercompare.co.uk/electricity-prices/

The service in the post I replied to charges £6 per week for up to 10Kwh, so best case 60pKWH, then 37p/kWh for each kWh beyond 10.

Which seems reasonable considering fast chargers at gas stations are about 45p/KwH.

I'm confused as to why they're inventing a new term to describe what you can already do with regular batteries. Heck even with permanently installed batteries, just plug a nintendo switch into your macbook over USB-C and guess which one will charge.

More likely we'll have small battery packs based on this technology to charge at home, store in the trunk and use on very long range journeys. What we're using small gas cans for now.

Refueling someone else in an emergency won't be a common use case by definition.

Selling the extra capacity at the corner of the street not common because it's not our job and we don't have time for that.

Maybe if every parking place has its own individual charging station people will plug in their cars in output mode, set how much they are selling and the network will buy it. This assumes that energy at home is cheaper than the price paid by the network.

Joules are the new bits.

If every vehicle could charge and return to the grid, the grid would need major improvements or else it will fail. The electric grid was not designed to have distributed power generation at a scale like this.

Would it? Intuitively it seems to me like distributed power generation and consumption would, in general, put less stress on the grid, as on average things will even out locally and you will not need to transfer large amount of energy over long distances. I realize my intuitive understanding could be completely wrong, and I would love to be explained how so.

As an example: My father (living in Germany) installed a 10 kWp photovoltaic system in spring last year along with a 10 kWh battery. In the summer months he was almost self-sufficient with that, as he could use his own solar power at night or on rainy days. The forecast for the whole year is ca. 65% self-sufficiency.

So local storage reduces the amount of both the power taken from and provided to the grid and therefore is beneficial for grid stability.

The problems start, when the storage is not local anymore, that is when the power surplus is not at the same location as the storage itself. A few examples: When the sun is shining your electric car is probably at the parking lot of your workplace and you can't store the output of your PV system. Or when it's a rainy day, you and also all other households in a wide area around you, wont' produce much power. Then the power has to come from somewhere else, e.g. a sunnier place, offshore wind or gas power plants.

So to use all the distributed storage effectively, you have to route the power around a lot, which probably makes changes in the power grid inevitable. How much one has to invest in the grid and how that relates to the overall power price is probably a very complex topic depending on a lot of variables including geography, power mix and cost of storage.

Yes, but without the storage this would be even worse, more storage would help here.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact