240V V2L from North American Ioniq 5

[zanydroid]

I’m learning about the capabilities of V2L/V2H cars of the current model year to see if it’s worth buying one of those cars instead of building or buying a backup power system for my house. I am aware there will likely be more options once bidirectional CCS is on the market, but that’s kind of too abstract and theoretical (a 2025-2030 thing). The ones on the market right now are just a tad shy of a good fit for my requirements. (Eventually) this post will get to whether there is turnkey solar / battery storage hardware that can be used to bridge the gap. The car I’ll investigate here is the Ioniq 5, which in North America outputs 3.6kW split across two 5-15R (*note: I'm not sure if 3.6kW capacity has been confirmed by users yet, I'm going by what I consider non-authoritative sources, namely car reviewers and the PR announcements from Hyundai).

I believe KIA/Hyundai E-GMP cars (Ioniq 5 etc) and F150 Lightning are the main MY 2022 cars that support V2L in North America (Chademo cars are bidirectional but I don’t believe the hardware is easy to buy in North America). I have no need for a Lightning, so that leaves just the Ioniq 5. With V2L the most viable topology is manual transfer switch for backup power, although it’s probably possible to hack something together with OpenEVSE and a hybrid inverter to achieve grid tie and fast transfer to standby power.

So let’s assume backup power case. For my house, I could probably fit backup gas heater, computers, and refrigerators on a 120V critical loads panel. However, I would not be able to start the primary air handler to circulate that heat, since that power is provided on the 240V mini-split circuit.

This car has two 5-15R, one in the cabin (on Limited trim) and one in an optional accessory that plugs into the J1772 port. In Europe and Korea the charging port (IEC and J1772, respectively, AFAIK) are ~230V L-N @ ~15A.

Obviously the overseas adapter might be used with a transformer to form 120/240 @~3kW (what efficiency might I expect). If it works, this has the advantage of working with the lower trim models. So let’s set that aside for now, I don’t think anyone in North America has gotten their hands on one yet.

So with the officially supported North American config, I could use an autotransformer to step up one of the 5-15R to 120/240 with <1.8kW capacity. That could allow the mini-split to boot up, but would be well below the MCA. It would have to be started in circulate only mode. Maybe it'll work...

Another configuration is to have two separate 120V critical loads panels, each with a transfer switch and connected to the car. This would waste a lot of space in the panels, but is easy. Not sure if this is code compliant...

So to get enough power to safely start up and configure the circulation fan (and perhaps even run the mini-split at low power, in conjunction with the backup gas heater), it would be great if I could combine the two 5-15R. However, I’m not sure what the best topology for this would be. I think it would have to be DC coupled, since we cannot trust how the inverters on the car are arranged (no guarantee that they’ll stay in sync, etc etc).

To achieve the DC coupling, I believe I can use current sharing power supplies:
6 to 8 Meanwell SDR-480P-48 48V/480W DIN mount, 3 to 4 connected to each 5-15R ($170 each)
https://www.meanwell.com/productPdf.aspx?i=811
These are UL 508 listed, which I assume is robust enough for home use.

The 48V output of these can then go into a standard off-grid 48V->120/240 inverter, which then feeds into a transfer switch.

Are there other ways to configure this? Is there an automatic transfer inverter that can be used instead? Perhaps there's an interesting config that uses the in-cabin 5-15R to feed into a solar system, while the car is fed power via L2 charger. This would not be able to power as much standby load, but it could be used for things like peak trimming, (very slow) arbitrage, and guaranteed carbon-free operation overnight.

 

[再生可能 エネルギー]

I would be very interested knowing if the car will be sleeping when the V2L/V2H mode will be used.

What I mean is that if the car will not be sleeping, then the 800 V DC to 12 V DC inverter
will be running all the time, otherwise the 12 V battery will be drained quickly.

The Teslas had issues with the 12 V battery, so now the new models are going in a very deep sleep mode when parked.
However, when parked with the Sentry mode recording activated, the DC/DC inverter has to stay running all the time,
thus the propulsion battery get also discharged in a noticeable amount after a full night of using the Sentry mode.

In the case if the IONIQ, several owners experienced issues with his 12 V battery getting drained when parked.
See this video: "IONIQ 5’s 12-volt battery is dead: here's how I jump-started it."

On this video "Monitoring the IONIQ 5's 12-volt battery: installing and using a battery monitor"
the owner at 18:12 explains that the IONIQ has an 'Accessory' mode not running the DC/DC inverter,
and the 12 V battery get discharged quickly, and recommend pressing the brake pedal to activate
the DC/DC inverter when seating inside the car when parked. As explain also in this video:
"12v battery monitor logs: Kia e-Niro’s Auxiliary Battery Saver doesn’t do much saving!"

 

[zanydroid]

I'm still researching whether I want the car. So I don't know what will happen in this case.

Sentry mode is running all the cameras + image processing stack, so that's probably in the 10s of watts. I don't know how efficient the DC buck converter would be, it's probably simpler than DC->AC inverter and could be optimized for low current draw of sentry mode.

I think this would be more of an annoyance, since it's not a fundamental issue with the power system itself (which limited to 120V @ 15A is for some V2L/V2H applications). I already have a compact jump start battery that I leave in my garage just in case, so for home V2L use I have no worry about the 12V battery running down. At home I can also trickle charge the 12V ; add other fixed infrastructure as needed to work around firmware deficiencies.

If I'm not at home and am aware of V2L running down the 12V battery, I can carry jump start battery I guess, and fallback to roadside assistance or friendly nearby driver giving me a jump.

 

[FourFox]

I have owned my Ioniq 5 since September. It is fantastic!
Like most EVs, the car spends most of its time (more than 90%) sitting under a car port doing nothing.

I am now investigating the installation of a Hybrid Solar inverter system to use the very attractive EV tariffs that are available in the UK.
They offer a reduction of approx 80% off electricity prices for using electricity from the grid from midnight to 5am. Some suppliers actually offer to pay you for the electricity used!

A significant drawback of a Hybrid installation is the high cost of the storage batteries required.

It occurs to me that I already own a brand new 76kWh battery. It sits under my car port!

I have done some research into the possibility of using this battery in a vehicle to load (V2L) installation.
I was surprised to find that there already was a huge vehicle to grid (V2G) installation at an insurance company in the Netherlands.
See 'Fullycharged.show/episodes/' 'Largest vehicle to grid project on the planet'.
They are mainly using ChAdeMO connections which offer an easier 2 way AC and DC communication with the EV. These are more common in mainland Europe. However you will notice in the video that there is indeed an Ioniq 5 (presumably with its original CCS Combo type 2 connection) participating in this project. The standard CCS Combo type 2 system has the capability to DC charge at 350kW (no kidding) but the Ioniq 5 is apparently restricted to only 240kW....that should give plenty of scope!

My next step was to contact Hyundai to find out what was modified on the Ioniq 5 to enable this mode of accessing the battery.
The reply from their technical department was unhelpful!
I don't fancy probing around with a multi-meter looking for an 800v DC source with a fuse rating of over 300 amps.

Any suggestions?

PS I fired-up both the internal 240v ac and the external V2L systems on my Ioniq 5 this week. It all worked fine, except it took some time to work out how to remove the external adapter from the charging socket at the end of my trial run (you just have to press the 'unlock' key on the key fob...duh). I may be wrong but it appears that the Hyundai internal inverter is powered directly from the 400v (or possibly 800v) system even when in sleep mode and the car is locked. The 12v service battery would run down in a few seconds if a significant load were imposed on the AC outlets if this was used as a primary source of power.
Each AC outlet is rated at 16A (odd for a 13A socket). They can both work together, so that is 32A 240ac capacity. The external socket adapter only accesses the AC and communication section of the car charging socket. The manual says not to use very heavy loads on this system like air conditioners and washing machines (presumably to limit high power start-up surges). It is ok to charge another EV using these AC outlets.....or a storage battery on my Hybrid system.

It seems a bit pointless to use a 800v DC EV battery - into an internal inverter - into a 250v Ac/48v DC battery charger - into an auxiliary 48v battery and onto a main house inverter.

Any suggestions.

 

[zanydroid]

Y

I have owned my Ioniq 5 since September. It is fantastic!
Like most EVs, the car spends most of its time (more than 90%) sitting under a car port doing nothing.

I am now investigating the installation of a Hybrid Solar inverter system to use the very attractive EV tariffs that are available in the UK.
They offer a reduction of approx 80% off electricity prices for using electricity from the grid from midnight to 5am. Some suppliers actually offer to pay you for the electricity used!

A significant drawback of a Hybrid installation is the high cost of the storage batteries required.

It occurs to me that I already own a brand new 76kWh battery. It sits under my car port!

I have done some research into the possibility of using this battery in a vehicle to load (V2L) installation.
I was surprised to find that there already was a huge vehicle to grid (V2G) installation at an insurance company in the Netherlands.
See 'Fullycharged.show/episodes/' 'Largest vehicle to grid project on the planet'.
They are mainly using ChAdeMO connections which offer an easier 2 way AC and DC communication with the EV. These are more common in mainland Europe. However you will notice in the video that there is indeed an Ioniq 5 (presumably with its original CCS Combo type 2 connection) participating in this project. The standard CCS Combo type 2 system has the capability to DC charge at 350kW (no kidding) but the Ioniq 5 is apparently restricted to only 240kW....that should give plenty of scope!

My next step was to contact Hyundai to find out what was modified on the Ioniq 5 to enable this mode of accessing the battery.
The reply from their technical department was unhelpful!
I don't fancy probing around with a multi-meter looking for an 800v DC source with a fuse rating of over 300 amps.
Any suggestions?

You need bidirectional CCS to DC couple the car. Bidirectional CCS is not coming out until CCS 3.0, which is supposed to be ~2025. I don't expect any car company to firmware update modify a 2025 model year car. You will not get the required safety listing for permanent installation in a house by modifying the car. Chademo has supported bidirectional DC since the start. I don't think you can expect Hyundai to know anything about configuring the car in V2H, except the engineers that (may) be working on getting the platform ready for CCS 3.0. They won't put you in touch unless you have clout or are a partner.

There are two products I'm aware of in North America, Wallbox Quasar 2, which is a AC <-> CCS device. They promise to patch it to support V2H. But of course your car probably won't get patched for bidirectional CCS. DCBel is a flexibly configurable hybrid inverter system, which has both CCS and Chademo options for DC coupling, plus MPPT. CCS to be patched in the future to support bidirectional. Both are ~$5K since that is the price range anyway of this size active power conversion device, but with the DCBel the converter can be reconfigured to do more than just AC couple a CCS vehicle.


Your 240V V2L is theoretically fine to "AC couple" to solar system. 80% efficiency on an EV battery is still a much bigger battery than most people install at home, and good enough for a few days. Highly doubtful that the V2L can AC couple directly to a grid tie solar since it doesn't do the solar control protocols. That's why I posted here, to ask about what some good options are to DC couple the system, convert voltage, and group the outputs to get at least 3kW of capacity. For your 240V L-N you can just use a transfer switch or plug directly into a generator input on a hybrid inverter.

I've read that the Ioniq 5 outlets are rated at 3.6kW total by the marketing material, but I don't really trust it since the spec was not listed in an official document AFAICT (I haven't come across anything written up as well as proper appliance or power supply specsheet, which would provide sustained and surge current). In principle, since the OBC is 10kW and the inverter runs the OBC in reverse (sharing the power transistors), it might be able to go higher.

PS I fired-up both the internal 240v ac and the external V2L systems on my Ioniq 5 this week. It all worked fine, except it took some time to work out how to remove the external adapter from the charging socket at the end of my trial run (you just have to press the 'unlock' key on the key fob...duh). I may be wrong but it appears that the Hyundai internal inverter is powered directly from the 400v (or possibly 800v) system even when in sleep mode and the car is locked. The 12v service battery would run down in a few seconds if a significant load were imposed on the AC outlets if this was used as a primary source of power.
Each AC outlet is rated at 16A (odd for a 13A socket). They can both work together, so that is 32A 240ac capacity. The external socket adapter only accesses the AC and communication section of the car charging socket. The manual says not to use very heavy loads on this system like air conditioners and washing machines (presumably to limit high power start-up surges). It is ok to charge another EV using these AC outlets.....or a storage battery on my Hybrid system.

The 16A rating is probably to be compatible with North America receptacles, 15A is the minimum.

You can probably connect the car to loads with soft start. In North America this is indicated by Rotor Lock Amps measurement. Inverter driven appliances have a better shot of working since they have built in soft start.


You can read some of my social media posts with analysis and replies here.

https://www.reddit.com/r/solar/comments/t9807n

https://www.reddit.com/r/solar/comments/t8ccns

And this one from someone else a while back:

https://www.reddit.com/r/solar/comments/i66bb0

 

[alexw]

Some hard facts in no particular order (some of which repeat zanydroid already said, sorry). I own an EV6, and have personally tested/measured/opened and observed all of these things except where indiciated. I have a USA car, so keep that in mind also:

• The EV6/Ioniq 5 use a bidirectional ICCU (integraded charge controller unit) to drive their V2L capabilities, i.e. can't be used while charging.
• The interior and exterior outlets are wired together. Not QUITE directly - I think the interior has a relay that can disable it independently based on whether a plug is physically present, probably for child finger reasons. I'm not actually certain the outlets share the same power path/phase, but I AM certain they share the same overcurrent limit (tested). It's 15A total, not 15x2. 1.8kW max USA, 3.6kW elsewhere where the standard mains is 230 or 240.
• The V2L output depends on the region of the car, not the connected (passive) adapter. Korea and USA share J1772 as a charge standard, so the 230V korean adapter plugs into USA cars. But you get 120V 60Hz out, not 230.
• I say (passive) adapter because (future? WIP? not sure) V2L standards will include (I believe) powerline communication for connection management. I have to assume a proper bidirectional EVSE that actively communicates by such a mechanism could get 240V out of the north american configuration.
• The V2L adapter is very simple and you can cheaply build your own. Multiple people have, though I'm struggling to resurface the threads. It's basically a J1772 (or type 2) plug and a couple resistors, VERY close to an actual J1772 cable but with the PP and CP pins shorted together to indicate "output please"
• CHAdeMO has always "supported" V2L, but I believe only DC-out V2L. I.e. there's standardized communication for the function, but you still need an EVSE with a fat stack of (expensive) inverters in it, a-la bidirectional CCS as well.

It's pretty unclear to me how Hyundai intends to support V2H or V2G on the current crop of cars. I'm certain it is/was intended, but absent robust standards maybe the intend to roll out a software update to support it, maybe not? The biggest issue in the USA is that the plug only has hot/neutral/ground, so it's impossible to support a backup V2H situation without an (auto-or-traditional)transformer external to the car. This is not necessarily an issue in Type 2 regions, but I don't know whether the ICCU supports 2 or 3 phase output either. I'd assume so but not yet publicly known.

As for solar harvesting, looked into that pretty extensively too. Some more relevant facts:

• The car uses negligible power in sleep mode, on the order of a percent every few weeks
• When charging, on, etc, the electronics draw minimum 300W.
• A tesla, by comparison, consumes about the same 300W, which is why Sentry mode is such a power sink.
• This is a huge killer for low-power charging efficiency. The minimum to even play ball is 300W burned right off the top doing no charging at all, so if you had a 600W array, you'd be talking 50% charging efficiency before even considering any other losses.
• The J1772 standard doesn't support indicated capacity lower than 6A (at least with basic PWM communication), so you can't even START charging with less than 720W, 300 of which as mentioned is wasted. Not sure how the Type 2 standard compares.
• This all implies that "DC slow charging" would be great, but absent a very specific power mode being supported by the car (like whatever the European solar roof Ioniq 5 variant does), you'd still be stuck with the 300W loss off the top.

 

[zanydroid]

Some hard facts in no particular order (some of which repeat zanydroid already said, sorry). I own an EV6, and have personally tested/measured/opened and observed all of these things except where indiciated. I have a USA car, so keep that in mind also:

• The EV6/Ioniq 5 use a bidirectional ICCU (integraded charge controller unit) to drive their V2L capabilities, i.e. can't be used while charging.
• The interior and exterior outlets are wired together. Not QUITE directly - I think the interior has a relay that can disable it independently based on whether a plug is physically present, probably for child finger reasons. I'm not actually certain the outlets share the same power path/phase, but I AM certain they share the same overcurrent limit (tested). It's 15A total, not 15x2. 1.8kW max USA, 3.6kW elsewhere where the standard mains is 230 or 240.
• The V2L output depends on the region of the car, not the connected (passive) adapter. Korea and USA share J1772 as a charge standard, so the 230V korean adapter plugs into USA cars. But you get 120V 60Hz out, not 230.
• I say (passive) adapter because (future? WIP? not sure) V2L standards will include (I believe) powerline communication for connection management. I have to assume a proper bidirectional EVSE that actively communicates by such a mechanism could get 240V out of the north american configuration.
• The V2L adapter is very simple and you can cheaply build your own. Multiple people have, though I'm struggling to resurface the threads. It's basically a J1772 (or type 2) plug and a couple resistors, VERY close to an actual J1772 cable but with the PP and CP pins shorted together to indicate "output please"
• CHAdeMO has always "supported" V2L, but I believe only DC-out V2L. I.e. there's standardized communication for the function, but you still need an EVSE with a fat stack of (expensive) inverters in it, a-la bidirectional CCS as well.

It's pretty unclear to me how Hyundai intends to support V2H or V2G on the current crop of cars. I'm certain it is/was intended, but absent robust standards maybe the intend to roll out a software update to support it, maybe not? The biggest issue in the USA is that the plug only has hot/neutral/ground, so it's impossible to support a backup V2H situation without an (auto-or-traditional)transformer external to the car. This is not necessarily an issue in Type 2 regions, but I don't know whether the ICCU supports 2 or 3 phase output either. I'd assume so but not yet publicly known.

Hey, really appreciate the details and for confirming all this.

Here is an English language thread I've been following on how to build a passive adapter.

V2L inner secrets (Not Many)

www.ioniqforum.com

Since I posted my message, there have been ~two V2X pilots announced by Hyundai group (in Europe, if I'm not mistaken). I haven't dug around to see what they use, since it's not super relevant to me in the US and I expect the juicy details to be trouble to hunt for :-(

The other thing I was investigating is bidirectional CCS, that is available with one of the F-150 Lightning, and VW has announced that they'll patch ID4s to support bidirectional CCS (IMO announcement of OTA is stronger than an announcement of a pilot, wrt actually being available to a lot of people. ofc it could still all be vaporware). CCS indeed uses power line communications to do all sorts of stuff. Haven't dug into it myself, and I don't think the protocols are public; I imagine they're logically similar to Chademo which has been bidirectional for longer and has had more DIY tinkering, to my understanding.

But the CCS standard has not been finalized, and based on my completely unscientific poll on reddit last week + reading DIY EV forum, nobody is anywhere near having hackable hardware, and it's likely that safety checks in the BMS on unidirectional CCS cars will prevent you from drawing current (IE, it would disconnect the battery on detecting the anomaly). My guess is the hardware will converge to <$2K when it becomes available, based on the cost of the shore-side charger for the F-150 and the cost of solar inverters. Plenty of listed solar inverters at 6kW, between $1500-2000 (and of course a ton of Chinese ones), and the difference here is basically "just" adding a PLC and software implementation [which I'm led to believe is non-trivial and requires a full engineering team to do properly].

Personally I'm betting on cars implementing just CCS2 and letting shore side equipment deal with actual V2G (utility integration) and V2H (grid forming etc). That feels like a better separation of concerns, and the shore side equipment will easily have a 15 year lifespan and is in the ballpark of what you would pay for solar tweaks to add storage inverter or panel upgrade anyway as a homeowner.

Is 3-phase OBC support required in Type 2 market? I guess maybe in the countries where 3-phase is prevalent at homes.

 

[zanydroid]

• The car uses negligible power in sleep mode, on the order of a percent every few weeks
• When charging, on, etc, the electronics draw minimum 300W.
• A tesla, by comparison, consumes about the same 300W, which is why Sentry mode is such a power sink.
• This is a huge killer for low-power charging efficiency. The minimum to even play ball is 300W burned right off the top doing no charging at all, so if you had a 600W array, you'd be talking 50% charging efficiency before even considering any other losses.
• The J1772 standard doesn't support indicated capacity lower than 6A (at least with basic PWM communication), so you can't even START charging with less than 720W, 300 of which as mentioned is wasted. Not sure how the Type 2 standard compares.
• This all implies that "DC slow charging" would be great, but absent a very specific power mode being supported by the car (like whatever the European solar roof Ioniq 5 variant does), you'd still be stuck with the 300W loss off the top.

Off hand, it feels like you could use a battery inverter to handle the low power charging cases, and the appropriately sized power system should be pretty modestly priced (at least on charging side, if you want to pull more than 300-600W on discharge then the price will escalate). What is the advantage of pulling all the system complexity of involving an EV?

 

[alexw]

PLC and software implementation [which I'm led to believe is non-trivial and requires a full engineering team to do properly].

Haha, definitely. The moment anything hardware becomes involved, that's basically the case. Even easy board spins are a whole thing. I'm guessing bidir CCS is "easy" on the ID.4 because they made the decision early to support plug-and-charge, implying PLC and an architecture that supports software defined features. Of course, it remains to be seen whether some minute aspect of the bidirectional standard pops up that their hardware isn't equipped to handle and tanks the whole thing. Probably not though.

Personally I'm betting on cars implementing just CCS2 and letting shore side equipment deal with actual V2G (utility integration) and V2H (grid forming etc). That feels like a better separation of concerns, and the shore side equipment will easily have a 15 year lifespan and is in the ballpark of what you would pay for solar tweaks to add storage inverter or panel upgrade anyway as a homeowner.

It's certainly "easier" and more flexible, but it just sucks from the standpoint of duplicating $1-2k worth of hardware (inverters). And like, what's all the fuss about E-GMP's "vehicle-to-stuff" bidirectional inverter if you're just going to buy one aftermarket? Also I think you're mostly right about cost comparisons between solar inverters and V2G, but there's still room for V2G inverters to cost a premium: gotta support 800V, instead of 400 or lower from a typical solar string.

Is 3-phase OBC support required in Type 2 market?

I don't actually know. I assume so, based on the fact that Tesla's inverters are all 3-channel and the NA variants just ship with some parts unpopulated to save cost. European (didn't actually watch this vid though...)

and North America:

(watching that video again, maybe they're actually identical and the depop'd parts aren't used in any locale?)

I also am not sure that 3-phase is "common" in residences in any market. I think it's more split along commercial vs residential charge points.

 

[alexw]

Off hand, it feels like you could use a battery inverter to handle the low power charging cases, and the appropriately sized power system should be pretty modestly priced (at least on charging side, if you want to pull more than 300-600W on discharge then the price will escalate). What is the advantage of pulling all the system complexity of involving an EV?

I don't really follow, sorry

 

[zanydroid]

I don't really follow, sorry

What's the reason to want to use an EV to charge at 300W-600W? You could get pretty far with a small Bluetti for that power level, and you don't have to pull in all the worries about what's running on the car.

 

[zanydroid]

Haha, definitely. The moment anything hardware becomes involved, that's basically the case. Even easy board spins are a whole thing. I'm guessing bidir CCS is "easy" on the ID.4 because they made the decision early to support plug-and-charge, implying PLC and an architecture that supports software defined features. Of course, it remains to be seen whether some minute aspect of the bidirectional standard pops up that their hardware isn't equipped to handle and tanks the whole thing. Probably not though.

The comparison I had in mind was to L2 EVSE. There's also the higher voltages involved, DC safety, and the complexity being on pulled into the charger rather than being dealt with in the OBC.

It's certainly "easier" and more flexible, but it just sucks from the standpoint of duplicating $1-2k worth of hardware (inverters). And like, what's all the fuss about E-GMP's "vehicle-to-stuff" bidirectional inverter if you're just going to buy one aftermarket? Also I think you're mostly right about cost comparisons between solar inverters and V2G, but there's still room for V2G inverters to cost a premium: gotta support 800V, instead of 400 or lower from a typical solar string.

Solar and fixed storage voltages are going up too.

You could argue also that $1-2K worth of OBC is overkill, and you're "consuming" one (paying more up front, perhaps also suffering the depreciation on resale) every time you switch cars. Vs buying it once in the home for 15+ years. In a perfect world, all OBCs will interop with your house identically. In an imperfect world, when you switch cars you have to worry about it interop'ing correctly on V2H level.

The bidirectional power system might be reconfigurable to buck from the EV operating voltage down to 400V (the E-GMP already has a boost from 400V to 800V onboard), maybe it's even required in CCS2 for maximum compatibility. This would be a simpler thing than grid coordination. Arguably you could get grid forming to work in the car, but based on forum threads I've seen I don't think UL1741 interoperability is super trusted, so why have a car involved in that when you might need the ability to make a lot of tweaks. Cars are pretty closed platforms to my understanding.

We'll have to see how future proof the E-GMP bidirectional inverter is for use cases other than buddy charging and V2L. AC coupling, across all the different power standards around the world, is tricky. So unlike charging side, if car wants to solve V2H unobtrusively for all possible markets it will have to deal with a lot of different cases.

I don't actually know. I assume so, based on the fact that Tesla's inverters are all 3-channel and the NA variants just ship with some parts unpopulated to save cost. European (didn't actually watch this vid though...)

and North America:

(watching that video again, maybe they're actually identical and the depop'd parts aren't used in any locale?)

I also am not sure that 3-phase is "common" in residences in any market. I think it's more split along commercial vs residential charge points.

This talks about a few cases:

https://www.reddit.com/r/electricians/comments/jm8e7c

By my understanding, in the US 120/208 is mainly used for some high density residential.

 

[alexw]

What's the reason to want to use an EV to charge at 300W-600W? You could get pretty far with a small Bluetti for that power level, and you don't have to pull in all the worries about what's running on the car.

Oh, to charge the EV itself. Like, if you wanted to use your car to energy harvest from your solar array rather than exporting to grid. The idea is, to do that efficiently under current circumstances, the car will take a big bite out of your generation simply by being on. If you had a 1kW array, you basically couldn't use that to charge the EV, despite 1kW being a perfectly decent amount of solar capacity. You'd be much better off with a powerwall despite already owning a big battery.

You could argue also that $1-2K worth of OBC is overkill, and you're "consuming" one (paying more up front, perhaps also suffering the depreciation on resale) every time you switch cars. Vs buying it once in the home for 15+ years. In a perfect world, all OBCs will interop with your house identically. In an imperfect world, when you switch cars you have to worry about it interop'ing correctly on V2H level.

Certainly a fair point. And in that comparison, the car's battery is the only true consumable, where the shore-side equipement, duplicated or not, will last quite a bit longer. But $2k for an EVSE is still a pretty high barrier to entry when you already own the car that can do the whole job itself, IF it can do the whole job itself. $2k is a standalone gas generator that can power your whole house as long as you have gas. Obviously that's not what I WANT, but if the goal is "power the house" it's hard not to compare the generator vs the car.

The bidirectional power system might be reconfigurable to buck from the EV operating voltage down to 400V (the E-GMP already has a boost from 400V to 800V onboard), maybe it's even required in CCS2 for maximum compatibility. This would be a simpler thing than grid coordination. Arguably you could get grid forming to work in the car, but based on forum threads I've seen I don't think UL1741 interoperability is super trusted, so why have a car involved in that when you might need the ability to make a lot of tweaks. Cars are pretty closed platforms to my understanding.

We're pretty deep into speculation here, but I agree with all your points. My baseline hypothesis is that Hyundai already designed-in 1) 400V out in anticipation of bidirectional CCS, 2) grid-forming (already has it) and grid-following (yet to be seen) capability in the ICCU to support V2G and V2H 3) the necessary PLC stuff for J1772-based V2L applications (vis-a-vis I think it's plug-and-charge ready and they just don't have the software integration sorted out yet). Of course, there's plenty of room for all of that hypothesis to be true, and still not come to pass in terms of delivered features, if the standards change enough by ratification that they can't software update their way to compatibility.

We'll have to see how future proof the E-GMP bidirectional inverter is for use cases other than buddy charging and V2L. AC coupling, across all the different power standards around the world, is tricky. So unlike charging side, if car wants to solve V2H unobtrusively for all possible markets it will have to deal with a lot of different cases.

Yep.

This talks about a few cases:

Yeah, I think I've heard Denmark it's the norm, but otherwise I wasn't aware it's common in Finland and Germany too.

 

[zanydroid]

We're pretty deep into speculation here, but I agree with all your points. My baseline hypothesis is that Hyundai already designed-in 1) 400V out in anticipation of bidirectional CCS, 2) grid-forming (already has it) and grid-following (yet to be seen) capability in the ICCU to support V2G and V2H 3) the necessary PLC stuff for J1772-based V2L applications (vis-a-vis I think it's plug-and-charge ready and they just don't have the software integration sorted out yet). Of course, there's plenty of room for all of that hypothesis to be true, and still not come to pass in terms of delivered features, if the standards change enough by ratification that they can't software update their way to compatibility.

I only know that it has the most basic grid-forming, comparable to a generator. I'm not aware of evidence that it can grid-form in the sense of AC-coupling with UL1741 inverters in a usable way when the grid is down. Here usable is intentionally a bit vague, for some people usable means the car is able to be topped off automatically with excess solar (no, it can't) so you can go days off grid without worry. For other people even having all the inverters shut down completely most of the time because they go into full anti-islanding mode is usable (I would bet this is how it works today).

It'll be interesting to see how the market will compare between requiring a simple shore-side box that's an auto-transformer + basic sensors, with all the smarts on the car, or a bidirectional charger. After factoring in normal installation from a licensed contractor, the price difference is probably not that much. If you're doing it DIY, sure, the price difference will be substantial. If you're a government or utility installing a few thousand of these, maybe the cost savings would really add up.

The fact that the market for AC coupled V2H will be quite fragmented also factors into the decision of which way the car companies want to go... they could just go for the 1 or 2 configurations corresponding to the lowest hanging/biggest market, which may not be one of the ones you need, and CCS for the rest.

 

[Bigbillsd]

Then what is the lightening giving us this year. I read it had V2H built in and it has CCS, is it prerelease of the 2025 version 3?

 

[zanydroid]

Then what is the lightening giving us this year. I read it had V2H built in and it has CCS, is it prerelease of the 2025 version 3?

You supposedly can buy from Ford a V2H capable DCFC for your house today. The web shop and price were posted last month. I presume it is some kind of pre-standard implementation, since the fundamental power coupling is the same — high voltage DC. To my knowledge it is the only CCS V2* implementation you can buy yourself as an end user, everything else is locked behind a utility, etc trial.

I don’t own the car nor intend to buy one, so you may want to ask on lightning forums.

 

[alexw]

You supposedly can buy from Ford a V2H capable DCFC for your house today. The web shop and price were posted last month. I presume it is some kind of pre-standard implementation, since the fundamental power coupling is the same — high voltage DC. To my knowledge it is the only CCS V2* implementation you can buy yourself as an end user, everything else is locked behind a utility, etc trial.

I don’t own the car nor intend to buy one, so you may want to ask on lightning forums.

This is roughly the same info I've heard as well - Ford's chosen solution for Lightning V2H is DC, buy-extra-inverters style a-la Wallbox Quasar V2. Probably compatible with that as well if I had to guess.

Kind of a bummer in my book given how much inversion capacity is built into the truck already.

OTOH, if you have the high-spec version with the 240V 7kW output or whatever it tops out at, I'm sure you can just plug a generator lead into a generator outlet in your home and get full-panel V2H with an isolation switch. Not quite as good as the hardwired V2H box, but better than the current options with I5/EV6

 

[zanydroid]

The F-150 has some pretty overkill power system specs, inline with it being a flagship halo product. The 240V V2L is has got to be sized to run some serious worksite equipment (often pretty yucky and picky inductive loads. I'm sure if the Lightning fails at a jobsite, Ford will get roasted and try to make it right, while if a E-GMP vehicle has issues with a heavy load, Hyundai is not really going to care). The Extended Range versions come with dual on-board charger (48A * 2) AND the Charge Station Pro (80A <-> CCS) discussed above.

 

[Dathes]

My setup is a Conext XW Pro with two EG4 Lifepower4 batteries 48v and a cheapo no-name LFP 25amp battery charger, all connected in parallel using a busbar. I can disable battery charging on the Conext and plug the V2L of my Kia EV6 into the battery charger. This way, the Conext drew power from both the EG4 LPs and the EV6 during a power outage two days ago. The display on the car's dashboard said 1.3Kw was being consumed while the charger can only do 48v*25a = 1200w.

I'm now checking if I can connect the V2L adapter directly to the Conext. Perhaps the generator 120v/240v split-phase can be wired for that as I believe the Conext can perform load balancing between different phases.

 

[kmschwem]

Y


You need bidirectional CCS to DC couple the car. Bidirectional CCS is not coming out until CCS 3.0, which is supposed to be ~2025. I don't expect any car company to firmware update modify a 2025 model year car. You will not get the required safety listing for permanent installation in a house by modifying the car. Chademo has supported bidirectional DC since the start. I don't think you can expect Hyundai to know anything about configuring the car in V2H, except the engineers that (may) be working on getting the platform ready for CCS 3.0. They won't put you in touch unless you have clout or are a partner.

There are two products I'm aware of in North America, Wallbox Quasar 2, which is a AC <-> CCS device. They promise to patch it to support V2H. But of course your car probably won't get patched for bidirectional CCS. DCBel is a flexibly configurable hybrid inverter system, which has both CCS and Chademo options for DC coupling, plus MPPT. CCS to be patched in the future to support bidirectional. Both are ~$5K since that is the price range anyway of this size active power conversion device, but with the DCBel the converter can be reconfigured to do more than just AC couple a CCS vehicle.


Your 240V V2L is theoretically fine to "AC couple" to solar system. 80% efficiency on an EV battery is still a much bigger battery than most people install at home, and good enough for a few days. Highly doubtful that the V2L can AC couple directly to a grid tie solar since it doesn't do the solar control protocols. That's why I posted here, to ask about what some good options are to DC couple the system, convert voltage, and group the outputs to get at least 3kW of capacity. For your 240V L-N you can just use a transfer switch or plug directly into a generator input on a hybrid inverter.

I've read that the Ioniq 5 outlets are rated at 3.6kW total by the marketing material, but I don't really trust it since the spec was not listed in an official document AFAICT (I haven't come across anything written up as well as proper appliance or power supply specsheet, which would provide sustained and surge current). In principle, since the OBC is 10kW and the inverter runs the OBC in reverse (sharing the power transistors), it might be able to go higher.

The 16A rating is probably to be compatible with North America receptacles, 15A is the minimum.

You can probably connect the car to loads with soft start. In North America this is indicated by Rotor Lock Amps measurement. Inverter driven appliances have a better shot of working since they have built in soft start.


You can read some of my social media posts with analysis and replies here.

https://www.reddit.com/r/solar/comments/t9807n

https://www.reddit.com/r/solar/comments/t8ccns

And this one from someone else a while back:

https://www.reddit.com/r/solar/comments/i66bb0

Sorry to resurrect an old thread here. I've been wondering about the V2H bi-directional EVSEs. Could this be done in a car model independent way? From my understanding there is some hand shaking communication between the EVSE and car to prepare for DC charging. If the car's happy it closes the relays to connect the DC port of the charge connector directly to the battery. Could an EVSE tell the car it wants to L3 charge, cause it to connect the battery, and then draw from the battery to power the 800VDC -> 240VAC inverter? The car firmware would see that the current was flowing out -- wonder if it'd open the relays?

 

[Grid Point Designs]

Some hard facts in no particular order (some of which repeat zanydroid already said, sorry). I own an EV6, and have personally tested/measured/opened and observed all of these things except where indiciated. I have a USA car, so keep that in mind also:

• The EV6/Ioniq 5 use a bidirectional ICCU (integraded charge controller unit) to drive their V2L capabilities, i.e. can't be used while charging.
• The interior and exterior outlets are wired together. Not QUITE directly - I think the interior has a relay that can disable it independently based on whether a plug is physically present, probably for child finger reasons. I'm not actually certain the outlets share the same power path/phase, but I AM certain they share the same overcurrent limit (tested). It's 15A total, not 15x2. 1.8kW max USA, 3.6kW elsewhere where the standard mains is 230 or 240.
• The V2L output depends on the region of the car, not the connected (passive) adapter. Korea and USA share J1772 as a charge standard, so the 230V korean adapter plugs into USA cars. But you get 120V 60Hz out, not 230.
• I say (passive) adapter because (future? WIP? not sure) V2L standards will include (I believe) powerline communication for connection management. I have to assume a proper bidirectional EVSE that actively communicates by such a mechanism could get 240V out of the north american configuration.
• The V2L adapter is very simple and you can cheaply build your own. Multiple people have, though I'm struggling to resurface the threads. It's basically a J1772 (or type 2) plug and a couple resistors, VERY close to an actual J1772 cable but with the PP and CP pins shorted together to indicate "output please"
• CHAdeMO has always "supported" V2L, but I believe only DC-out V2L. I.e. there's standardized communication for the function, but you still need an EVSE with a fat stack of (expensive) inverters in it, a-la bidirectional CCS as well.

It's pretty unclear to me how Hyundai intends to support V2H or V2G on the current crop of cars. I'm certain it is/was intended, but absent robust standards maybe the intend to roll out a software update to support it, maybe not? The biggest issue in the USA is that the plug only has hot/neutral/ground, so it's impossible to support a backup V2H situation without an (auto-or-traditional)transformer external to the car. This is not necessarily an issue in Type 2 regions, but I don't know whether the ICCU supports 2 or 3 phase output either. I'd assume so but not yet publicly known.

As for solar harvesting, looked into that pretty extensively too. Some more relevant facts:

• The car uses negligible power in sleep mode, on the order of a percent every few weeks
• When charging, on, etc, the electronics draw minimum 300W.
• A tesla, by comparison, consumes about the same 300W, which is why Sentry mode is such a power sink.
• This is a huge killer for low-power charging efficiency. The minimum to even play ball is 300W burned right off the top doing no charging at all, so if you had a 600W array, you'd be talking 50% charging efficiency before even considering any other losses.
• The J1772 standard doesn't support indicated capacity lower than 6A (at least with basic PWM communication), so you can't even START charging with less than 720W, 300 of which as mentioned is wasted. Not sure how the Type 2 standard compares.
• This all implies that "DC slow charging" would be great, but absent a very specific power mode being supported by the car (like whatever the European solar roof Ioniq 5 variant does), you'd still be stuck with the 300W loss off the top.

 

[zanydroid]

Sorry to resurrect an old thread here. I've been wondering about the V2H bi-directional EVSEs. Could this be done in a car model independent way? From my understanding there is some hand shaking communication between the EVSE and car to prepare for DC charging. If the car's happy it closes the relays to connect the DC port of the charge connector directly to the battery. Could an EVSE tell the car it wants to L3 charge, cause it to connect the battery, and then draw from the battery to power the 800VDC -> 240VAC inverter? The car firmware would see that the current was flowing out -- wonder if it'd open the relays?

Oops missed this question, took a few months off from this forum.

CCS2 is supposed to have a profile that will connect the battery (with potentially DC-DC conversion to an agreed on voltage, probably best to output a 600V or lower voltage to be more compatible with standard inverters and 600V wiring methods. Not sure what the line diagram looks like, and I don't know where to get the standards for free).

As for trying to do this with CCS, when I checked on DIY EV forum, very few people (zero in fact) in the business were willing to help people figure out how to do this. Some folks I talked to on other forums thought the BMS or other safety mechanisms on the car might get cross. Maybe a few cars would work in this hacked mode without complaining.

 

[byteharmony]

CCS2 is supposed to have a profile that will connect the battery (with potentially DC-DC conversion to an agreed on voltage, probably best to output a 600V or lower voltage to be more compatible with standard inverters and 600V wiring methods. Not sure what the line diagram looks like, and I don't know where to get the standards for free).

Given how cost effectively high voltages can be transmitted wouldn't it be amazing if a standard were to come into existence for 300 to 500v DC? Think about it, we could run a 10 AWG THHN wire pair right next to your AC panel and jack it into your next generation inverter (like a sol ark) that would be smart enough to charge your EV and use it as a disaster avoidant generator.

If we could just keep voltage levels to lower than THHN I think this could be come standard. Given that there are many more cheap cars made that could use these lower voltages than expensive ev's that may need higher voltages, this seams like a possible future reality. Could even move us in the direction of getting rid of AC charging of EV's.

Another option would be to have those really high voltages come into special car chargers that have DC BUCK and boost from the 500Volt max DC transition lines to the required voltage for the vehicle? The less stuff in the vehicle itself the lighter, faster, cheaper, etc it becomes.

 

[zanydroid]

Given how cost effectively high voltages can be transmitted wouldn't it be amazing if a standard were to come into existence for 300 to 500v DC? Think about it, we could run a 10 AWG THHN wire pair right next to your AC panel and jack it into your next generation inverter (like a sol ark) that would be smart enough to charge your EV and use it as a disaster avoidant generator.

If we could just keep voltage levels to lower than THHN I think this could be come standard. Given that there are many more cheap cars made that could use these lower voltages than expensive ev's that may need higher voltages, this seams like a possible future reality. Could even move us in the direction of getting rid of AC charging of EV's.

Another option would be to have those really high voltages come into special car chargers that have DC BUCK and boost from the 500Volt max DC transition lines to the required voltage for the vehicle? The less stuff in the vehicle itself the lighter, faster, cheaper, etc it becomes.

Incidentally I ran into some new-to-me info about the Ford Charger Pro, and started this thread on Reddit.

That charger appears to be little more than a L2 EVSE + CCS2 trigger interface. It connects to a hybrid inverter via RS485 and 600V DC wiring. #8 THHN for about 100A worth of backup power.

There is a pointer in there too to dcbel's single line diagram.

https://www.reddit.com/r/evcharging/comments/yjqwe0

Personally I doubt there will be a DC coupled standard any time soon. Based on my limited knowledge, barely anything can interop over serial these days. Also UL likes to qualify an integrated system (battery + inverter).

AC coupled system probably supports more interop due to UL1741SA, albeit janky because there's only so much that can be signaled that way.

 

[byteharmony]

Incidentally I ran into some new-to-me info about the Ford Charger Pro, and started this thread on Reddit.

That charger appears to be little more than a L2 EVSE + CCS2 trigger interface. It connects to a hybrid inverter via RS485 and 600V DC wiring. #8 THHN for about 100A worth of backup power.

There is a pointer in there too to dcbel's single line diagram.

https://www.reddit.com/r/evcharging/comments/yjqwe0

Personally I doubt there will be a DC coupled standard any time soon. Based on my limited knowledge, barely anything can interop over serial these days. Also UL likes to qualify an integrated system (battery + inverter).

AC coupled system probably supports more interop due to UL1741SA, albeit janky because there's only so much that can be signaled that way.

On the electrical side, technically this is great news. A simple voltage monitor that senses a low voltage on a 48v DC battery could flip a relay connecting the x00v power source to a DC buck device with constant current. Once it detects your programmed top voltage it disconnects the relay.


For a 20 amp DC device I could see those electronics coming in at under $100 if mass produced.
40 amps 150
60 amps 200
Etc.

We just need standards for home power storage and we are promptly heading that way ?