Tesla To Open Up 10% Of Superchargers To Other Cars, But It’s More Complex And The Plan’s Misguided

Apr 18, 2023

Teslas charge at a busy California supercharger. Soon, federal grants will let non-Tesla CCS cars ... [+] charge at 10% of these stalls

A recent announcement from the White House touts a partnership with EV companies to push deployment of EV charging infrastructure. Most notable was the inclusion of Tesla, which has been left out of many government EV meetings and plans, and which runs its own charging network — which while exclusive to Tesla, has more stalls than the other charging networks combined.

The announcement states Tesla will open a small fraction of its charging stalls to non-Tesla cars, which use the CCS charging plug used by all cars except Teslas, the Nissan Leaf and a few legacy cars. (However, there are only about 1/3rd as many CCS cars on the roads as there are Teslas.) This has generated excitement because the harsh reality for owners of CCS cars is that the Tesla network is not just larger but widely viewed as superior, and not just a little bit superior. Because Teslas can use the CCS network with an adapter whose price was just lowered to $175, they have access to around 28,000 stalls, while CCS cars have about 12,000 stalls with a reputation for low reliability.

The details are complex, though. Here are some important points to consider:

While I have been highly critical of Tesla's FSD and many other aspects of the company, there is almost universal consensus that Tesla's charging network is better than the others, not not just a little better but a lot. I’m going to sound like a Tesla Stan fanboy because the difference is just that stark. It's actually a little bit baffling, because Tesla has no special secrets to make this so, all the aspects of their design have been public and well understood for almost a decade, and their network was large long before the others were built.

The White House boldly touts that their plan, with over $7B in subsidies will result in the following attributes for the improved charging experience. To quote from the White House declaration:

What's bizarre is that these goals describe what the Tesla network already is. Tesla built it without subsidies — at least on stations — but the smaller network, built with massive subsidies, is a mess, rarely meeting those goals. It seems ironic that the federal government things a bigger subsidy is the way to get the alternate network up to speed.

A comparison of the size of Tesla's combined slow and fast charge cable and the CCS fast charge ... [+] plug.

It's not a coincidence. The problem is that EV charging is not a business as yet, and it's difficult to make it one. People imagine it being a business like a gas station, selling commodity fuel at a small margin and snacks for most of the profit. EV charging doesn't work like this at all, and it can't be easily shoehorned into the gasoline mode of operation. EV charging is a service business. The underlying commodity — kWh of electrical energy — costs around 3 cents/kWh wholesale at the power plant. By the time you pay 40-55 cents/kWh at a fast charger, you’ve paid almost all the cost for transmission, charging equipment and other associated costs. And not much profit. Some charging stations are free. Stations at home are about 13 cents/kWh national average (though recently up.) Public slow stations range from free to 55 cents, but typically are around 25 cents.

If you mistakenly think it's like gasoline, it's a world where almost everybody has a slow gas station at home for $1/gallon, public slow stations are sometimes free but usually around $2/gallon, and fast stations along the freeway and for those who don't have the $1 station at home are $3-$4/gallon. Many hotels throw in a fill-up for free, too. How do you make a business in that world?

Tesla first decided not to. Their initial charging network was free to customers for the life of the car. They paid for it out of car sales, because people bought the cars, knowing they could take road trips and charge for free. The initial chargers were only on the freeways, not in most big cities. They weren't a business, they were there to sell cars — though selling cars was a great business.

Later, they started billing, but said it was on a break-even basis. The earlier customers still got it free for the life of their car. More recently they have raised the price to a higher level which presumably makes some profit, and funds the expansion of the network.

Even billed for, the network still does it job selling cars. The simple truth is if road trips are important to you, you have to hate Tesla a lot to buy anything else. That was 100% true a couple of years ago, and is slowly becoming less true, but it's still a large factor. This is why people are surprised that Tesla will let other cars use their crown jewels — and why they are only opening up 10% of them.

Here are the things that are better about Tesla's network, and why:

In sort, "it just works" and it's easier than pumping gas.

This is true to some degree even with slow charging. Tesla's home charging units are the best value of any major manufacturer. Their mobile connector is the best portable charger and just $230 (it used to be free with the car.) Unlike most of the others, you can get it to support 8 different plugs you might find on the road in RV parks and garages and welding shops and more. Their home wall charger at $425 has wifi and charge sharing and costs less than most major brands of J1772 chargers. Most people can actually just use the $230 unit in their homes. (And Tesla accessories are usually expensive as would be expected for cars in this price range.)

In other words, again, Tesla figured out charging, showed the world, and the world ignored them and screwed it up. It's that dramatic a difference.

Charging providers can apply for NEVI grants for spaces that close out a 50-mile gap on the highways. A NEVI grant requires at least 4 stalls. Each must be able to do 150kW at the same time. They must take credit cards, support CCS, and publish their availability to show up in charging maps. No membership may be demanded (though the price can be lower with a membership.) Uptime must be at least a (fairly sucky) 97% which means 262 hours of downtime is allow each year. Support for "plug-and-charge" a committee-designed gnarly protocol to duplicate Tesla's easy charging experience, is to be required.

A lot of the grants will be NEVI money. The numbers above suggest that Tesla won't be making every stall take CCS cars, so they may just designate 4 stalls as NEVI stalls in a larger station of 8 to 12 stalls. (Only 10% of Tesla stations have fewer than 8 stalls.) Tesla stalls don't have screens or card readers because they don't need them. Tesla may have to add those at extra cost. Tesla makes heavy use of their in-car system and their app for charging — for NEVI CCS customers using credit cards or another app, they will have to do without. The app helps you manage everything and tells you when your charging session will be ending so you can come back and move your car. It's easy to understand the NEVI rules about not requiring a special app — the need to download more than a dozen apps to use the many different charging networks is frustrating, particularly if you come to a new station in a region without cell service — but a better answer would be to mandate a requirement that all station and networks talk a basic local protocol (bluetooth, NFC or internet if available) which any app or in-car software can use to do the basics.

The NEVI rules show 20th century thinking. You want an app to manage charging. You don't sit around waiting while you charge — you eat or do other things. That's why screens and credit cards are anachronisms, and should not be put into law in the modern era. The plug-and-charge protocol, while complex, is a good idea and I am sure Tesla will work to support it.

The 150kW requirement at the same time is meant to assure good service, but it's actually a big burden. Tesla does not design it this way because cars all start taking power at a high rate, but quickly decline By the time they are nearly full they may be drawing only 20kW. Because you are always charging a mix of freshly arrived empty cars taking tons of power and cars who have been there a while, the right approach is to share the power around lots of stalls, not guarantee it for four particular ones. Tesla's stations full of 250kW stalls don't have anywhere near 250kW or even 150kW provisioned for every stall at once. That's super expensive and pointless. It's much better to have lots of stalls and no waiting, even if in the very rare times the station is full of cars with empty batteries, you might not give everybody full power. It's a better experience for everybody. The NEVI rules muck that up.

Tesla will be forced to designate 4 of the stalls as special, and give them an unfair share of the power. Most of the time it probably won't be noticed, but some of the time it will make people fight over who gets the priority stalls.

In Oregon, the NEVI grants require one station be 350kW, which means at the 800 volt level supported by a few cars, but not used by Teslas. However, Tesla's newest v4 charger, meant to deliver as much as a megawatt to their semi-truck, handles 1000v. A speed of 350kW should be used fairly rarely, as it costs quite a bit more to provide it.

Tesla may also make a play, in time, to challenge the requirement that subsidized chargers use CCS. Last year Tesla released all intellectual property on its own connector, which is widely viewed as the best connector, and is the most widely used i the USA, on far more cars and more stations than CCS. They are encouraging other carmakers to use the connector, though only Aptera (which may never ship a car) has done so. For other carmakers, the connector would be secondary to getting access by their cars to Tesla's full network. If Tesla does get other carmakers to use this connector, it will be a competing standard to CCS, and a more successful one in terms of cars, though not in terms of carmakers. Tesla could argue that laws demanding CCS arbitrarily benefit the smaller, inferior, more expensive standard, and there would be a fair bit of merit to that claim. Adapters can be made available at low cost to allow interoperability.

This is why you have engineers design your charging station, not political committees.

Tesla stations have very short charging cords. Teslas all have the charge port in the very back left corner of the car, and the cord barely reaches that if you park as far back as you can. These cords will not reach the vast majority of other cars, unless they have their port in the same place (or the front-right corner.) If Tesla wants to serve CCS cars, they don't just need an adapter, they need a longer cord. Tesla's new cords are special, with liquid cooling to keep them thin and light. These short cords are part of how they have made their stations simple and cheap.

If Tesla wants to play games, they could actually just keep these cords and offer CCS adapters nobody can actually use. They will meet the legal requirement to support CCS — it will be the fault of the other carmakers that they didn't put the port in an accessible place. I doubt they will do this, but it's an option. They can also put a fat surcharge on their price, but give Tesla drivers (or anybody who buys a club membership for a fat price) the price they charge them today.

We could see a world where if you are a CCS driver, you can go to Tesla stations, but only 4 of the stalls will charge you, and they will have short cords that don't reach your car, and the price is high if you don't pay a fat monthly fee. Tesla could get the subsidies without really letting any competitors onto their network. But they say otherwise — they say that they really want to enable all electric cars, not just Teslas. They say their membership to get the price down will be just $1/month. So they will probably not do any of these tricks.

Even so, the stations might not be so enticing. Only 30% of cars are not Teslas, so CCS drivers will be in a minority at Tesla stations. If only 4 of the stalls work for them, usually they will be full of Teslas already, especially if they have artificial priority for getting power. And only 10% of stalls will support CCS at all.

Update: The first stalls with Tesla's CCS adapter, called a "Magic Dock" have appeared near the supercharger factory in Buffalo. They have the standard cord length so will not reach many CCS cars, unless the cars park sideways, dock blocking many of the stalls. One suspects some conflict may arise. It is not yet possible to activate the dock and get the CCS adapter out of it.

It should be noted that this battle went differently in Europe and China. In Europe, a law required that any station have at least one stall with CCS. Tesla used Europe's "Type 2" connector at first, which is similar to the USA's J1772 for slow charging, though they adapted it to also do fast charging. However, various factors in Europe made them decide to switch to using CCS2 — a system similar to USA CCS but based on the type 2 connector. They switched their superchargers to offer CCS2 and recently, and allowed non-Tesla cars to charge at their stations. However, due to the short cords, not all cars can do so, and some are known to park sideways (blocking several stalls) to make the cord reach.

A 2-stall EVgo charger being used by a Chevy Bolt at a grocery store. Note the long cables — this ... [+] Bolt could never reach the Tesla cord, even with an adapter.

Charging is a difficult business. Tesla built the largest network to sell cars. Electrify America built the 2nd largest network to pay penalties over dieselgate. Chargepoint, the largest independent network, doesn't sell electricity at all — they sell equipment and manage the network for their widely distributed array of customers. FLO is similar. Volta doesn't sell electricity, they give it away and put ad screens in parking lots. Only a few of the smaller companies, like EVgo, are trying to be like gasoline companies and make a profit from selling energy.

And most of them, except Tesla and a few smaller ones, have only been able to put in these no-business stations due to subsidies. A large fraction of the non-Tesla level 2 chargers also were paid for by subsidies — and that's why many of them are in silly places and very rarely used.

This is the reason that most of the networks have poor reputations for reliability. If it's not a business, there is little incentive to keep it operating well. These new rules hope to demand 97% uptime, and perhaps they will succeed, but that's not that great a number if it's for stations or even for stalls. (This is another advantage of large stations, because hardware failures in individual stalls don't cause much in the way of customer problems.)

In addition, if charging is a business — or at least is in support of a business like selling cars and making drivers happy — then charging stations are more likely to be in useful locations, have pleasant charging experiences, and competitive prices. If the business of charging is mostly paid by subsidies, the stations will serve that master, rather than car buyers. If the government wishes to kickstart the market, the best plan is not to dictate where and how to run chargers, but to create incentives to run it as a sustainable and viable business that meets the core public goal — namely facilitating the transition to emission-free driving and overcoming the advantage of the century-long head-start the highly subsidized oil industry has in deploying infrastructure.

The long term answer must be "where drivers want them to be, to provide the best recharging experience" and possibly, "to enable drivers to feel they can easily go anywhere in their cars that they could go with gasoline cars." The latter was Tesla's core plan, to make people feel comfortable buying a Tesla.

It turns out we haven't been putting chargers in these places. They have gone where subsidies and regulations push them to be. They have gone in places that want to "feel" green. My own conclusion from driving and observing EV driving is that there are these types of locations that are most useful:

The challenge for governments is that almost all charging must be installed on private lots, because that is where cars park, unless they park on the street. Governments are unsure of how to give subsidies to private facilities in a fair way that serves the public interest, but again, there is no choice — charging must be where people park in order to work. Perhaps loan programs would be better than grants, with the grants given to drivers on an income-based formula to pay any profits charged on the loan-funded private stations.

However, this does not mean regulations should identify these places — rather that the market will pay for these locations (except apartments and curbs, rural spots and some offices) on its own with the right programs. This list provides advice on how to judge if things are going well. What you don't want is to subsidize the following:

Such locations meet neither of the two business needs — making drivers feel free to go where they want in an EV, and providing good charging experiences that don't take time out of the day. It should also be clear that show chargers in the right parking places (#1 and #2 the the list of useful places) should represent the vast majority of charging stations. Such charging takes no time from a driver's day — it usually happens when sleeping or working — and is low cost to install and run, low stress on batteries and needs minimal new power infrastructure. If there are to be subsidies, >90% of them should apply where >90% of the stations belong.