On

March 30 the Biden Administration released scant details of big spending in the upcoming $2T infrastructure bill to promote EVs. Included in that budget are supply chain incentives, purchase incentives (similar to or replacing the $7,500 tax credit available to people buying other than a Tesla TSLA or a Chevy Bolt.) Also planned is money to help a roll-out of 500,000 EV charging stations.

EVs are definitely the future, and they’re definitely a boon for the economy and the environment. They’re so good in fact, that they don’t need incentives on this scale, so if they’re going to happen, they should be done with care. There are over 80,000 public EV charging stalls in the USA now and a large fraction were installed because of subsidies and for emotional reasons, and many of them are nearly a complete waste — rarely used, and often in a state of disrepair. Stations are needed, but they must be valuable, so that they are well utilized and people work to maintain them.

The waste can be pinned on broad misunderstanding how how charging stations are used. Such understanding often only comes after you own an EV. Prior to owning one, people tend to imagine it works like gasoline, where you drive around until you are low, hunt for a station and fill up. They imagine driving an EV is full of “range anxiety” and worrying about how slow charging is, and it had some of that in its early days when cars were sold with less than 100 miles of range.

The same misunderstanding leads people to expect charging stations to be big and fast and thus expensive, which makes people imagine that huge amounts of money and big subsidies are necessary to make it happen. Most of the time, charging is done at night or at the office, while you sleep or work. The average car only drives about 35 miles/day, which means charging infrastructure only needs to provide very low average power, and only rarely is there a hurry to do it. Sometimes there is a hurry, which must be solved, but that’s for rare uses and rare days. It barely matters if it takes 30 minutes or 9 hours to recharge if you do it while you sleep or work.

MORE FOR YOU

Home Level 2 charging costs from $300 to $1200 in most homes, if they have modern electric panels near the garage. Older homes can face a cost of over $5K. The average public station install is $6,000 today, and often much more, but this could be much less. Level 3 fast stations are $50,000 today but should also drop, but 250kw stations are much more. Level 1 charging is already present in most homes or can be installed for a few hundred dollars in most cases.

As a result, the charging infrastructure should be broken down as follows:

  1. Cheap, slow charging at homes, hotels, offices, commuter lots, apartment parking lots and curbsides on some streets — the vast bulk of charging.
  2. Expensive, fast charging or swap for intercity travel
  3. Expensive fast charging or swap for the some fleets and the minority of urban drivers who don’t have access to #1, or on rare days of high use.
  4. In the future, a range of charging speeds for robocars

It’s important to understand the vast bulk of charging infrastructure is of type 1, though type 4 will grow in the future. This is where most of the power will be delivered. Most importantly this is the cheap infrastructure, which means the cost of this buildout is much lower than some people expect.

The cost difference can be immense. Slow charging can be done from ordinary household plugs, or better from slight variations which can deliver 2-3x the power with only modest extra cost in fresh installs. DC Fast charging stations and the power to run them can cost hundreds of thousands of dollars. Getting a power connection that can deliver hundreds of kilowatts to a megawatt is usually a large project, with high costs in part because of the safety issues in delivering that much power.

The first wave of charging infrastructure wasn’t planned well. Public stations consisted at first almost entirely of pricey Level 2 chargers, often put in useless locations where people only parked for short periods. Tesla made better choices, giving Level 2 charging gear for free to hotels, and installing its own network of superchargers on intercity routes. For a long time, only Teslas could do practical cross-country road-trips. Only now have the “standardized” CCS/ChaDEmo fast charging networks grown out to allow that, and even so, there are many reports of poor reliability and maintenance of this network, since most stations were created for other than business reasons, and thus no incentive to maintain them.

This is the danger in an infrastructure plan which will pay for 500,000 stations. It is important that stations be put in for a business reason — to sell electricity at a profit, to attract drivers who will buy other things at a profit (which is actually a big driver of gas stations) or to help sell cars (as Tesla Superchargers are, since they started off free to all Tesla drivers.) When they are put in for other reasons — compensation for the Dieselgate scandal, VC money, in order to get government grants or to present a “green” public image, there is not just a risk they will not be located in useful places. There is the risk that nobody will care about maintaining them if they break.

For those who own private parking, and those who charge in offices and commuter lots, slower, cheap charging is the most effective solution. Nothing wrong with having faster, full Level 2 if it’s cheap and easy to put in, but it’s not strictly necessary if cost is a factor. The problem is we have a body of car owners who don’t own their parking spot, can’t get charging at work and need a solution. For them, the solution can be:

  1. Fast charging stations can be used a bit like slow gas stations. If they are located next to places people frequently stop for about 30 minutes, they can fast charge 1-2 times/week when running errands. More rarely, they might need to just sit and wait, the price they pay for being in this situation. However, in the modern world with phones, this can be productive time or leisure time.
  2. Over time, apartment parking lots and curbsides on no-driveway streets can get low cost group charging systems (described below.) They can be forced or even subsidized to do so.
  3. Later this decade, many electric cars will feature the ability to drive themselves slowly, over short distances, to local charging depots when the car is parked and not in use The car will arrange its own charging at the depots — to the owner it will just seem they have a car that always is ready to go. A few times a year they might not get enough and will use method #1 above.
  4. Also in the long term, cheap/slow charging should be installed in a growing fraction of public parking spaces. This charging will exist not simply to keep cars charged, but also to use cars as repositories for excess solar power from 7am to 3pm.

Cheap, slow group charging

To vastly reduce the cost of bulk charging infrastructure, charging lots or streets can be designed around the average needs of cars. As noted, that’s just 35 miles, or under 10kwh of energy per day. Almost all public charging stations today supply from 3.3kw to 7kw, as most cars cant do AC charging at greater rates than that. The 5-7kw range is enough to be expensive, but not enough to give meaningful power during a short errand.

On the other hand, when cars will park 8-10 hours, as they do at night or at the office, you only need 1kw on average. That’s much cheaper, and in fact is easily provided by any household plug with a dedicated circuit breaker.

When putting together charging for a lot, a common approach is to install a small number of 7kw chargers. Those are enough to replenish the average car with 80 minutes/day, but people pay for them because they also want it to handle those rare dates when you need 250 miles which takes the full 8 hours. In a parking lot, though, there will typically not be enough of these chargers for all the cars, so even though they can handle the average car in 80 minutes, drivers must do a dance, moving their cars once full, and waiting for a free spot. No fun.

A better lot design for 40 cars has 40 slots, but only a small number can provide the full 7kw. Those are allocated to the few cars each day which actually need that. The rest of the spots can be a mix of 3kw and 1.5kw stations. These can be wired using the same wire thickness used in household plugs, which is safer and low cost. (3kw stations just need slightly different wiring to use 240v on the same wires as 1.5kw stations.) Each kw represents about 4 miles/hour of recharge speed for typical cars.

A group lot like this — which could exist at an office, in an apartment building or even along the curb of a street — would be a “smart lot.” Cars wanting to park and plug in would talk to the charging controller and tell it how much charge they need, and how long they will park. Cars that need a lot of charging would be assigned the fast stations. Most cars would be content with the slower ones, where they would stay plugged all day or all night, with no need to move when full or wait for a spot.

A special option exists for commercial buildings and lots which can get the industrial 3-phase 480v power. This power service provides 277 volts per line. Already many EVs can use 277v, and almost all could be designed to do so in time. Bump the wire thickness to 12AWG (still commonly used in homes) and you can provide 4.4kw with such service with minimal wiring, allowing almost all cars to take any spot and meet their needs, restoring 140 miles in a single day or evening. The total lot would manage the power to stay within its service limits. Again, because the average car only needs 1kw, you could service a 38 car lot with the same 200 amp service of a typical residential home today. (You can only use 80% of the total for charging cars.)

If the EV supply shifts heavier, with more trucks, their needs are higher though 240v-15a is normally more than enough, and 277v-20a woudl do even better.

The moral of the story: You can put in charging to meet the needs of lots of cars, and it doesn’t need to be very expensive.

Solar Consumers

If you’ve read about renewable power, you will have read about its one big wart — solar and wind are generated when the wind blows and the sun shines, not when people necessarily need the power. As we build out tons of renewable power, there will be times when there is too much of it, with nowhere for it to go. That’s where all these charging stations where cars park all day or all night can come to the rescue. If there are lots of low cost stations, you will eventually find one almost anywhere you park for more than an hour or two. If cars plug into these stations — even if they aren’t in a particular hurry for a charge — they can be set to opportunistically charge when there is a surplus of renewable power that needs to go somewhere that is being sold at a good rate.

Indeed, for the renewable power suppliers, they may find this ability to sell their surplus power to cars very worthwhile — much cheaper than contracting with dedicated energy storage facilities. The savings could be enough that they would pay to install low cost charging on curbsides and in parking lots, just for the chance to sell power through them that otherwise would have to be discarded.

Robocar charging

Another option will arise soon that completely rewrites the rules of charging infrastructure. In the 2020s, more and more cars will be released that have the ability to drive themselves. While everybody is waiting for cars that can take passengers on the roads at full speed, it is much simpler to make a car able to make short trips on uncrowded roads or in parking lots, particularly at night. While it is debatable if Tesla will soon release their “full” self-driving system, it’s quite reasonable that Tesla might produce a car able to move on its own for a 2 mile neighborhood trip at 2am, or within an employee or commuter parking lot. They might even produce a software upgrade that allows their current cars to do this safely.

Such a car would, when in need of power, book a time at a bulk charging station and travel there for it. If you owned such a car but did not have a parking space with charging, every few nights it would wander off to a charging station, probably some parking lot near a power substation. There, it would get plugged in by robots or even a human “plug jockey” and pick up a charge. It could also do this while parked at work during the day, to a charging depot not far away, or even in a corner of the employee parking lot. In time, the cars will get better and be able to go a bit further away at all times of the day, by keeping to quiet roads and going slow — they are not in a hurry.

For the owner, it’s the ideal situation. A car that’s always magically charged, without you having to do anything. That’s way better than a car where you have to find a gas station 1-2 times/week. These charging depots would probably use faster chargers, but they would get very high utilization.

Such a facility is even better for the application of receiving surplus solar power. When there is a surplus, the word can get out to all the cars, “drive over now and get some cheap surplus power.” The owner need only approve that the car is not needed for an hour, or that they don’t mind a 10 minute delay in summoning it for a surprise trip.

In this world — which could come well before any big infrastructure plan gets out of the planning stages — you don’t have to build much charging infrastructure at all. You put in in places where high-wattage power is already cheap to obtain. The cars work like magic, only needing charging on a long highway road trip. (Even then, you’ll just stop at any restaurant and while you eat, the car will wander off and fill up.)

In time, cars may also switch to a new charging plug interface, one located underneath the car so that the car (which is a robot, after all) can do almost all the work of plugging itself in, making these stations simple.