Jacques Mattheij

Technology, Coding and Business

The Problem With Electric Vehicles

Electric Vehicles (EVs) are all the rage these days. They’re wicked fast off the mark, they are nearly silent at low speeds and they are perceived as ‘green’, what better way to make yourself feel good than to get an Electric Vehicle as your next car.

Now, as much as I like ‘greentech’ I see a whole pile of issues with EVs that won’t be easily wiped off the table using appeals to emotion (such as acceleration speed of ‘fun’ of driving) or a (somewhat misplaced) sense of improving the environment.

First of all, if you really want to improve the environment the best way to do so is to not buy any vehicle at all. Every car, electrical or not, even when it is never used is an enormous expenditure in both energy (and therefore greenhouse gas emissions) and raw materials simply because it has to be manufactured. So if you want to do the best possible thing for the environment cutting down on your traveling and working hard to avoid the purchase of a vehicle are the two best things that you could do when it comes to transportation.

But fine, you need a car, for whatever (presumably valid) reasons. I’m trying to imagine a world where all cars are electric and what kind of impact this would have on the way transportation is arranged right now in the world as we know it.

Some Background

For starters, lots of the infrastructure that we use to power our current transportation needs comes in the form of the global movement of various fractions of Petroleum, which is rather messy (it’s a liquid, and spills are not rare at all and the environmental impact of removing it from the Earth’s crust is huge) and impractical compared to say transporting energy through wires. The bad news (the messiness, the environmental damage and the fact that this requires the physical movement of vast amounts of material) is somewhat balanced by the good news: it can be stored for a long time without degradation if stabilized properly, it can be carried around to places where no other infrastructure besides roads exists, the energy density of the end-products as used in transportation (mostly: gasoline, diesel and liquified petroleum gas or LPG for short) is very high compared to existing battery technology.

In a world where all the cars and trucks are electric you’re going to have to roughly supply your average highway with infrastructure comparable to the energy consumption of the cars on that highway (or the cities around it). So a massive shift from one form of energy carrier (everything based on petroleum that ends up in the transportation sector) to electrical energy (which will require not only generating capacity in the form of nuclear, fossil fuels or renewables, but also the infrastructure to transport it) would be required. This translates into a massive increase in the carrying capacity of the power grid compared to where we are today. To give a rough estimate, according to this page the energy consumption of the United States transportation sector (which will not be typical for the world as a whole but it gives an indication) is approximately 28%. The vast majority of that comes on account of Petroleum based consumption, but of course it will also already include those EVs that have already been deployed. Regular ICE’s are terribly inefficient, as low as 18 to 20% on average. On paper, EVs do much better than this but in practice that advantage is somewhat dimished. There are for instance the charging losses to contend with, around 20% or so. Electric motors (the prime mover in an EV) are much better than ICEs in converting energy to motive power at 80% or better, so less energy would be required from the charger to the vehicle to achieve a similar distance traveled. But because it is the total system efficiency including generating losses at the power plant and transportation losses in the grid in the end that theoretical 80% ends up being much lower (if the power source at the generating plant is a gas turbine for instance the efficiency of the generator is in the very best case 60% but more realistic would be about 50% and further losses in the grid would be another 6.5% or thereabouts. Total system efficiency would then be .5 x .8 x .935 x .8 or roughly 29%. Still much better than the ICEs this all replaced but not quite as good as it looked initially. So 68% (20 (ICE) / 29 (EV)) of that original 28%, or 20% extra energy would have to be generated by the utilities and transported through the grid to the charging stations in order to accommodate all transportation to be electric.

The problems

Problem 1: Transportation will load the grid and generating capacity in rather nasty ways

Our current infrastructure has been created with several large factors driving it: industrial use tends to create a fairly steady baseline because companies have been given incentives to use electricity steadily (so in continuous processes) and at night rather than in irregular patterns during the day, household and office use adds a an element of larger variation but is a relatively small fraction of total energy use. The major generating capacity in most countries is provided by natural gas or coal burning plants and nuclear power plants. Load Following describes the process with which a power plant can increase or decrease its output power responding to increases in demand. If everybody starts charging their EVs this will cause the grid and generating capacity to be very heavily loaded during times when domestic power consumption is traditionally rather low, and in places where such quantities of power are not currently available. Your average town does not have the power infrastructure to deal with an extra draw of a whole bunch of commuters arriving home roughly around the same time (say, between 5:30 and 7 pm) and all of them plugging their cars in to recharge. The maximum power draw when recharging batteries is right at the beginning of the charge cycle and this will compress a whole pile of consumption into a relatively small amount of time.

Problem 2: Rapid charging is actually not so rapid, highway re-charging stations will have to be much larger than current gas stations

Because re-fueling using gasoline or diesel is incredibly rapid (typically: < 5 minutes from start to finish to get a full tank including payment) gas stations tend to be relatively compact. A typical gas station on the highway can have anywhere from 6 to 14 pumps and will process a hundred cars per hour or more without a problem. If the time to ‘refuel’ increases to an hour or so then you’d need much more space in order to allow all the vehicles to be on-site for that whole time, and you’d have to keep those people busy for that time as well. Personally I would not care for any extra delay during my travels, being forced to wait while my car recharges is not something that appeals to me, and I assume that people that are underway usually want to get to their destination rather than to be forced to take hour long breaks in order to get their vehicles ready for the next leg.

Problem 3: Gas stations are not generally in the neighbourhood of electricity generation stations.

The fuel that a fuel delivery truck burns and all the other energy expended to get a certain quantity of fuel delivered to a gas station is similar in nature to the losses in powerlines to charge EVs. They’re called ‘transportation losses’. These are for the grid as a whole at a rather acceptable 6.5%. But if we’re going to draw power in very large quantities (a few hundred super chargers in the same number of locations that we currently have gas stations at) those losses will likely go up due to the increased average distance between consumers and producers, making the total system efficiency of EV’s worse they would seem to be today. Also, even if this is deemed to be an acceptable solution we’d have to still make that power available, which will require fairly massive investments in infrastructure in order to accommodate the power draw. Every highway would be more or less automatically accompanied by a bunch of power infrastructure, not unlike the trolleybus systems, but with higher voltages and using periodic re-charging of batteries rather than a continuous contact between the vehicle and the power infrastructure (note that the Trolleybus system is rather clever and side-steps the charging loss issue). Of course we have lots of experience with building power grids and this is definitely something that could be done but it will require time to build it.

Problem 4: Re-charging will not work nearly as well when vehicle utilization goes up due to sharing

A typical family car will be parked in excess of 95% of the time, which leaves ample opportunity in an ‘ownership’ situation for re-charging. But as vehicle ownership shifts to sharing a vehicle between multiple users during the day vehicle utilization will increase (which is a good thing!), but opportunities to re-charge will be reduced and per-vehicle consumption of energy will increase. Overall there will be fewer vehicles but the number of passenger kilometers (or miles, if you wish) will be roughly the same, or might even go up when coupled with such emerging technologies as self-driving cars. So there will be substantially less opportunity for re-charging between rides.

Problem 5: We don’t actually have all this infrastructure yet

If a substantial chunk of our energy consumption due to transportation needs is going to shift from being directly petroleum based to being mostly based on the timely delivery of electrons in vast quantities and to a very large number of locations then we will have to invest massively in both generating capacity and grid capacity. In all the EV articles I read this fact seems to be wiped off the table as a footnote or it isn’t even mentioned at all. The shift from hydrocarbons to electrons for transportation will take many years of planning and building to accommodate, and even if EVs are ‘hot’ right now the total number of EVs sold is still relatively low (exceptions: Norway and the Netherlands). Now, NL is a really small country (about 100 x 200 km), so things like range due to limited battery capacity and cost of grid infrastructure are fairly low impact here, and in Norway they have vast amounts of hydropower, which explains some of why these two countries are ahead of the pack. In other countries that are not so fortunate it will take a substantial investment in power generating capacity and in grid infrastructure in order to make large scale deployment of EVs a reality.

Problem 6: Range

In small countries and around cities range is not really an issue. But with trucking and things like holiday trips and other longer distance travel a lack of range can become a real issue. A typical EV on the market today has a range of a few 100 miles at best and a re-charge time that is not compatible with charging-while-traveling. Gasoline storage density, universal availability and the speed in terms of extra range per minute of recharging/refueling translates into a massive convenience advantage for ICEs which EVs with present-day technology simply can not match.

Problem 7: Trailers

A very large number of the cars out there have trailer hooks, they’ll tow anything from trash to the local dump to other vehicles, horses, boats, caravans and so on. EVs as a rule are not set up for this kind of use and adding a trailer to an EV will dramatically impact the range (which is usually not exactly super to begin with). To compete successfully on all fronts with ICE based vehicles trailer hitches would have to be a factory option for EVs and adding one should not void your warranty or cause you to be operating your vehicle in an illegal manner. For now if having a trailer behind your car is a must you’ll have to be very careful when selecting one because chances that this is an option are slim.

Problem 8: Service

ICEs are relatively well understood but they’re finicky to maintain. Servicing an EV is in principle a lot simpler than servicing an ICE based vehicle simply because the systems are inherently less repairable so the accent will be on replacing modules rather than repairing them. A typical EV uses a DC circuit at a few 100 volts driving an inverter which in turn passes multi-phase power to the electric motor. Such a circuit, while in principle repairable is well outside the range of capabilities your average garage has, but it can be easy to troubleshoot the source of the problem (battery, power electronics, motor or wiring) and replacing the broken module with a refurbished one is something that a competent mechanic should be able to do. This does require a willingness of the manufacturers to open up their service channels, with the failure rate of some EV powertrains it is not at all imaginary that your EV could fail in a location where the brand of your choice has no representation.

Problem 9: Tax Breaks

In quite a few countries the owners of EVs get significant tax breaks, either when purchasing the car or during the lifetime of the car in the form of reduced ownership taxes. This is a good way to stimulate a shift and to get a larger pool of ‘early adopters’ but in the longer term this system will not survive. The reason for this is quite simple: governments will make any such measures income neutral for themselves and as the balance shifts from ICEs to EVs the EVs will have to carry a proportionally larger chunk of the tax burden. With some luck by then the economies of scale will have brought down the cost of batteries and EVs in general so that there will still be an economic incentive to switch to electrical.

Some Possible solutions

Not all of these problems have timely or easy solutions, some of the (especially the infrastructure and power generation issues) will likely take decades to fix, so I don’t see the world switch to EVs in massive quantities in the next few years, I think the change over will be very gradual with technological innovations driving each successive cycle of adoption (battery tech improvements would be one major driver of adoption).

One stop-gap hybrid solution for a number of these issues would be a gas station that has a small power generation station powered by diesel fuel. This would be an intermediate step where an absence of infrastructure would be overcome but it will likely lead to higher prices-per-charge than could be obtained through larger scale initiatives. Still, better some charge than none at all. An even smaller scale solution could come in the form of a ‘range extender trailer’ delivering power to the vehicle from a self contained unit which would be not much more than a fuel tank, an ICE and a small generator on two wheels attached to the trailer hitch of the car, however the above note on trailers applies to such a solution. It may also be interpreted as a way around certain tax rules that give the owners of EVs an advantage, if you then add a range-extender to your car you’ve just re-created a hybrid electric vehicle, which may be subject to completely different taxation.

In the longer term the infrastructure will catch up with the demand that all-out electrical transportation would create, battery technology is hopefully not yet at the end of its development and charge times may be further reduced. If EVs would break 800 km range in production vehicles at affordable prices all re-charging could be done overnight and a large part of the range-anxiety problem would simply go away. Taxi services, commuting, local transport and so on are all excellent use cases for EVs today, and going electric for those situations probably makes more sense than doing this with ICEs. Going longer distances a hybrid is right now probably the best you can do and it will be a while before EVs will be able to compete across the board with ICE’s. You’ll know when we have achieved parity because gas stations will start closing, or will replace pumps with EV super charging stations en-masse.

Another possible solution to some of the issues related to mass charging of vehicles would be that you set a timer to tell the car how long it has to charge, this could then be used to randomize the start-time of the charge cycle so that there won’t be a ‘thundering herd’ problem causing overload of the grid by starting the charge cycle of a large number of vehicles within a relatively short amount of time. This would also increase the ability of utilities to adapt their capacity to the demand because it would come on rather slower.

HN Submission/Discussion
If you read this far you should probably follow me on twitter: