Why the smart charging of electric vehicles is not like your average prisoner’s dilemma
Most readers will be familiar with the dilemma that asks whether two prisoners should betray each other or remain silent to their investigators.
Most readers will be familiar with the dilemma that asks whether two prisoners should betray each other or remain silent to their investigators. Although the total penalty is minimized when both prisoners remain silent, it is in their individual interest to betray the other. The strategy for each prisoner of betraying each other is a Nash equilibrium, an important concept from the game theory that concerns itself with the mathematics of such dilemmas. Given the simple rules of this game, both prisoners could only do worse for themselves in another strategy.
A less fictitious dilemma: the energy transition and climate change
A less straightforward example of how game theory applies to everyday life is the issue of man-made climate change. An often heard argument in the West is that it will take us a dazzling amount of money to achieve a significant reduction in our country’s CO2emissions, while we’re not sure it will be enough and in time. And is there any point in doing anything costly if other, far bigger countries just continue polluting the same space? It seems to be in every country’s individual interest to let another part of the world solve climate change. Again, betrayal appears to be our Nash equilibrium, albeit a rather deadly one for the parts of our civilization with the fewest resources and the most vulnerable climates.
With governments stuck in game theory, it seems about time to let the markets solve the equation. After all, if we change the penalties and rewards of the game, we can shift the Nash equilibrium in our advantage. That’s exactly what’s happening in the Netherlands for a small piece of the puzzle.
The rise of electric vehicles and the electric infrastructure market
Faced with the inevitable adoption of electric vehicles, a group of grid operators, consisting of several parties responsible for Dutch high and low voltage grids, saw it coming that the Dutch electricity grid in its current state could not handle a majority adopting Tesla’s and Nissan Leafs. Their first dilemma: act or react? They acted. Then came a second dilemma: should every operator, large or small, try to solve the puzzle for itself with proprietary knowledge, or should they work together as one with open systems?
Dutch grid operators are fully separated from energy providers and while energy providers have to compete with each other, grid operators enjoy local monopolies and do not compete with anyone. That probably made it easier to cooperate. Together they founded Elaad (which would translate to E charge) and from that non-profit organization, open protocols found their way into the world. Electric vehicle chargers now talk to IoT platforms and these platforms talk to each other. This has led to cross-border interoperable systems that use protocols (OCPP, OCPI, OSCP) that were originally developed in the Netherlands.
By working together and sharing knowledge, the Netherlands have developed a leading position in knowledge and implementation of smart charging infrastructure. True, other countries are more successful in adopting (Norway) or producing (China) electric vehicles, but the years of experience and very scalable knowledge that is concentrated in the Netherlands already prove to be very valuable. Dutch companies (GreenFlux, NewMotion, Alfen, Eneco eMobility, PitPoint, EVBox), focusing on electric vehicle charging and integrating this into a smart grid, can create value thanks to earlier investments by the Dutch governments and said grid operators. Remember the flop of the plug-in hybrid Mitshubishi Outlander that was heavily subsidized in the Netherlands and the UK? Yes, they did little (or better: no) good for the environment, but they needed chargers and we provided them.
We have chargers but do we have capacity?
We now have the highest density of charging stations in the world, here in the Netherlands. After getting everyone on board with open protocols, the next challenge revealed itself soon. How are we going to supply all these chargers with electricity, with a grid that is arguably very stable, but wasn’t built for our all-electric ambitions?
The problem that we face now is not that the grid does not have the capacity to carry the additional energy necessary to charge electric vehicles. The real problem is that this extra demand is concentrated at times where we already saw peak demand without a single electric vehicle. At dinner time, when the vehicles are plugged into the net and the potatoes are put on the induction cooker, the grid will bend and break. That is, if we do not anticipate.
This sudden demand does not just play on a national level, but also very locally. Buildings have scarcely been adapted for future adoption of charging infrastructure and new energy economies. Expanding the local grid capacity often requires expensive cables, larger transformer stations and hence capital expenditures running in the millions for larger buildings.
The availability problem: my kingdom for a charger
This is what we call the availability problem. We want more electric vehicles on the road and they need to be charged. The perfect charge locations are where the vehicles are deliberately idle – not along the high road but at home or at the office. Providing these locations with enough charging capacity is costly, unless smart solutions are developed for it. This requires some creativity and that is what we see all around us in the market. Ideas have come from many corners and led to interesting projects and startups.
Solutions for the availability problem can be divided in two categories. The first class aims to optimize the usage of existing infrastructure by reducing idle time. Paradoxically, the second embraces idle time and the flexibility it generates.
Many solutions lean more on the social aspect of the problem. Whatsapp groups have been established to request charging opportunities; a feature that was later integrated with public charger information into dedicated apps. A disadvantage of social solutions over (partly) technical solutions, is the dependency on other users. This relation is just fine if the other users are your neighbor or colleague, but when they are unknown to you and not familiar with the same app, some of the effect will be lost. Will some colleague you barely know and who works on the 7thfloor of your office interrupt a meeting, walk seven stairs, unplug his EV, find a new, scarce parking spot and head back, so that you can charge? Here we arrive at game theory again, with a high reward for the colleague who found a charging spot. The Nash equilibrium will favor the early bird.
Besides social initiatives, another solution in the idle time reducing class are parking tariffs. As soon as an electric car no longer consumes power because its battery is fully charged, the charger notices this, informs the connected platform and the tariff per minute is increased. This is no punishment for cars that do not charge as fast as some higher-end vehicles, but it is a penalty to those occupying costly infrastructure without consuming it. As a consequence, it merges parking and charging facilities into one service. This creates one less transaction, which is a seemingly desirable simplification. It is however not always easy to understand when, where and how these tariffs come to place, not even with the rise of great smartphone apps. Some parking lots will have those services combined, but others will not, which will inevitably lead to parking tickets for confused users. On top of that, having to park your car somewhere else to prevent ramping tariffs is not very user-friendly!
Harvesting flexibility with smart charging
The second class of solutions to the availability problem is part of a far bigger trend: the Internet of Things (IoT), or more specifically the Internet of Energy (IoE). Armed with highly scalable and blazingly fast cloud infrastructure, smart data and tools to exchange information with users, grid operators, smart meters and energy markets, the Internet of Energy is here to stay.
When idle time, or better, flexibility is considered an asset, smart charging enters the stage. Smart charging’s primary application is currently found in smaller systems with one problem owner: a building’s facility manager, who is often charged with finding a cheap way to solve the availability problem for that building. Smart charging is a broad term that is somewhat inflated, but it often means load balancing: distributing charging capacity to multiple vehicles in a smart way, such that the total available capacity for charging is never exceeded. When expanding the bottleneck by increasing the building’s grid capacity is too costly, smart charging can still increase the number of vehicles that charge simultaneously.
Cloud-based smart charging systems can upgrade a local system’s capacity to charge up to ten times as many vehicles during the day, without requiring drivers to walk down the stairs of their office to do a well-timed parking lot swap with someone they’ve never met. It is important to note that smart charging effectively accelerates charging. This effect becomes more obvious when the power capacity for charging is dynamically adjusted during the day. When charging is allowed to accelerate when a building uses less power for other purposes, impressive performance gains are possible, nearly doubling the average charge rate per EV while still respecting the local grid constraints.
Since GreenFlux came to life in 2011, we have applied smart charging systems on an international scale. From this, we learned that electric vehicle drivers are very open to smart charging. It allows them to lower their charging costs, for example by charging only when they get reduced electricity tariffs. Smart charging at their office will greatly enhance their charging options. Quite importantly though, they prefer the insurance of being able to check on their charging, or even control it. GreenFlux’s Charge Assist App makes this possible.
Power to the user, through a simple but effective app
By giving drivers an app that displays their real-time charging information, range anxiety is soon a concept from the past. Integrating solar, wind and home batteries (which will just be second-hand car batteries) with a smart IoE creates a powerful system. A system that is controlled by the electric vehicle driver.
By bundling IoT, cloud computing and smart human interaction design, we offer drivers the chance to control the charging of their vehicle in the simplest way possible. We provide them with the opportunity to obtain a free charger, simply because our cloud-based smart charging can handle ten times as many charging sessions at the same time. When they return to their cars after a day at the office, their battery is fully charged – which they could have already known from the Charge Assist app on their phone.
Now you could argue that this is all nice if your behavior is predictable, but when you need to leave for a lunch meeting, how do we know you need your car fully charged a bit earlier? Truth is we don’t, since we don’t believe in shaky connections to your Outlook calendar or constantly asking you questions about your car’s battery. We found something better though..
How our Smart Charge Assist changes the equilibrium
Here we finally circle back to the prisoner’s dilemma that turns out not to be such a big dilemma this time. We provided users of the Charge Assist app with one simple button, that simply informs our smart charging cloud of the user’s need for the fastest charge we can possibly give. And we will take care of that request. Yet – and we guarantee this! – we still refuse to blow up any circuit breakers. Now, since using this button is free and unrestricted, would not every user be better off by always pressing that button?
That’s the question we asked ourselves during the largest Smart Charging trial ever in the worldand it turns out that good old John Nash’s equilibrium lies not where you might expect it. If you know your vehicle will be fully charged when you’ll need it, why go through the trouble of pressing a button? The hundreds of users in the trial had all used the button to try it out, but after that they pressed it only in 5% of their charge sessions when they suspected they might need a quicker charge. The button was their insurance, but most likely they would be alright either way.
It turns out that, with Smart Charging, optimizing for the total also optimizes for the individuals. There are no prisoners.
GreenFlux offers its Smart Charge Assist solution with proven smart charging technology and the free Charge Assist app as part of its Microsoft Azure IoT electric vehicle charging platform. Visit www.greenflux.com for a full overview of our solutions for charge network operators and electric mobility service providers.
Article by Nico Spoelstra
 As an interesting example of how this new development is explored, Elaad researches the IOTA tangle (not to be confused with, but very similar to blockchain) to control charging with micropayments.
 Although grid operators have been researching algorithms to control power intake by cars on regional or national levels, this is far more challenging than the building-case because of regulatory reasons.
Of course this is already much aided by fast charging stations that are becoming more ubiquitous and incredibly powerful.
 This can be further increased by our OSCP-based connections with solar and battery systems.
 Unfortunately, car OEMs are not so fond of open protocols and do not tell the charger’s connected platform about their state of charge. The proprietary protocol ISO15118 will help here, to some extent.