Article: Thursday, 30 June 2016
Fully charging an electric vehicle can use more energy than the average household consumes in a whole day. So when drivers come home from work and all ‘plug in’ to recharge at the same time of day, the demand for power could destabilise the power grid, says Konstantina Valogianni of Rotterdam School of Management, Erasmus University (RSM). In her PhD thesis, she describes how self-learning algorithms can reduce peak demands on the grid and bring down the costs of electricity for consumers. She defends her thesis on 30 June.
Electric vehicles (EVs) make transport more sustainable, but they come with a potential downside if not managed properly, says researcher Valogianni. The power grid and power plants designed to supply the daily needs of households actually struggle to deliver all the electricity needed to recharge so many cars at the end of the working day – especially when they all plug in at 18:00. If this happens, the demand for electricity soars and regular power plants don’t have the capacity to fulfil it.
Additional power plants are fired up to provide this extra power – a polluting and expensive operation that drives up the market price for electricity. If EV ownership grows as expected, the number of EVs requiring evening recharging could make peaks of electricity demand unmanageable, with potential for blackouts.
Valogianni says that peak loads can be smoothed out by better managing how EVs charge, for example by recharging them during the troughs of low demand, or by not recharging more than is needed for normal use. Another method is to let stationary EVs sell stored energy back to the power grid in periods of high demand.
There are essentially two ways to manage EV charging, she continues. The manager of the central grid can manage demand in a top-down manner by adjusting electricity prices to which the market responds by buying more or less electricity. On the consumer side, on-board computers in EVs can help the driver to decide the best time to recharge or sell energy back. In her research, Valogianni designed and tested algorithms that support both approaches. She found that combining the two approaches holds a lot of potential for the future.
In one of her studies, Valogianni developed an algorithm that runs as intelligent software and can make charging decisions in the car. In the most optimistic scenario with 1.000.000 all-electric cars on the road, Valogianni's real-world test data suggest that using the algorithm would reduce the market price for electricity by 55 per cent on average.
Valogianni’s software constantly learns about the driver’s preferences by observing how the EV is driven, when it is charged and how much electricity the household uses. The software relates this to fluctuations in energy prices during the day and then decides if it’s the right time to recharge, or to sell back surplus electricity to the grid, Valogianni says.
The intelligent software can even decide to not fully charge the EV if it expects the owner to drive only a short commute in the morning. This way the algorithm finds a balance between ensuring the battery contains enough energy, while lowering the cost of charging. Valogianni stresses that it is a flexible system. The user can always set the EV to override any decision and do a full recharge if there’s going to be a long drive the next day.
Another way of reducing peak loads caused by EV charging is to let the central grid manager determine the best time for recharging to happen. This means some customers would have to wait for less busy times before recharging their EVs. The challenge here, Valogianni says, is to keep everyone happy by not allowing delays in charging to become uncomfortable for consumers, while at the same time successfully reducing stress on the power grid.
Valogianni found that an almost-real-time energy auction can help to solve this problem. In such a system, energy providers would communicate their current energy prices to customers with the help of a mobile app. By raising prices at times of high demand, and lowering them at times of lower demand, they can ‘nudge’ their customers to choose the best time to charge their EVs. After modelling and a field test with the app, Valogianni found this system has the potential to be fair to all customers and keep their EVs charged, while reducing peak loads on the power system.
Valogianni says that the algorithms she developed for consumers and for grid managers would produce even better results in combination. Such a hybrid system would learn individual consumer preferences and give grid managers enough influence over demand to reduce unwanted peak loads.
Valogianni, K. (2016, June 30). Sustainable Electric Vehicle Management using Coordinated Machine Learning (No. EPS-2016-387-LIS). ERIM Ph.D. Series Research in Management. Erasmus University Rotterdam.
Science Communication and Media Officer
Corporate Communications & PR Manager
Rotterdam School of Management, Erasmus University (RSM) is one of Europe’s top-ranked business schools. RSM provides ground-breaking research and education furthering excellence in all aspects of management and is based in the international port city of Rotterdam – a vital nexus of business, logistics and trade. RSM’s primary focus is on developing business leaders with international careers who can become a force for positive change by carrying their innovative mindset into a sustainable future. Our first-class range of bachelor, master, MBA, PhD and executive programmes encourage them to become to become critical, creative, caring and collaborative thinkers and doers.
