Strategies for Smart Charging of Electric Vehicles
MetadataShow full item record
The electrification of the transport sector is put on the agenda as an important means to reduce the greenhouse gas emissions in Norway, EU and other parts of the world. Chargeable vehicles have the potential to reduce the greenhouse gas emissions both because they are four to five times more energy effective than today s petrol-powered vehicles, and because the electricity they use can come from renewable production. Questions regarding sufficient access of conventional oil and increasing challenges with local emissions will also contribute in the electrification of the transport sector .The electrification of the transport sector brings with it some challenges. The charging of the electric vehicles leads to an increase in demand. If no sort of smart control of the charging is put in place, it could be safe to assume that many vehicle owners would plug-in their vehicles when they come home from their last journey of the day, leading to an increase in power demand at the time when there already might be a peak in demand. This charging regime is often referred to as dumb charging .This thesis focuses on smart strategies for charging of electric vehicles and presents two smart strategies for charging of the electric vehicles, the profit maximization scenario and the power factor control scenario. In these scenarios it is assumed that a control system is put in place so that specific chargers can be instructed to change the power factor and to begin and stop charging. The technical specifications of this control system has not been studied, it is simply assumed that such a control system is put in place. In the profit maximization scenario, it is assumed that technology facilitating the electric vehicles to discharge energy back to the grid is put in place. The electric vehicles are then instructed to charge when the electricity prices are low and discharge back to the grid when the electricity prices are high. Since the demand and electricity price pattern follows a similar pattern (high prices when the demand is high, and low prices when the demand is low), this procedure should have a smoothing effect on the demand pattern with the right choice of parameters.In the power factor control scenario, all electric vehicles are assumed to be plugged-in and start charging as soon as they return from the last journey of the day. When there are voltages in the grid below a certain value, electric vehicles charging at that time is instructed to reduce the charging power factor. By reducing the power factor, reactive power can be injected into the grid.To demonstrate the principles, algorithms have been developed and implemented in MATLAB for both the dumb charging scenario and the two smart charging scenarios. A model of the IEEE 13 node test feeder has been modelled in MATPOWER, and power flow simulations of the three charging scenarios have been run, assuming a vehicle adoption of 50%. The algorithms success has then been rated according to their capability to avoid low voltages on the demand nodes in this test case. For the purpose of this thesis it is assumed that the designed model is a distribution network in Norway. Therefore the base demand pattern used in the simulations follows the actual base demand pattern for a chosen 24-hour period in Norway.The results from the simulations showed that the when all vehicles were charged as soon as they return from their last journey of the day, it resulted in a major peak in demand, and thus a reduction of the voltage between 18:00 and 21:00.Both smart strategies presented in this thesis improved the voltage profile. While the lowest node voltage in the simulations of the dumb charging scenario was 0.915 pu, the lowest node voltage in the profit maximization scenario was 0.932, and the lowest voltage in the power factor control scenario was 0.934 pu.In the profit maximization scenario, the demand profile, and thus the voltage profile, did however not smooth out as much as expected. Since the electricity prices changes when there is a change in demand, the algorithm is designed to find the charging schedule for one vehicle at the time, changing the electricity prices in between. A further study on this scenario should focus on finding a better relationship between changes in demand and electricity prices.In the power factor control scenario, the minimum voltage allowed was not avoided at all nodes in the designed test case. The simulations showed that one of the nodes experienced a voltage violation. This scenario did however only assume that the vehicles were only available for grid services the first eight hours after the return of the last journey of the day. If the vehicles were available at other times too, this voltage violation could however been avoided.