Algorithms provide faster and more efficient electric vehicle charging

Thursday, October 4, 2018

As electric cars become more common, charging large numbers of electric vehicles could cause problems with the stability of the local electricity grid. This could be avoided by utilizing smart control charging, wherein the charging power is altered over time by some algorithm. It is also possible to use the grid more effectively by using smart charging. With the right algorithms for charging, researchers want to ensure that the power grid can be kept in balance.

Joakim Munkhammar is an associate professor, research leader of the Electric Transport Group, Uppsala University and involved in Swedish Electromobility Centre. He and his research colleagues are working to find the optimum models for smart charging of electric vehicles.

“Smart” in this case means charging is controlled via algorithms, to achieve a variation in the otherwise expected charge pattern. The advantage is that smart electric car charging could minimize the risk of overloading the local power grid, while during low load times, the charging model will be able to utilize more capacity for charging, compared to standard charging without control.

“Conventional uncontrolled electric car charging can lead to problems with the local power grid. Alternatively, it may be suboptimal not to use remaining resources in the local power grid for electric car charging. This research project is based on developing models or algorithms for smart electric car charging that will allow the charging to avoid local problems in the grid, while at other times they can charge cars faster than with uncontrolled charging”, says Joakim Munkhammar.

The major challenge in the field is to create algorithms that work well and are robust, and that connect directly to existing industry technology. Therefore, the research is closely linked to both the energy and automotive industry, in the form of Vattenfall RnD and CEVT.

“The idea is to link this closer to the industry than before, and connect to CEVT and Vattenfall RnD by creating algorithms that can be implemented in existing technology. The way forward is to create models or algorithms for this, validate them against data and try to implement in actual systems.”

The hope is that the smart charge algorithms developed by the researchers will be used directly in the automotive industry. In this particular study, focus is on passenger cars and home charging, but Joakim Munkhammar points out that the algorithms can also prove useful for other vehicle types.

What are the results from the research so far?

“In a study, presented at the E-mobility Symposium in Stockholm in October, we have found that the amount of problems with the local power grid expected to occur with home charging is minimal. Therefore, the need for smart charging here is largely non-existent. The results, which are preliminary, are calculated based on electricity consumption without heating and are a comparison between electricity consumption from households with or without electric car charging and a given fuse level. Later, we will calculate what happens with the power grid. It may be particularly interesting on block level or city level.”

“We also found that electric car charging could be done faster for most of the year’s days, for virtually all simulated cases with different installed charging power.”

The results can consequently be of importance to the individual car owner, but also in a broader perspective.

“Smart electric car charging is designed to utilize resources more efficiently, in this case the local power grid. Therefore, all parties could win by implementing such algorithms. For society, range anxiety for electric car drivers can decrease by shorter charging time and for electricity grid providers the stability of the local grid could be enhanced, and it could also help to keep the local power grid in balance, even with for example local solar production, which could otherwise create problems.”

What are the needs of further research?

Aside from neighborhood and city level of home-charging public and workplace charging are the next challenges in the research. Then, research on the power grid’s impact from electric car charging is required, and the combination of electric car charging and intermittent sources of energy, such as wind or solar power, is important to determine how the algorithms are to be designed. Furthermore, it is a challenge to create transports for future smart cities, where this is only the first step”, concludes Joakim Munkhammar.

/ Daniel Karlsson