Smart charging solution implemented in Vilnius: urban districts are opening up for EV charging

The growing demand for power for EV charging in urban neighborhoods can be addressed through smart grid balancing based on artificial intelligence and machine learning. With this solution, the power of the charging stations is balanced based on real-time network load data. Drivers of electric vehicles in Vilnius' Pilait? district have already tested this technical solution, implemented by Lithuanian national DSO (Energijos Skirstymo Operatorius (ESO) and Inbalance.

The broader application of the innovative solution in Lithuania will satisfy the growing power demand and allow the installation of charging stations where it would not be possible under conventional methods due to grid limitations. It will save funds for the development and reconstruction of the electricity network and ensure more efficient use of the existing electricity network capacity.

Due to the uptake of electric vehicles in Lithuania, the demand for power for charging is estimated to increase several times by 2030. The country will need an estimated hundreds of millions to meet this need to expand and reconstruct the electricity network. According to Inbalance calculations, smart balancing can reduce the scope of investments required for network development and reconstruction by 80 percent.

"During the pilot project, we tested a technical solution that can be applied throughout Lithuania, especially where there may be a lack of available power and a large investment in the reconstruction of the electricity network would be required. This opens up the possibility of deploying charging stations in densely populated areas while maintaining the stability of the electricity grid. With the help of smart grid balancing, it is possible to achieve the result without additional investment - to charge an electric car within a reasonable time using the grid capacity efficiently," says Ovidijus Martinonis, Head of Network Development at ESO.

Special equipment was installed in the substation in Pilait?, which monitors its load parameters in real-time and transmits them to the system. The software uses artificial intelligence and machine learning algorithms to process data, develop charging behavior models, and predict low and peak network consumption. According to the models and forecasts, the software controls charging power in real-time: if the consumption load is high and approaches its limit, the charging power is temporarily reduced. If the consumption is low, the charging can be speeded up.

"Thanks to the technologies used, the microgrid - the distribution network chain that includes the final substation and electricity consumer inputs - can be balanced intelligently. Most importantly, the user does not experience any lack of power - no restrictions on consumption or charging stations. Such balancing solutions significantly reduce the investments required to develop and reconstruct the network," says Nerijus ?iaulys, Chief Technology Officer of Inbalance, a company that manufactures and manages intelligent charging infrastructure.

According to N. ?iaulys, implementing this advanced technology also justifies the primary goal of the transition to EVs - to protect the environment. "Reducing the need for reconstruction and network expansion, in other words, not doing unnecessary work, also means leaving a smaller footprint and consuming fewer resources, which is necessary to ensure that the development of the EV charging network is sustainable. The financial savings could be channeled to other areas, further creating added value for the environment," he said.

According to O. Martinonis, the pilot project results reveal that such a solution could already be used in multi-apartment parking lots, using the free input power of the buildings, in public charging stations, using free transformer power, and elsewhere.

The pilot project lasted about three months. The testing was implemented with Inbalance charging stations connected directly to the ESO-operated distribution network.