The capacity of the grid must rise to integrate volatile renewable sources into the energy supply. By using existing lines better depending on weather conditions, the need for new lines can be reduced. For this purpose, Karlsruhe Institute of Technology (KIT) researchers are developing self-learning sensor networks to model weather cooling based on real data. The power transmission of the line can be improved in favorable conditions.
The rapid expansion of the use of renewables–northern wind power, southern photovoltaic power–and the growing international trade in energy are leading to increasing energy transmission grid requirements. The current grid infrastructure must be expanded considerably to carry power from producers to consumers, stop a temporary shutdown of power-generating plants from regenerative sources, especially at a high wind intensity and make sure high security of supply in general. This is linked to long-term licensing and high costs.
By better utilizing existing overhead lines, the need for new transmission lines can be substantially reduced. This can significantly increase energy transport depending on the weather, for example, atmospheric temperature, solar radiation, wind speed, and wind direction. This rise can be accomplished without going beyond the maximum permissible temperature of the conductor and without the distance from the ground.
Within PrognoNetz researching and business partners will establish broad sensor networks with smart sensors, which are located near to each other and close to overhead lines to evaluate weather conditions accurately, unlike conventional weather stations. The sensor networks are tough ambient conditions resistant and provide critical data wirelessly to the control center. The new algorithms will enable the sensors to learn themselves. Based on the data measured, accurate power load predictions for hours or even days are generated automatically. ITIV team is working on forecast models based on artificial intelligence and a laser-based wind sensor whose precise measurement is greater than conventional rigidly mounted sensors. Furthermore, unmanned drones are used to install and maintain the weather sensors on the power stations.
The meteorological self-learning network to be established within PrognoNetz will immediately be used for current TransnetBW high-voltage lines and machinery. To assure the best use of existing power grids, this AI-based network adapts operation at all times to the weather conditions. This means that under favorable conditions, power transport can be increased by 15 percent to 30 percent.