Semantic web promises a smarter electricity grid

(PhysOrg.com) -- Dispersed wind farms and solar panels on people’s homes are posing new challenges for managing power grids that were designed when all electricity was generated in centralised plants. A new semantic web technology promises a solution.

Managing increasingly dispersed energy resources - from far-flung wind farms on remote mountaintops to racks of solar panels on suburban homes - is crucial if the electricity they produce is to be used as efficiently as possible.

Today, wind farms, solar and other renewable resources generate around 7 percent of the European Union’s electricity, a figure that is set to rise to 20 percent by 2020. But unlike conventional fossil-fuel burning or nuclear power plants, few of these dispersed energy generators feed information into grid management systems.

So grid operators can’t be sure what generators are connected, whether they are in operation and, if so, how much power they are producing. This could become a bigger problem, once these power sources start to account for a larger portion of Europe’s energy supply.

“Grid management systems were designed around a large number of clients and a few suppliers, but now there are increasing numbers of suppliers. This requires a new level of communication and management system,” explains Bernhard Schowe-von der Brelie, a researcher at the FGH research institute in Mannheim, Germany.

Schowe-von der Brelie led a team of researchers who have made considerable progress towards developing a solution, not only for managing distributed resources in an electricity grid, but also for communicating with autonomous systems, devices and sensors across any network.

Self-describing network

Working in the EU-funded S-TEN project, the team developed a generic framework for novel ICT architectures and applied semantic web technologies to make networks ‘self-describing’ so that each component - be it a volt meter on a wind turbine or a thermometer on a weather station - autonomously publishes information about what it is, where it is, and what it does.

Because semantic data can be understood by machines as well as humans, the approach should lead to more efficient automated grid management and better decision-support for human operators. Smart power grids, efficiently supplying a town or city from locally generated electricity and then feeding it into a wider supply network, could therefore be more easily and cost-effectively set up.