Why local load balancing is vital for managing the grid in the new energy world

Posted on: October 6, 2019 by Hervé Barancourt

Wind. Solar. Tidal. Waves. When it comes to energy production from renewable sources, we’re often at the mercy of the forces of nature. Their intermittent energy production can be unpredictable.

In any energy network, balancing production and consumption is critical. But how do you do that with unpredictable energy sources? In this blog post, I explore a more efficient option for balancing the load on the energy grid in a world where the energy mix is becoming increasingly diverse.

The end of the centralized energy grid

Traditionally, the TSO (Transmission System Operator) is responsible for balancing the power at a high voltage level, managing centrally the energy produced by power stations sitting on the high-level grid. That all changes when energy production comes from renewable sources, particularly intermittent production sources.

More than 90% of intermittent energy production happens locally on the medium or low-level grid, the part managed by the DSO (Distribution System Operator). This energy is then fed into the TSO’s part of the grid, which regulation stipulates the TSO must balance.

But the latency that occurs when local energy from intermittent sources is fed into the high-level grid can make load-balancing difficult for the TSO. Moreover, studies have shown that when load is balanced centrally and more than 35 percent of the energy produced comes from intermittent energy sources, there is a significant risk of the grid becoming challenging to manage and even unstable.

Feeding energy produced at the edge of the grid into the central TSO grid can lead to significant technical losses – losses due to the resistance of equipment such as feeders, cables, overhead lines and transformers that arise on any electric network. The further the power has to travel, the greater those technical losses.

Local load balancing on a decentralized energy production

Latency and technical losses can be reduced by decreasing the distance power needs to travel. This is where we come to the idea of balancing production and consumption locally – or, simply put, ‘local load balancing.’ With local load balancing, energy production and demand is managed locally by the DSO, which means it no longer needs to travel very far from where it’s produced using the high voltage network.

To balance locally-produced energy with local energy consumption, DSOs need very good local forecasting of production and consumption, at the municipal level, for example. Machine learning will play a vital role, utilizing knowledge of the local weather, the topology of the local grid and end-users’ consumption profiles to forecast production potential and consumption in the near term and ensure correlation between demand and production.

Local load balancing also allows DSO to avoid the 35 percent limit for energy production from intermittent sources. Studies have shown that it will enable an increase in the use of intermittent energy sources without risking the grid becoming unmanageable or even unstable. Because it avoids latency, DSOs can react to changes to demand in pseudo-real-time.

Machine learning for balancing load at the edge

We have verified the concept of local load balancing in the SMAP smart grid demonstrator led by Enedis and Auvergne-Rhône-Alpes Energie Environnement. The SMAP project is built around France’s first photovoltaic village, the rural municipality of Les Haies in the Rhône valley. The village sits in the Pilat Regional Natural Park and has 810 inhabitants. Around 500 m2 of photovoltaic panels cover public and private roofs.

Together with several partners, we demonstrated how new infrastructure and communication technologies can enable the dynamic management DSOs need to balance intermittent local production with local demand. We deployed machine learning, for instance, to forecast production in intermittent energy sources accurately, turning a local electrical system based on an intermittent energy source into an energy system that is predictive, communicative and controllable.

The pilot shows a future of energy production enabled by information technology. For the TSO, it confirms the possibility of a well-balanced decentralized national grid where technology at the edge balances local production and consumption.

Local load balancing is a critical new concept that utilities will need to adopt to ensure the grid remains manageable when more than 35 percent of energy produced comes from local intermittent renewable sources. Read our Journey 2022 ‘Resolving Digital Dilemmas’ publication, researched and written by the Atos Scientific Community, to understand other emerging concepts that E&U will need to grasp to succeed in the new energy world.

Share this blog article

About Hervé Barancourt
Head of Smart Grid Strategy at Atos Worldgrid and member of the Scientific Community

Hervé joined Atos in 1990. He is responsible for the definition and implementation of Atos strategy for smart grids and smart metering. Hervé studied at the Centrale Lille Technical School and holds a master’s degree in engineering.
Follow or contact Hervé