Europe's distribution grids are the bottleneck of the energy transition. Electricity demand is set to rise about 60 percent by 2030, 40 percent of Europe's distribution grids are over 40 years old, and the European Commission's Grid Action Plan puts 584 billion euros of grid investment on the table this decade, much of it in distribution. But reinforcement takes years, and EVs, heat pumps, and rooftop solar are landing on local feeders now.The gap between when the load arrives and when the grid can be reinforced is a forecasting problem. If a distribution operator can see where and when a feeder or transformer will be overloaded, it can manage congestion, dispatch flexibility, and target reinforcement where it pays, instead of reinforcing blind.
The problem: load arrives faster than the grid can be built
Distribution congestion is local. It shows up at a specific feeder or transformer when EV charging, heat-pump heating, and solar feed-in cluster in the same neighborhood at the same hour. The data that would reveal it, smart-meter consumption, sits with the operator under GDPR and cannot be freely centralized. So most operators plan on coarse averages and find overloads after they happen.
- Demand rising about 60 percent by 2030, concentrated in electrification
- 40 percent of distribution grids over 40 years old
- Reinforcement takes 4 to 10 years, connections take 2 to 3
- Smart-meter data is personal data and cannot move freely
The result is blind reinforcement, avoidable curtailment, and overloads caught too late.
The approach: forecast capacity at the feeder, where the data lives
LST-E is a specialized forecasting model, an LSTM trained on smart-meter time series with weekday and time-of-day features, that predicts load at the feeder and transformer level. It is part of Helios, our purpose-built energy control plane, which runs on Lascaris, the sovereign decision fabric. Every forecast, and every action taken on it, is policy-checked and written to an audit trail a regulator can reconstruct.
It runs federated, inside the operator, so consumer meter data never leaves the grid. The forecast moves, the data stays.
What it predicts
- Where and when feeders and transformers will exceed capacity
- Available headroom by location, so connection requests go where the grid can take them
- The impact of EV and heat-pump clustering before it bites
- Where flexibility or storage defers reinforcement, and where reinforcement is unavoidable
How it works
- Connect, not migrate. LST-E runs against smart-meter, SCADA, and weather data in place.
- Forecast locally. The model trains and predicts inside the operator's boundary. Only the forecast leaves.
- See the overload coming. Capacity forecasts flag feeders and transformers at risk, by time and location.
- Act on the record. Lascaris policy-checks and logs every dispatch, curtailment, or connection decision.
- Scale by feeder. Start with one substation, extend across the network without re-architecting.
Why this matters for distribution operators
- Defer reinforcement: target the 584 billion euro buildout where it pays, not by guesswork.
- Cut curtailment: visibility of headroom directs connections to where the grid can take them.
- Stay compliant: meter data stays with the operator under GDPR, and energy is a critical sector under NIS2 and CER.
- Hold up to scrutiny: every forecast and decision is auditable.
Why Scalytics
Built by utility veterans. Our founding team led digital transformation at E.ON, architecting IoT and grid systems across connected energy assets. LST-E and the federated approach come from that work, not from a whiteboard. LST-E connects to existing meter, SCADA, and historian systems, federated by design and on the record by default.
For system-level demand and balancing, see Demand Management and Grid Intelligence. For grid and OT security under NIS2 and CER, see Federated Cybersecurity Analytics.
Sources
- EU Grid Action Plan, COM/2023/757 (2023): https://energy.ec.europa.eu/topics/infrastructure/european-grids_en
- ACER, grid congestion and curtailment cost in the EU: https://acer.europa.eu/monitoring/MMR/crosszonal_electricity_trade_capacities_2024
- LST-E, federated energy forecasting model: https://github.com/scalytics/LSTM-NNW/tree/main/LST-E
