Smart local control could save billions in grid expansion costs
The Netherlands risks spending more than €300 billion on expanding electricity grid - “Grid congestion is actually similar to traffic jams.”
Published on September 4, 2026
Bart, co-founder of Media52 and Professor of Journalism oversees IO+, events, and Laio. A journalist at heart, he keeps writing as many stories as possible.
The Dutch electricity grid does not only need to become stronger, but above all smarter, according to a new report by Zenmo Simulations. Local energy systems – known as ‘holons’ – could reduce peak loads, shorten connection queues and substantially lower the societal costs of grid congestion. The projected savings run into the billions, although the researchers stress that their calculations are not forecasts.
The Netherlands risks spending more than €300 billion on expanding its electricity grid over the next 25 years. At the same time, businesses, residential developments and renewable energy projects sometimes have to wait years for a connection. According to the report Holons – local flexibility as a solution to grid congestion, installing more cables and building more substations will not be enough. A considerable part of the problem could be solved by using existing grid capacity more intelligently.
The report was written by researcher Auke Hoekstra and produced by Zenmo Simulations on behalf of Energy Innovation NL and the Netherlands Enterprise Agency. Its central idea is that electric cars, heat pumps, solar panels and batteries should respond not only to national electricity prices, but also to the capacity available on the local grid.
“Grid congestion is actually similar to traffic jams,” the researchers write. “Traffic jams do not occur on all roads simultaneously, nor can they all be accurately predicted a year in advance: we need traffic reports for the electricity grid.”
Peak demand is the problem, not average consumption
That comparison gets to the heart of the issue. Grid operators must design their infrastructure around the highest peaks, even if these occur for only a few hours or quarter-hour periods each year. According to the report, fixed off-peak hours or time blocks therefore offer little help. They do not take into account local weather conditions, generation, consumption, or outages.
Dynamic grid tariffs could. If a transformer in a particular neighbourhood is expected to become overloaded the following day, the local tariff would temporarily rise. Cars would charge later, heat pumps would warm homes earlier, and batteries would feed electricity back into the grid. Tariffs would remain unchanged elsewhere, where sufficient capacity was still available.
The distinction between time and place is essential. Low national electricity prices can actually aggravate local congestion. When strong winds make electricity cheap, batteries and electric cars may all start charging simultaneously. That may benefit the energy market, but not necessarily the local grid.
What is a holon?
The researchers take the idea one step further. They propose an electricity system composed of autonomous but interconnected units known as holons. A home or business can be a holon, but so can a neighbourhood, business park, municipality or province. Together, these layers form a ‘holarchy’.
Each holon first tries to balance demand, generation and storage internally. Electricity is exchanged with a higher level only when necessary. A home battery would therefore support the household first, while batteries and charging points across a neighbourhood could subsequently respond collectively to the local grid's capacity.
The report compares this structure to the internet. Information there is not controlled from one central location either. Routers make local decisions according to shared protocols, allowing the system to continue operating when one component fails.
According to the authors, a similar model could make the electricity grid not only more efficient but also more resilient. During a major outage, local areas with their own generation and storage could continue operating temporarily. Citizens and businesses would also gain more control over their energy supply.
Three future scenarios
Zenmo modelled three scenarios. Under NoFlex, appliances do not respond flexibly to prices or grid loads. Under MyFlex, they respond to energy prices but not to local congestion. According to the researchers, this broadly reflects the direction in which the Netherlands is currently heading.
Under HolonFlex, appliances take account of both energy prices and local grid capacity. Users might specify when an electric car must be fully charged or which indoor temperatures are acceptable, after which the system would operate automatically.
The difference is substantial. In the simulation, peak demand per detached or semi-detached home falls from 3.5 kilowatts under NoFlex and 2.7 kilowatts under MyFlex to 1.5 kilowatts under HolonFlex. According to the report, this would allow more than twice as many homes to be accommodated within the same grid capacity.
For the Netherlands as a whole, HolonFlex results in a 23.2% lower peak load than MyFlex. Over the period from 2020 to 2050, the model calculates total societal costs of €1.24 trillion under NoFlex, €612 billion under MyFlex and €421 billion under HolonFlex. HolonFlex therefore saves approximately €200 billion compared with a system that responds solely to energy prices.
Not a blueprint or forecast
These figures should be interpreted cautiously. The report itself describes the simulation as pioneering work. Its results depend heavily on assumptions about the adoption of electric vehicles, batteries and heat pumps, the rate at which users participate and the societal costs of delayed grid connections.
Remarkably, Zenmo also calculates significantly lower costs for physical grid expansion than the grid operators themselves. The report cannot explain this difference based on publicly available data. This underlines both the uncertainty surrounding the results and the need for greater transparency.
The study does not claim that grid expansion will become unnecessary. It does, however, show that flexibility should no longer be treated as a supplementary measure. According to the simulation, adding one kilowatt of grid capacity for a household costs around ten times as much as avoiding the need for that capacity through smart HolonFlex control.
Technology is not the biggest obstacle
The main barriers are institutional. Grid operators must share more detailed information about local grid loads. The Netherlands Authority for Consumers and Markets must enable congestion incentives that vary by time and location. The Netherlands must also make better use of European frameworks for local energy communities and Congestion Service Providers.
The report therefore presents seven recommendations, including prioritising the procurement of flexibility over grid expansion, providing better support for business parks and exploring the creation of an independent Grid Transition Institute.
The real message behind the multibillion-euro calculations is therefore a governance one: the Netherlands still treats grid capacity too much like a static product. A smarter electricity grid requires not only copper, transformers and technicians, but also price incentives, shared data and room for local collaboration. Without that shift, billions risk being invested in infrastructure that is fully utilised during only a handful of peak periods each year.