What Pythia Sphere tested with The Green Village (TGV)
Are we being too careful with the electricity grid?
It may sound like a strange question. After all, reliability is one of the biggest successes of modern power systems. Grid operators carefully assess failure scenarios, apply safety margins and plan diligently to keep the lights on. Reliability is crucial and should indeed stay as the starting point.
But today’s challenge is different. Heat pumps, EV chargers, solar panels and batteries are being connected faster than the grid can be reinforced. Across the Netherlands this has led to connection queues and growing grid congestion.
The question is no longer simply “What is the maximum possible peak?” It is whether our current planning methods leave valuable capacity unused. Instead of treating every unlikely peak as equally important, can we quantify how often limits are exceeded, how long these exceedances last, and what practical measures could reduce that risk?
This is the main idea behind the TGV field lab.
The goal is not to take more risk and hope it works. The goal is to create insights into the possible risk so that we can use the energy infrastructure even more efficiently.
At The Green Village, TU Delft’s field lab for sustainable innovation, Pythia Sphere tested this with real measured data. The site is a useful example because it has buildings, local energy technologies and measured profiles. The data is used to show the local electricity system as a story, not only as a table.
The workflow is:
- Start with measured import, export and generation profiles.
- Clean the data and extend/fill missing values using weather conditioning.
- Show the results as heatmaps, load-duration-frequency curves and battery scenarios.
These steps provide a transparent probabilistic decision-support framework for connection design, DSO discussions and connection requests.
Heatmap
The heat-map answers an important question: what is normal, high or extreme at a certain moment? Here, we do the analysis at the Point of Common Coupling (PCC), i.e. the point where the TGV community is connected to the main power grid.
What you can do with this graph: Identify when the grid is most likely to experience high loading during the year and distinguish between normal operating conditions and rare extremes. This helps target reinforcement or flexibility where it delivers the greatest benefit.
Load-Duration-Frequency Curve (LDF)
The load-duration-frequency curve (also known as an Intensity-Duration-Frequency curve in the water sector) answers another important question: how often is a capacity limit crossed, and for how many hours?
What you can do with this graph: Quantify the trade-off between connection capacity and operational risk. Instead of planning for a single worst-case peak, you can evaluate how often a capacity limit is exceeded, how long exceedances last, and which mitigation measures offer the best value.
The results show that accepting the risk of just 10 capacity exceedance events per year, each lasting only 15 minutes, can reduce the required connection capacity by up to 20%.
This potentially allows for more nodes to be included within the same grid capacity budget. If this represents too much risk for your business operations, a firm/non-firm contract with the DSO could be a suitable solution. If you want to stay fully in control of your operations, a flexible option such as a battery may be a better fit.
The figure also shows the potential impact of battery storage. In the extreme case, a battery could shave all peaks, allowing the business to operate steadily with a much lower required connection capacity. However, this would require an incredibly large battery, for example 500 kWh with a 6-hour discharge duration, resulting in a 57% connection-limit reduction.
By contrast, a small 10-kWh battery with a 1-hour discharge duration only achieves a 1% reduction. A 50-kWh battery with a 1-hour discharge duration, however, already achieves a 35% peak reduction. The optimal solution should therefore be determined through a cost-benefit analysis, considering the cost of battery flexibility, the operational cost of business interruptions or blackouts, and the cost of not obtaining the desired connection capacity.
This makes the battery discussion much more concrete. A battery is not just “good” or “bad”. It can help with short peaks, or it may not help enough for long congestion periods. Sphere shows that difference, so battery sizes can be compared properly.
This is the value of the TGV case: it turns grid scarcity into something concrete. Not only “there is no capacity”, but “this is the risk, this is when it happens, and this is what a solution changes.”
For readers who want the bigger argument behind this, Netcongestie als risicovraagstuk explains why congestion is not only an infrastructure problem, but also a question of what risk society accepts, who decides that, and how that choice is made explicit.
The Green Village case demonstrates that once risk is quantified, conversations about congestion become far more constructive. Instead of asking whether the grid is “full,” we can ask when capacity constraints occur, how severe they are, and which interventions provide the greatest benefit.
The shift from worst-case assumptions to quantified risk is what will enable more informed and more efficient grid planning in an increasingly electrified world.
Curious to see this project live? Click here.
References
- van der Heijden, Palensky and Vergara-Barrios – Netcongestie als risicovraagstuk
- The Green Village
- Netbeheer Nederland – Feiten en cijfers
- Netbeheer Nederland – Capaciteitskaart
- ENTSO-E System Operation Guideline
- Copernicus Climate Data Store – ERA5
- Dutch Delta Programme – Flood risk management
- Giraldo et al. – Risk-Aware Flexibility Areas for EV Charging Pools
