Fast, precise and simple energy system models.
We design our models for AI-enhanced, GPU-accelerated optimization. The aim: detailed studies that run fast enough to explore the futures that matter.
The team.
A TU Delft spin-off with roots in optimization under uncertainty, machine learning and high-performance solvers.
A group of TU Delft students builds and tests the platform with the core team, so the work stays close to the research it comes from.
Meet the full team →Why we can make it better.
We populate our models with the most accurate representations of renewable energy and conventional generation
Modeling renewables is hard and matters greatly for the accuracy of the model.
The European fleet of conventional generators is accurately represented in our model
The European grid is represented in detail
We combine new model formulations with high-performance computing and AI. The aim is to solve large energy system models in a fraction of the usual time.
Run large European power system models in minutes instead of overnight runs.
Full model fidelity, without simplifying away what matters.
Using modern cloud infrastructures and our knowledge about optimization models.
We turn solver output into results that you can compare across many scenarios, so you see how much the answer depends on uncertain inputs.
A modern and fast viewer of results, not raw solver files.
Dive into relevant statistics and compare hundreds of scenarios.
Actively looking for pilot partners.
We take on pilot projects with partners who bring a question from their own system. For example:
Wind farm capture rates
Which share of the average market price does a wind farm earn, across many weather years and future scenarios?
European resource adequacy
Does the European power system keep up with demand in a cold, windless winter, and which resources close the gap?
Electricity price forecasts
Which prices can you expect in your bidding zones, in the short and in the long term?
Built on research at TU Delft.
Our team-members are involved in supervision and research at TU Delft at various levels: PhD projects, MSc thesis projects, internships, etc. Below is a selection of recent work with relevance to Greenville. Note that these are independent scientific publications, with no formal role of Greenville.
Wake models come in different varieties: fast, analytical models that can be used for wind farm optimization and high-fidelity models that can compute wake effects in great detail but are too slow to be used in optimization loops. Graph Neural Networks, trained on high-fidelity model data, offer a way out of this impasse.
Why it mattersShows the potential of Graph Neural Networks for fast and accurate wind farm wake modelling2026Fundamental Approaches to Dutch Electricity Price FormationEnergy system optimization models can be used to model electricity prices, but are known to produce “too flat” price curves. This thesis explores various techniques to get more realistic prices.
Why it mattersExplores how energy system models can yield more realistic price forecastsSee more research →Built in Europe, run in Europe.
Model code, training data and compute stay on European infrastructure. The critical infrastructure of the energy transition should not depend on software you cannot inspect or hosting you cannot reach.
