Open source Python codes related to the Master thesis "Scandinavian Market Modelling" of Thibaut Richert & Mattia Balidini

Abstract:

Over the past decades, modelling of energy systems has constantly gained momentum and became trendy through the development of more and more sophisticated tools. BALMOREL, WILMAR or else MARKAL-TIMES are some of the best models nowadays available for large scale energy system modelling. They are used for many purposes such as future scenario analysis, study of technical evolution of the energy system or investigation on new policies design. These are advanced, complicated and not always open-source models that try to reflect as close as possible the reality of large and complicated systems. Scandinavian Power Market Modelling modelling from the start a market model reflecting the reality of the Nord Pool day-ahead market and provide it to the public open-source. Understanding the mechanisms of the price formation in this particular market is the base of the presented Thesis. Key features identified as main drivers in the day-ahead prices calculation were identified and studies: Unit commitment, Demand response, Hydro power scheduling and Combined heat and power plants modelling. Results have highlighted the paramount importance of hydro-power scheduling in the day-ahead market due to the high share in the total installed capacity of this technology for Nord Pool (60%). Combined heat and power have shown to be one of the main drivers as well due to the strong correlation between the heat and the electricity prices. Unit commitment and demand response modelling were proven to be interesting features but very specific and too advanced for a tool constructed from the beginning such as the one proposed in this study. The final model developed showed great results with Root Mean Square Errors below 10 €/MWh and R^2 values around 0.30 depending on the country considered.