New video: Stochastic Planning vs Minimax Regret
"Network planners" are companies like - National Grid, UK Power Networks, SSEN, Northern Powergrid in the UK. - Enedis and RTE in France. - e-distribución and Red Eléctrica in Spain. - e-distribuzione and Terna in Italy; - E.ON, Netze BW, and 50Hertz in Germany; - Stedin and TenneT in the Netherlands; - PG&E, Con Edison, Duke Energy, and ERCOT in the US etc. These companies have to think about reinforcing the electricity grid years in advance. E.g. when should we upgrade a line? Which part of the network needs reinforcement first? etc Network Planners use a suite of optimisation models to make these decisions, including: - Deterministic optimisation (solving across one scenario ) - Stochastic Planning (SP) ( many scenarios , weighted by their probabilities) - Least-Worst Regret (LWR) ( ignores probabilities and protects against the worst-case regret). Each framework can point to a different "optimal" investment strategy for the same grid. Knowing why they disagree is essential. This 13-minute video comes from a consultancy project I delivered, training an energy company on the differences between LWR and SP. The company wanted an intuitive understanding of how two key inputs (probabilities and social costs) drive the differences in the optimal solutions these two frameworks produce. In the video, I walk through one worked example: a simple case focused on EV-driven demand growth, where we compare what SP and LWR recommend under different probability assumptions and different social cost values. The results show that in some cases the two frameworks agree completely, in others they point in opposite directions. This is just one example of the kind of analysis that helps planners choose the right framework for the right decision. The video is in course 120. And the new video is "5.1. Effect of Social Cost & Probabilities". The attached slide offers a summary of key results.