Modeling heterogeneous and non-rational energy preferences within renewable energy communities

Description

The operation and planning of modern electric power systems face major transformations nowadays, due on the one hand to the increasing share of renewable generation in the electricity mix, and on the other hand to the liberalisation of the electricity supply chain. The main challenge with electricity systems consists in the fact that generation and consumption must be physically equal at every instant in order to maintain system stability, since electrical energy is currently non storable at a large scale. Extreme problems of coordination must thereby be solved by modern system operators, which are furthermore complicated by the fact that they do not own the generation (and consumption) assets, since the liberalisation of the electricity sector. The coordination is in that context performed through multiple market platforms, on which the market actors can interact at different time horizons by exchanging/offering different products (energy, reserve, capacity, etc.).

Currently, new modes of exchange of electricity tend to emerge at the local level, which question the market structure depicted above. This is motivated firstly by the proliferation of decentralized renewable energy resources (owned by small end-users or prosumers), following the ambitious environmental targets promoted at the European and worldwide scale, for which a more efficient coordination could be achieved locally. The increasing willingness of the citizen to play an active role in the electricity supply chain is another important driving factor. The literature speaks of’consumer-centric electric systems’, for which the end-user is placed at the centre of the electrical energy value chain. The European Commission has for instance recently formalized the concept of Renewable Energy Communities (RECs), which are constituted by organized groups of consumers/prosumers (i.e. consumers who can also produce electricity) allowed to exchange electricity locally without resorting to the traditional wholesale/retail market structure.

The local coordination of energy exchanges within a community, and between a community and the external electricity system, relies on advanced mathematical tools, such as optimization and game theory. Currently, most of the literature considers that 1. the members of a given community all share the same objectives (e.g. minimizing the electricity bill) and that 2. all the members behave as rational economic agents, according to the expected utility theory. Even if it may be true for large market actors who often interact with market platforms, this is not the case for most members of RECs, who are typically small domestic end-users connected to the Low Voltage electricity network, or small-to-medium companies connected to the Medium Voltage network. In that way, agents may show heterogeneous energy preferences (some actors may want to minimize their GHG emissions even if the electricity bill increases), as well as non-rational preferences (in the sense that their actual decisions will deviate from the solution which maximizes the expected utility).

In this postdoc, we aim at:

1. extending standard centralized optimization formulations to game theoretical models in order to take into account the heterogeneous preferences of the members of Renewable Energy Communities,
2. modeling the non-rational behaviour of the agents of a Renewable Energy Community using Prospect Theory. Prospect theory is a theory of behavioral economics and behavioral finance that was developed by Daniel Kahneman and Amos Tversky in 1979. The theory was cited in the decision to award Kahneman the 2002 Nobel Memorial Prize in Economics. Based on numerous experiments, contrary to the expected utility theory (which models the decision that perfectly rational agents would make), prospect theory aims to describe the actual behavior of people.