Empowering Prosumers: Incentive Design for Local Electricity Markets Under Generalized Uncertainty and Grid Constraints

TL;DR

The paper tackles integrating uncertainty from stochastic renewables into local electricity markets by introducing a probabilistic Locational Marginal Pricing framework built on a convex chance-constrained OPF with the lindistflow grid model. Uncertainty is propagated via general Polynomial Chaos ($gPC$), enabling day-ahead price distributions and real-time prices to be obtained by evaluating the PC expansions, while ensuring grid feasibility with probabilistic constraints. It also advocates passive balancing by end-prosumers through two strategies—rule-based and DP-based—so that market operation remains scalable and simple for numerous small participants. The approach is demonstrated through four case studies, including a 179-bus network, showing improved pricing signals, congestion relief, voltage constraint management, and computational scalability, with practical implications for integrating prosumers and DERs into distribution grids.

Abstract

Since the 1990s, widespread introduction of central (wholesale) electricity markets has been seen across multiple continents, driven by the search for efficient operation of the power grid through competition. The increase of renewables has made significant impacts both on central electricity markets and distribution-level grids as renewable power generation is often connected to the latter. These stochastic renewable technologies have both advantages and disadvantages. On one hand they offer very low marginal cost and carbon emissions, while on the other hand, their output is uncertain, requiring flexible backup power with high marginal cost. Flexibility from end-prosumers or smaller market participants is therefore seen as a key enabler of large-scale integration of renewables. However, current central electricity markets do not directly include uncertainty into the market clearing and do not account for physical constraints of distribution grids. In this paper we propose a local electricity market framework based on probabilistic locational marginal pricing, effectively accounting for uncertainties in production, consumption and grid variables. The model includes a representation of the grid using the lindistflow equations and accounts for the propagation of uncertainty using general Polynomial Chaos (gPC). A two-stage convex model is proposed; in the day-ahead stage, probability distributions of prices are calculated for every timestep, where the expected values represent the day-ahead (spot) prices. In the real-time stage, uncertainties are realized (measured) and a trivial calculation reveals the real-time price. Through four instructive case-studies we highlight the effectiveness of the method to incentivize end-prosumers' participation in the market, while ensuring that their behavior does not have an adverse impact on the operation of the grid.

Figures (9)

• Figure 1: Grid for case-studies 1-3, which is a modified version of the medium-voltage distribution network benchmark developed by the CIGRÉ Task Force C6.04.02 conseil_international_des_grands_reseaux_electriques_benchmark_2014.
• Figure 2: Upper left and right: power flow schedule at the slack bus and arbitrage price seen by the ESS in bus 10. Lower left and right: average cumulative regret and average ESS Energy level across scenarios.
• Figure 3: Case study on local congestion and resulting PLMP. The congestion occurs at timestep 10, when the load in node 4 increases. The resulting PLMP show a congested system.
• Figure 4: Distribution of nodal voltage magnitudes and the topology of the considered grid. The probabilistic evolution of the nodal voltage magnitudes is shown above every bus. The x-axis represents the 24 timesteps in a day and the y-axis represents the nodal voltage magnitude. All axes have the same scaling. The inserted histogram shows the distribution of the nodal voltage in bus 9 for the 12-th timestep, when the network experiences voltage congestion due to increased PV production.
• Figure 5: Distribution of PLMP and the topology of the considered grid. The probabilistic evolution of the PLMP is shown above every bus. The x-axis represents the 24 timesteps in a day and the y-axis represents the PLMP. All axes have the same scaling.