To reduce carbon emissions the transformation of the heating sector is of great importance. In particular, district heating networks in combination with renewable energy generation and waste heat play a major role. This includes lowering the operating temperatures and transforming the existing networks to decentralized structures. Incorporating this, new operating strategies are needed.

We consider a global optimization approach that aims at finding cost-optimal operating strategies. The resulting problem is based on nonlinear physical equations for describing the network state and binary variables to determine flow directions in the network pipes. The problem is solved with the solver SCIP. Due to the MINLP structure the solving process leads to high computational costs. Therefore, we consider methods that aim at reducing the number of variables and nonlinear constraints, making use of the underlying network structure.

The developed methods are evaluated with numerical results based on a real district heating network.

Large-scale energy system optimization modeling is becoming an increasingly important field, since we need to harness larger shares of variable renewable energy. The main challenge is to include many detailed models of energy assets, while maintaining computational tractability. For example, energy storage systems have become a promising option to increase power system flexibility, but the standard MILP models that describe optimal investment and operation of these storage units, possibly including the optional capacity to provide up/down reserves, do not scale well. A common practice is therefore to relax the integrality constraints, but this results in LP relaxations that allow simultaneous charging and discharging, while this is not feasible in practice. In this talk, we discuss how we can use MILP theory to improve the tightness of such formulations, and how this can benefit energy system optimization models. Our methodology uses the disjunctive nature of the storage optimization problems to derive the convex hull of its solutions for one time period. When included in multi-period large-scale energy system models, these tighter MILP models are expected to scale better, and their LP relaxations are expected to better prevent simultaneous charging and discharging. We demonstrate these two benefits with illustrative case studies of a unit commitment problem and a transmission expansion planning problem. The improved MILP formulations and LP relaxations can be used for many different types of energy storage systems, as well as transmission lines.

Distributed energy generation (DER), such as roof-top solar panels, are expected to become prolific in the net zero electricity distribution system of 2050. The intermittency of DERs poses challenges for operating the distribution system. At scale, however, the intermittency of DERs could be leveraged to provide operational flexibility to higher voltage levels of the power system, including the transmission system. This work develops a new optimization model to characterize the flexibility region afforded by DERs in the distribution level, and analyzes opportunities for use at the transmission level. IEEE distribution test systems are used as case studies of the optimization formulation. This research demonstrates the value of flexibility regions and how to operate active distribution networks to optimize that value.

This paper evaluates the potential effectiveness of Critical Peak Pricing (CPP) as a demand response strategy for peak load shaving in electricity grids, with a specific focus on Quebec's electricity grid. The study employs a unique approach by integrating prosumer behavior into the analysis, accounting for the adoption of distributed energy resources such as photovoltaic panels, batteries, and electric vehicles. Through comprehensive simulations and the application of reinforcement learning algorithms, we analyze the effectiveness of CPP programs, both in mass and targeted offering scenarios. The results reveal that while CPP is effective in incentivizing load shifting, its efficacy diminishes with increasing prosumer participation, leading to new peaks. To counteract this, we propose targeted, dynamic pricing strategies demonstrating significantly improved performance and extended viability. The study also highlights the influential role of batteries and electric vehicles in peak load reduction, suggesting a need for focused policy and incentive structures.