Multistage Expansion Planning of Active Distribution Systems: Towards Network Integration of Distributed Energy Resources

Over the last few years, driven by several technical and environmental factors, there has been a growing interest in the concept of active distribution networks (ADNs). Based on this new concept, traditional passive distribution networks will evolve into modern active ones by employing distributed energy resources (DERs) such as distributed generators (DGs), energy storage systems (ESSs), and demand responsive loads (DRLs). Such a transition from passive to active networks poses serious challenges to distribution system planners. On the one hand, the ability of DGs to directly inject active and reactive powers into the system nodes leads to bidirectional power flows through the distribution feeders. This issue, if not adequately addressed at the design stage, can adversely affect various operational aspects of ADNs, specifically the reactive power balance and voltage regulation. Therefore, the new context where DGs come into play necessitates the development of a planning methodology which incorporates an accurate network model reflecting realistic operational characteristics of the system. On the other hand, large-scale integration of renewable DGs results in the intermittent and highly volatile nodal power injections and the implementation of demand response programs further complicates the long-term predictability of the load growth. These factors introduce a tremendous amount of uncertainty to the planning process of ADNs. As a result, effective approaches must also be devised to properly model the major sources of uncertainty. \nBased on the above discussion, successful transition from traditional passive distribution networks to modern active ones requires a planning methodology that firstly includes an accurate network model, and secondly accounts for the major sources of uncertainty. However, incorporating these two features into the planning process of ADNs is a very complex task and requires sophisticated mathematical programming techniques that are not currently available in the literature. Therefore, this research project aim to develop a comprehensive planning methodology for ADNs, which is capable of dealing with different types of DERs (i.e., DGs, ESSs, and DRLs), while giving full consideration to the above-mentioned two key features. To achieve this objective, five major steps are defined for the project. \nStep 1 develops a deterministic mixed-integer linear programming (MILP) model for integrated expansion planning of distribution network and renewable/conventional DGs, which includes a highly accurate network model based on a linear format of AC power flow equations. This MILP model can be solved using standard off-the-shelf mathematical programming solvers that not only guarantee convergence to the global optimal solution, but also provide a measure of the distance to the global optimum during the solution process. Step 2 proposes a distributionally robust chance-constrained programming approach to characterize the inherent uncertainties of renewable DGs and loads. The key advantage of this approach is that it requires limited information about the uncertain parameters, rather than perfect knowledge of their probability distribution functions. Step 3 devises a fast Benders decomposition-based solution procedure that paves the way for effective incorporation of ESSs and DRLs into the developed planning methodology. To this end, two effective acceleration strategies are proposed to significantly enhance the computational performance of the classical Benders decomposition algorithm. Eventually, Steps 4 and 5 propose appropriate models for ESSs and DRLs and integrate them into the developed planning methodology. In this regard, a sequential-time power flow simulation method is also proposed to incorporate the short-term operation analysis of ADNs into their long-term planning studies. \nBy completing the above-defined steps, the planning model developed in Step 1 will be gradually evolved, so that Step 5 will yield the final comprehensive planning methodology for ADNs.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.