Operation and Control of Active Distribution Networks

Abstract

The Distributed Generation paradigm has encouraged the growth of small and medium-scale power generation systems located close to the demand centers. As a consequence, in recent years the traditional distribution networks with unidirectional power flows are in a transition towards what is called Active Distribution Networks (ADNs), with the insertion of distributed energy resources (DERs) in different nodes of the network. These transformations in the electrical systems bring new problems and challenges. The insertion of DERs in ADNs generates the presence of bidirectional active power flows. This issue, couple with a high R/X ratio of medium and low voltage lines and the intermittency of renewable generation sources, can generate fluctuations and deviations in the voltage profiles of the feeders. For this reason, the development of methodologies for the design and implementation of an adequate control system is of special importance. The study of control in ADNs is a multidisciplinary area of research since it involves fields of study such as design, modeling, control, optimization, monitoring, and supervision, among others. One of the main challenges is to approach these fields in a comprehensive manner. In fact, the global study of the problem will result in important advantages such as cost reduction, increased efficiency, flexibility and system reliability. According to the previous definitions, we proposed to address this problem with methodological tools from the Process Systems Engineering (PSE) discipline. The control requirements in ADN are diverse and operate on different time scales, so they can be addressed from a hierarchical control structure point of view. In particular, in this work we will address the analysis, design and implementation of control systems for ADNs based on the Plant-Wide Control (PWC) theory, which is a research subline within the PSE. A systematic strategy for the voltage control problem is proposed to find control structures that optimize the necessary hardware requirements for its implementation with good dynamic performance. The general PWC design procedure is formulated as an optimization problem where the cost functions depend only on the steady-state models of the network. The methodology provides a systematic criterion for the selection of the controlled variables (measured variables), compared to the strategies present in the literature, which are generally based on heuristics. On the other hand, the optimal operation of the active distribution network can be divided into two different optimization problems: Unit Commitment (UC) and Economic Dispatch (ED). The UC problem determines the start-up and shut-down schedule of all dispatchable units to supply the electric demand minimizing the total operating cost, while the ED determines, for each hour of the planning horizon, the actual power output of each of the committed generating units necessary to supply the demand and comply with the limits imposed by the network. Here the two problems are solved simultaneously to achieve the day-ahead optimal operation of active distribution networks with distributed generation and energy storage. For this purpose, a Mixed Integer Linear Programming formulation using MOST (MATPOWER Optimal Scheduling Tool) is proposed. All the proposed methodologies are evaluated using the standard IEEE 33-bus distribution network, to which we incorporate distributed generation and energy storage for a more profitable and flexible case study.