Redundancy in Communication Networks for Smart Grids

Traditional electric power grids are currently undergoing fundamental changes: Representative examples are the increase in the penetration of volatile and decentralized renewable-energy sources and the emerging distributed energy-storage systems. These changes are not viable without the introduction of automation in grid monitoring and control, which implies the application of information and communication technologies (ICT) in power systems. Consequently, there is a transition toward smart grids. IEEE defines smart grid as follows: "The integration of power, communications, and information technologies for an improved electric power infrastructure serving loads while providing for an ongoing evolution of end-use applications" . The indispensable components of the future smart grids are the communication networks. Many well-established techniques and best practices, applied in other domains, are revisited and applied in new ways. Nevertheless, some gaps still need to be bridged due to the specific requirements of the smart-grid communication networks. Concretely, a challenging objective is to fulfill reliability and low-delay requirements over the wide-area networks, commonly used in smart grids. The main ``playground" for the work presented in this thesis is the smart-grid pilot of the EPFL campus. It is deployed on the operational $20kV$ medium-voltage distribution network of the campus. At the time of the writing of this thesis, the real-time monitoring of this active distribution network has been already put in place, as the first step toward the introduction of control and protection. The monitoring infrastructure relies on a communication network that is a representative example of the smart-grid communication networks. Keeping all this in mind, in this thesis, the main topic that we focus on, is the assurance of data communication over redundant network-infrastructure in industrial environments. This thesis consists of two parts that correspond to the two aspects of the topic that we address. In the first part of the thesis, we evaluate existing, well-established, technologies and solutions in the context of the EPFL smart-grid pilot. We report on the architecture of the communication network that we built on our campus. In addition, we go into more detail by reporting on some of the characteristics of the devices used in the network. We also discuss security aspects of the MPLS Transport Profile (MPLS-TP) which is one of the proposed technologies in the context of smart grids. In the second part of this thesis, we propose new solutions. While designing our campus smart-grid network, we analyzed the imposed requirements and recognized the need for a solution for reliable packet delivery within stringent delay constraints over a redundant network-infrastructure. The existing solutions for exploiting network redundancy, such as the parallel redundancy protocol (PRP), are not viable for IP-layer wide-area networks, a key element of emerging smart grids. Other solutions (MPLS-TP for example) do not meet the stringent delay requirement. To address this issue, we present a transport-layer solution: the IP-layer parallel redundancy protocol (iPRP). In the rest of the thesis, we analyze the methods for implementing fail-independent paths that are fundamental for the optimal operation of iPRP, in SDN-based networks. We also evaluate the benefits of iPRP in wireless environments. We show that, with a help of iPRP, the performance of the communication based on the Wi-Fi technology can be significantly improved.

Related material