Abstract

The electrical grid is a huge and complex system which represents a critical infrastructure. Due to this fact, the electric power industry has traditionally adopted a conservative attitude regarding changes. As a result of that, the electrical grid has experienced very few breakthroughs for decades and currently is not prepared to face novel challenges, such as properly integrating DERs (Distributed Energy Resources) or proactively controlling the energy demand by means of the so-called DR (Demand Response) programs, which mainly derive from nowadays society concerns on global warming and climate change. Upgrading traditional electrical grid to the so-called Smart Grid represents one of the most complex engineering projects ever and will certainly drive the next wave of research and innovation in both the energy and the ICT (Information and Communications Technology) sectors. The road towards the Smart Grid will mean an unprecedented revolution especially at the power distribution and customer domains, since the unpredictable and uncontrollable nature of renewables will impose the coordination of generation and consumption points in almost real time. M2M (Machine-to-Machine) communications allow networked devices to communicate between them without further human intervention. What in the very beginning seemed to be a tailored solution for telemetry applications, has become a communications paradigm itself, addressing the myriad of applications existing and yet to be in the wide context of the Internet of Things. As a matter of fact, M2M communications represent one of the main pillars of the Smart Grid in that they will enable the bidirectional real-time exchange of information between the consumption and generation facilities to be monitored and controlled, and the information systems where the optimization processes run. There is a plethora of communications technologies and protocols available within the scope of M2M communications for the Smart Grid. Hence, research is needed in two directions. On the one side, it is required to evaluate how different communications architectures and technologies meet the specific requirements of the Smart Grid before undertaking the important investments needed to deploy this kind of infrastructures on a large scale. On the other side, it is crucial to develop common data models which serve as reference to future horizontal or wide scope protocols which expand across different domains or areas. This thesis aims to tackle these issues. The main goal of the thesis is to contribute to the area of M2M communications architectures tailored to the power distribution and customer domains of the Smart Grid. In order to achieve this overall objective, first we carry out a survey on the most relevant standardization activities developed in parallel to this thesis and on the most outstanding technological and research trends within the Smart Grid area, identifying gaps and challenges. Second, we propose a novel M2M communications architecture to support energy efficiency and optimum coordination of DER (Distributed Energy Resources) within the so-called energy-positive neighborhoods, which are neighborhoods which ensure a substantial part of their consumption by local generation based on renewables. The proposed architecture comprises three network segments, for the sake of flexibility and scalability, and combines different communications technologies to meet the specific communications requirements of each of them. Next, we model formally the domain of knowledge of energy efficiency platforms for energy-positive neighborhoods by means of an ontology developed in OWL (Ontology Web Language), with the aim that it becomes a reference data model for the application of M2M communications to this context. Thus, this ontology has been made public through the EC (European Commission) eeBuildings Data Model community, so that other researches can re-use it and extend it. We also propose a methodology that can be applied, in general, to characterize any communications overlay deployed on top of an infrastructure devoted to any purpose. Following this methodology, we model the traffic of the proposed M2M communications architecture in realistic large-scale scenarios. The main goal of this model is to ensure that potential works based on it actually mean and bring value to the interested parties. Although the model is tailored to the Portuguese power distribution grid, since it is based on actual data provided by EDP (Energias de Portugal), it can be easily adapted to other scenarios by suitably tuning the appropriate parameters. Taking this model as reference, we finally evaluate the core of the proposed M2M communications architecture twofold. On the one side, we analyze the impact of using IPSec (Internet Protocol Security) or TLS/SSL (Transport Layer Security/Secure Socket Layer) as VPN (Virtual Private Network) technologies on the operational costs of a potential energy efficiency platform which relies on the proposed M2M communications architecture. To the author’s best knowledge, no similar studies are available in the state of the art. The main conclusion of this analysis is that using TLS/SSL along with data aggregation is the best option to minimize operational costs at neighborhood level. On the other side, we evaluate by means of simulations the performance of IEEE 802.11b, using as metric the goodput (i.e., throughput at the application layer), and GPRS (General Packet Radio Service), using as metric the transmission time. The first conclusion of these simulations is along the line that IEEE 802.11b meets the requirements in terms of goodput of the NAN (Neighborhood Area Networks), which is of special interest to the Smart Grid community taking into account the low cost and wide adoption of this technology. The second conclusion of such simulations is that GPRS meets the requirements in terms of bandwidth of the backhaul network, thus confirming that it represents a very attractive technology considering that it is the most mature and widely deployed cellular technology in Europe. Presidente: Jürgen Jähnert; Vocal: David Fernández Cambronero; Secretario: Francisco Valera Pintor