ENERGOS: Integral smart grid management Yoseba K. Penya

Juan Garbajosa

Mariano Ortega and Eloy Gonz´alez

DeustoTech - University of Deusto Donostia, Basque Country [email protected]

Technical University of Madrid (UPM) Madrid, Spain [email protected]

Indra Software Labs Madrid, Spain {mortega, egonzalezort}@indra.es

Abstract—The smart grid revolution demands a huge effort redesigning and enhancing current power networks, as well as integrating emerging scenarios such as distributed generation, renewable energies and the electric vehicle. This novel scenario will present a huge flood of data that can be only handled, processed, and exploited with the help of cutting-edge ICT technologies. We present here the ENERGOS project that fully addresses this evolution, and detail accurately how it will deal with the challenges posed by the smart grid paradigm.

I. I NTRODUCTION Hundred years from their initial development, deployment, and spreading, power networks must face nowadays a new revolution. With the peak of petrol production soon in sight, the continuous increment in the demand world-wide as well as the need to reduce contaminating emissions, efficiency at all levels is the key to the optimal, sustainable use of the resources [1]. But energy efficiency is not the only goal to be achieved: a number of new models, methods, and possibilities are already imposing different requirements that the classical power networks just cannot fulfill. The paradigm smart grid (term coined in contrast to the old allegedly non-intelligent ones) refers to the future electrical networks in which these novel scenarios and ICT (Information and Communication Technologies) will go hand on hand to achieve this excellency. Historically, the electrical grid evolved and extended nationwide, government controlled, and centralized. The liberalization of the electrical markets broke this tight shape, opening the way to transnational energy business and corporations. The inheritance, however, remained: diverse network specifications, models, protocols, standards in use, architecture and markets. . . preventing in this way the design and development of a unique smart grid approach [2]. As the consumption growth unsparingly chases the generation capacity, there has been a hectic research around smart grid as the tool to the efficient energy management and consumption. The focus has been laid on certain aspects of the problem, usually either the logical or the physical part, introducing to this end either of the computing paradigms. We present here the ENERGOS project [3] that fully addresses the optimal deployment and management of smart grids in both levels, logical and physical. Though ENERGOS is not the first effort to employ the so-called Semantic Web Technologies [4], [5], it is the primer to use them thoroughly in its architecture, fully exploding the advantage that this model offers.

It should be mentioned that IEEE1 has started an umbrella initiative that, using current IEEE resources, intends to push ahead the smart grid concept. To this end, they have created a portal including a smart grid conceptual model, publications, and conferences related to the topic, standards, education, and a survey of public policy that considers Australia, China, European Union, India and USA. The remainder of the paper is organized as follows. Section II discusses related work. Section III provides an introduction to the smart grid domain and its challenges. Section IV details the main research areas addressed by ENERGOS. Section V discusses the architecture that allows distributing the intelligence all over the grid and its compounding parts. Finally, section VI concludes. II. R ELATED W ORK Many of the existing smart grid projects cope with different aspects of communications for supporting the power grid. Among these, the project DLC+VIT4IP [6] is focused on the design, implementation, evaluation, and test of the communication infrastructure using the existing power distribution network. This allows the interconnection of sensors and actuators to implement (new) intelligent services for energy consumption observation, energy management, and efficiency. The project HiPerDNO [7] envisions the future electricity distribution networks will consist of millions of smart meters, small-scale embedded generation, and responsive load will generate vast amounts of data requiring near to real-time analysis; this project deploys high performance and grid computing technologies. The INTEGRIS project [8] proposes the development of a novel and flexible ICT infrastructure based on a hybrid power-line communication-wireless integrated communications system able to fulfil the communications requirements foreseen for the future electricity networks. The project PRIME [9] is focused on the development of a new open, public and non-proprietary telecom solution for supporting smart metering and smart grids in general. The IECSA initiative of the EPRI [10] is also focused on the data exchange infrastructure for the power grid. The OpenNode project [11] is focused on the supporting communication, including middleware, of a power distributed grid. The project [12] aims at defining a set of draft standards, to support distribution of energy in general terms (e.g. including gas distribution) and 1 Institute

of Electric and Electronic Engineers http://ieee.org

using existing standards, such as the IEC 61334 series for power-line communication standards. Other projects, such as [13] studies several aspects of energy efficiency in power grids. The project Miracle [14] intends to develop an approach on a conceptual and an infrastructural level that allows energy distribution companies to balance the available supply of renewable energy sources and the current demand in an ad-hoc fashion. The network [15] copes with large-scale integration of domestic and distributed micro generation and promote improvements in energy efficiency through the implementation of innovative ICT solutions into local smart power grids. The action COST [16] fosters cooperation aiming at energyefficient alternate solutions to share IT distributed resources. One of the main difference of these projects with respect to ENERGOS is that ENERGOS presents a more holistic perspective; the advances in single aspects (e.g. communications) must be integrated and coordinated with the progress in other aspects such as a real time platform for controlling and monitoring the power grid. III. D OMAIN DESCRIPTION A. Approaching Intelligence to the grid: an opportunity The electric network landscape is changing. This landscape refers to how energy can be generated and the characteristics of demand. The commitment European 20-20-20 to reduce 20% CO2 emissions, to reduce 20% the global consumption, and to have 20% of the consumption from renewable energies, makes it necessary to integrate in the future more and more of renewal energies in the low and medium voltage/tension network. In the coming years tens of thousands of new installations for renewable power generation such as solar, micro-generation or mini-wind energy will have to be integrated. This will radically change the way distribution is performed. We can guess that new scenarios and situations will come up. The current power grid structure includes a vertical design that goes from a huge number of meters whose load data is gathered manually or remotely into SDCs (Substation Data Concentrator) though in sometimes, as in the USA, this component is replaced by direct access to the meters or access through a gateway. The upper level is composed of the secondary substation connected to the these SDCs and industrial or high-voltage meters. Above, the primary substations control the transition from the transport network to the distribution one. Finally, on top of the pyramid sit the Utility control and business systems. Moreover, currently human operators are in charge of managing current power network infrastructures. In this task, operators are supported by SCADA, EMS (energy management system), and DMS (demand-management system). The overall architecture core is a Central Controller Centre. Power network operation needs to be safe, reliable and, as long as possible, cost effective. The intelligence for a proper operation is mostly provided by human operators: the intelligence required for this relies, basically, on humans. However power networks are undergoing, as outlined in section III, a deep transformation

in the sense a distributed bidirectional flow power generation is approaching to us. To manage adequately these networks it is necessary to empower intelligence, and this can be achieved by a wide application of information technology devices. As a result, the Central Controller Centre will evolve towards an architecture based on distributed intelligent devices. This evolution will concern the technical and, as important, the business sphere. Another issue is demand. The objective to take to consumers and active role, This active role will be only possible as long as consumers are aware of the real consumption, and they pay the actual price for the actual consumption. This will allow us to implement an a Active demand management, and for sure will drive us to a scenario of more power efficiency. For this home energy devices, that can set up a dialogue between the producer and the consumer is needed. At present smart meters are are already being installed, but that is a first step. A third issue are electric cars. Electric cars will introduce a new kind of consumers, consumers that may need a large amount of energy at certain days (e.g. a bank holiday). This may collapse the whole system, if the system is not ready for that. At the same time, cars may be also become donors of energy. Therefore the electric grid has to evolve to a smart grid where intelligence play a key role. We can devise a layered architecture where intelligence can be introduced at different levels: network engineering, control platform, communication and integration, real time, technical systems and management systems. Communications systems work as a bridge between the network engineering,and the control layer and the real time and technical management systems. That is, the network engineering is supported by a number of control platforms (distributed generation, substations, transformation centres, PHEV - new suppliers, and smart meters). The other layers provide, based on real time systems kernels, subsystems for managing the distributed generation, network analysis and planning, control and monitoring centres, and remote-management and remote-measurement. On top of this corporative markets of energy and generation systems, network maintenance, and operation will be standing. Finally Commercial systems can also use this information. This layer architecture can be understood as an architecture in which information is flowing upwards and downwards. This information flow relies on the communication and integration systems: they are the bridge that joins two words, the network world, that is dramatically disperse across the geography, with a huge number of nodes, and the other world of computers departments, also distributed but much less de-localized. Standardization is essential to guarantee a reasonable degree or interoperability. This interoperability should be defined in such a way that an open market could be there. For this reason the IEC Technical Committee 57 has produced the specification shown in Fig. III-A, illustrating the principal components of the smart grids and the main related standard protocols. Though an extensive description is out of the scope of this

paper, it should be noted that this very general framework is understood in the following way within the project as follows: Three layers have been identified: highlighted, as follows: the communications bus, a real time middle-ware at the data acquisition layer, and a business bus, at the Common Information model [17], [18] layer.

IV. R ESEARCH AREAS With the scenario and requirements drawn in the previous section in mind, we put forward here the research areas in which ENERGOS is divided and their goal. •

B. Smartgrids in the praxis The power network classical architecture was a stack in whose bottom was the generation of the energy, above the transport, then distribution, and finally consumption. Liberalisation of the energy market added a new layer for the utilities but, still, whereas the energy flowed bottom-up, data did upbottom. This picture has been blurred by the apparition of new technologies and models with mixed roles. First, distributed generation has broken with the unique directionality of the energy and now, it may flow both ways. Second, smart grids will enable data interchange from the client to the utility, and vice versa (e.g. to transmit a price signal), and between all the members of the systems. Third, micro-grids and the electric vehicle (EV), for instance, out-shape this stack by merging features of all the layers. Micro-grids, on one hand, combine low-scale energy generation and storage to provide electricity to their customers, and may be connected to the grid or work as an isle, therefore they contain generation, distribution, consumption, and, optionally, utility and transportrelated functions. EVs, on the other hand, may act as energy consumers and generators, in both roles introducing an additional obstacle: the ability of changing places (i.e. are not attached to a certain distribution grid). With these premises in mind, the smart grid presents four distinct layers: •







Application level: Supporting diverse business applications, distribution and transport network operation (e.g. failure detection, load forecasting, etc.), retailer applications (tariff customisation, remote billing, etc.), as well as transversal applications (e.g. demand-side management programs). Data layer: Dealing with processing the new flood of data, data protocol interoperability and standardization, valuable information extraction. . . Communication layer: Enabling data interchange among the different participants and communication protocol interoperability. Power layer: Concerning the optimization of current network structures and components (e.g. smart metering), integration of emerging technologies (e.g. distributed generation and storage, electrical vehicle, etc.), and developing of new tools.

The three upper layers conform the logical level, where as the bottom one, the physical level. Table I shows a brief list pointing up the new needs and requirements that each of these areas poses within the smart grid paradigm.











Demand forecasting and management: This area deals with short-time load forecasting in distribution networks, including load forecasting per substation, overall load forecasting (i.e. transport-grid-wide), applying statistical time-series modelling and Artificial Intelligence (AI) methods-based prediction (e.g. Neural Networks). Micro-grid automatic supervision and control: This area addresses the islanding problem of micro-grids. If a micro-grid presents distributed generation and the transmission grid experiences problems and fails to provide energy, the distributed generation may avoid the blackout within the micro-grid but it also may lead to abnormal values of frequency and voltage yielding risks for injuries and damage to equipment [19]. Real-time data acquisition and processing: As a core part of the project, this area focuses on the challenge of acquiring data (some of it in real-time), process and transform it into valuable information, and dealing with network events (e.g. alarms). To this end, this area defines an unique, extensible ontology (based on the CIM model [4]) and proposes a distributed architecture of several intelligent nodes understanding the part of the ontology (called profile) related to their activity. In this way, all the previous data becomes enriched information, supporting in this way the inference of new, unknown knowledge. Smart meters: This area addresses the optimal design of the smart meters, their computing capabilities and how to integrate them in the aforementioned semantic structure. It also deals with the interoperability of the metering standard protocols. Advanced operation and learning environments: this area studies new approaches to support the network operation. This includes the use of intelligence for supporting operation, embedding expert knowledge; and also simulation of the grid. Operation planning This area covers algorithms for the power network operation planning, including long term TABLE I N EEDS AND REQUIREMENTS IN EACH SMART GRID AREA .

Generation Transport

Distribution Utility Micro-grids Elec. Vehicles Consumption

Planing, supervision, control Proactive maintaining, operation planing, network operation, deliverance quality, advanced operation bureaus, distributed generation integration, primary system automation, failure detection, security, safety Short-time load forecasting, spatial forecasting, operation, low-voltage data acquisition Remote billing, tariff customization, new electric market formats, demand-side management programs Operation, safety, security, grid balancing, fault management, distributed generation integration, optimized network topologies, reactive energy control Mobility and demand, charge management Demand-side management, load balancing, microgeneration integration

Fig. 1.





TC57 Reference architecture with smart grids standards.

planning, and demand estimation and generation. Network operation: The scope of this area, somehow complementary to operation planning, includes algorithms for the network automatic operation, crisis contingency and management, distributed agent coordination, information filtering of the information coming from the grid, and algorithms for social and economic impact estimation. Infrastructures for re-charging and management of electric vehicles: This area is one of the examples of the holistic vision of the project. Nowadays, electric vehicles are already part of the future grid. Therefore, it is important to analyse the infrastructure required for car recharging and the impact on the grid: new control algorithms and models are required.









Communications: Communications is an important area; the related work in section II has clearly highlighted this fact. Therefore, ENERGOS will study and analyse new protocols at smart grid service level, PLC technologies for smart grid implementation, WDM technologies, wireless technology, communication network topology, quality of service analysis and, finally, safety and trust. Methods and techniques for signal capture: This area includes both signal capture and degradation, and image deployment. This area also takes care of alternatives to conventional for monitoring methods. Automation of primary power equipment: This covers automation of transform centres, and network quality improvement. Standards: Standards are recognized as an essential

part of industrial research and technology consolidation and transfer. Therefore they are present in ENERGOS. There is a clear intent to transform research results into standards proposals, both either to submit new standards to international and national bodies, or to suggest improvements too current existing standards. V. D ISTRIBUTED I NTELLIGENT A RCHITECTURE As we have seen, there is no centralised architecture possible since gathering the complete information, if feasible at all, would entail too a high cost in infrastructures and in time. Therefore, we have to find a trade-off between distributing the intelligence and the costs associated to this decision (e.g. in embedded platforms that support it). In ENERGOS, the key stone for this challenge is the so-called PGDIN ( Power Grid Distributed Intelligence Node), placed from the Utility downwards until the SDC (Substation Data Concentrator) as shown in Fig. 2. In the bottom of the system, we find the Smart Meters, which besides publishing the consumption that they register, are also able to receive information and react consequently (for instance, when receiving information on energy prices). The PGDINS are virtually interconnected following the natural structure of the grid. Therefore, for instance a primary substation PGDIN gets information from the secondary substation PGDINS attached to it (and, eventually, also from SDCs that gather information from industrial smart meters), processes that information and offers it to the upper level (in this case, the PGDIN at the Utility). They use a message-oriented middleware (MDM) to assure that no packet is lost (other common protocols such as TCP or VP are not connection oriented) in real-time; therefore, they can be sure that every information sent arrives to the destiny within the established maximum time. Specifically, ENERGOS relies on the OMG DDS standard [20], which offers a widely parametrisable Quality of Service (QoS). The sharing of the information is achieved in two ways: first, it offers a SOA (ServiceOriented Architecture) interface to retrieve and query usual meter information (e.g. active and reactive power consumption in different time horizons, historical load data, etc.). Second, it is also connected to a CEP (event-oriented bus) in order to communicate events such as alarms, status, and information that, due to its nature, requires special handling (e.g. oil temperature of the substation, security sensor data within the secondary or primary substation, etc.). These events are processed in upper layers and, if needed, forwarded to the next level. The intelligence is distributed since the PGDIN can take decisions autonomously based on a local knowledge base (persisted semantically [21]), and a semantic reasoner [22]. The collaboration between PGDINs may also happen to resolve specific situations that, for instance, require comparing or combining data from more than one PGDIN. Furthermore, these interactions can be also understood from an interoperability perspective problem in which all the PGDINs share a common

Fig. 2. Distribution of the intelligent nodes all over the ENERGOS architecture.

ontology (more accurately, the ontology backed by the IEC and the EPRI for the smart grid [23]. VI. CONCLUSION With the objective of increasing the energy efficiency and simultaneously reduce contaminating emissions, the smart grid paradigm has emerge from the combination of a redesign of the old power networks; a radical new manner of dealing with the data they generate is being developed. The ENERGOS project, with a holistic vision, is addressing all parties involved in the smart grid. Relying on the new design of the network, and in combination with the introduction of information technologies is enabled the implementation of distributed intelligence; results are both in the technical and in the business side. From this, an improved energy management can be achieved; as important, new business models can be applied. The society may get a profit from both, as their combination will turn out in a better use of energy. Future grids will present classical and distributed generation, transport networks and micro-grids, distributed storage and Electric Vehicles being both consumers and generators, enhanced grid-wide deliverance security against blackouts, more accurate load forecasting (i.e. more accurate network operation planning), client-side energy efficiency programs, etc. The ENERGOS project is currently deploying and testing its approach. ACKNOWLEDGEMENTS This research work has been sponsored by the Spanish MEC (ENERGOS Project TIN2009-00000). R EFERENCES [1] U.S. Department Of Energy, “Grid 2030: A national vision for electricitys second 100 years,” 2003.

[2] NIST Framework and Roadmap for Smart Grid Interoperability Standards, Release 1.0, draft ed., National Institute of Standards and Technology, U.S. Department of Commerce, 2009. [3] E. Consortium, “Project ENERGOS: Technologies for automated and intelligent management of energy distribution networks of the future.” [Online]. Available: http://innovationenergy.org/energos/ [4] A. Wagner, S. Speiser, and A. Harth, “Semantic web technologies for a smart energy grid: Requirements and challenges.” [5] A. Crapo, X. Wang, J. Lizzi, and R. Larson, “The semantically enabled smart grid.” [6] DLC+VIT4IP, “Project DLC+VIT4IP: Distribution Line Carrier: Verification, Integration and Test of PLC Technologies and IP Communication for Utilities.” [Online]. Available: http:www.dlc-vit4ip. org [7] HiPerDNO, “Project HiPerDNO: High Performance Computing Technologies for Smart Distribution Network Operation.” [Online]. Available: http://dea.brunel.ac.uk/hiperdno [8] INTEGRIS, “Project INTEGRIS: INTelligent Electrical Grid Sensor communications: INTelligent Electrical Grid Sensor communications.” [Online]. Available: http://fp7integris.eu [9] PRIME, “Project PRIME = PoweRline Intelligent Metering Evolution.” [Online]. Available: http://www.prime-alliance.org/ [10] EPRI - IECSA, “Project Integrated Energy and Communication Systems Architecture.” [Online]. Available: www.epri-intelligrid.com [11] OpenNode, “Project OpenNode: Open Architecture for Secondary Nodes of the Electricity SmartGrid.” [Online]. Available: http: //www.opennode.eu [12] OPENmeter, “Project Open Public Extended Network Metering.” [Online]. Available: http://www.openmeter.com/

[13] SEESGEN, “Project SEESGEN-ICT: Thematic Network to encourage energy efficiency in SmartGrids.” [Online]. Available: http://seesgen-ict. cesiricerca.it [14] MIRACLE, “Project MIRACLE: Micro-Request-Based Aggregation, Forecasting and Scheduling of Energy Demand, Supply and Distribution.” [Online]. Available: http://www.miracle-project.eu [15] ICT4SMARTDG, “Project ICT4SMARTDG: ICT for smart distribution generation.” [Online]. Available: http://www.ict4smartdg.eutc.org [16] COST Action IC0804, “Project IC0804 - Energy efficiency in large scale distributed systems.” [Online]. Available: http://www.cost804.org/ [17] The Common Information Model for Distribution: An Introduction to the CIM for Integrating Distribution Applications and Systems, 1st ed., EPRI, 2008. [18] Distribution CIM Modeling of Two North American Feeders, 1st ed., EPRI, 2009. [19] O. Samuelsson and N. Strth, “Islanding detection and connection requirements.” [20] M. O. de Mues, A. Alvarez, A. Espinoza, and J. Garbajosa, “Towards a distributed intelligent ict architecture for the smart grid,” in Proceedings of the 9th International Conference on Industrial Informatics (INDIN), 2011, in press. [21] A. Pe˜na and Y. K. Penya, “Distributed semantic repositories in smart grids,” in Proceedings of the 9th International Conference on Industrial Informatics (INDIN), 2011, in press. [22] A. Espinoza, M. O. de Mues, C. Fernandez, J. Garbajosa, and A. Alvarez, “Software-intensive systems interoperability in smart grids: A semantic approach,” in Proceedings of the 9th International Conference on Industrial Informatics (INDIN), 2011, in press. [23] IEC 61968 Application integration at electric utilities - System interfaces for distribution management. Part 11: Common Information Model (CIM), 1st ed., IEC, 2010.