Creating new ICT networks on Smart Grids service

Transcription

1 Creating new ICT networks on Smart Grids service ( Laura Corbeira (Endesa Servicios) Guillermo Ravera (La Salle Universitat Ramon Llull) Novel solutions of the information and communication technology as the backbone of Smart Distribution Darmstadt, April 2011

2 Outline The INTEGRIS Project Project objectives Requirements Examples of Smart Grid Use Cases ICT Challenges Ideas/hints for possible solutions 2 2

3 The INTEGRIS project Project full title: INTelligent Electrical Grid Sensor communications Grant agreement no.: Date of initiation: 1st February, 2010 Duration: 30 months The research leading to these results has received funding from the European Union European Atomic Energy Community 7 th Framework Programme (FP7/ FP7/ ) under grant agreement n

4 INTEGRIS Objectives INTEGRIS is a novel and flexible ICT system, based on the integration of PLC and WSN technologies, able to completely and efficiently fulfill communication requirements for future Smart Grids. BASIC OBJECTIVE : To define and develop an integrated ICT environment able to efficiently encompass the communications requirements that can be foreseen for Smart Grids. SCIENTIFIC AND TECHNOLOGICAL OBJECTIVES : Research on the efficient integration and interoperability of PLC and wireless technologies (WSN, IEEE802.11n, RFID) Research and development of an autonomous self healing ICT system with QoS guarantees for Smart Grids. To contribute to the development of IEC technology within the electrical distribution network OTHER RESEARCH LINES: Multilevel security framework Application of Distributed Systems techniques to Smart Grids Application of Cognitive techniques to Smart Grids 4

5 Requirements A Smart Grid needs a wide communications network, robust and flexible. The requirements of the communications network depend on the applications. FUNCTION TYPE FUNCTION CLASS SEN mangement function Advanced meter function Active Protection Functions (@ HV/MV MV/LV level) Command Control & Regulation Monitoring & Analysis Advanced Meter & Supply Management function (Commercial functionalities Active Demand End to End Information Exchange and DR functions Management CLASS MODE Value/signal APF CMD Block & trip Signal O/C command Load shedding Peak shaving MON Analogical & Digital TVPP AMS IEM Energy meas., Supply mngt. Command, Alarm signals Energy meas., CVPP/Load Other signals Transfer time (*) <=20ms <=2s >=2s <=5m <=10s <=5m <=5s Availability in Grid State Normal Perturbed Crisis Normal Perturbed Crisis Normal Perturbed Crisis Normal Normal Perturbed Reliability level Very High High High Low Medium 5

6 Smart Grid Use Cases 1/2 INTEGRIS project will design Smart Grid applications for the power distribution network, and will implement some of them in field trials, in order to proof the validity of the developed ICT infrastructure. 6

7 Smart Grid Use Cases 2/2 INTEGRIS project will design Smart Grid applications for the power distribution network, and will implement some of them in field trials, in order to proof the validity of the developed ICT infrastructure. 7

8 ICT challenges in Smart Grids 1/2 The future of Electric System comes from the construction of an adaptive, optimized, integrated and distributed Intelligent Distribution Network, interactive with consumers and markets New kind of services: Services with very low delay and very high availability at the same time Services difficult to be amenable to flows Services with no connection establishment (Always connected) Wide array of requirements Services over L2 (Goose messages) Extending the concept of Smart Grids to distribution networks: Distribution networks are complex, spread over the territory and, in great part, buried or underground. The topology of the power network is different in each country. Difficulty of designing an ICT infrastructure spreading all over the Power distribution network having the required high availability and low delay. 8

9 ICT challenges in Smart Grids 2/2 The future of Electric System comes from the construction of an adaptive, optimized, integrated and distributed Intelligent Distribution Network, interactive with consumers and markets Networking challenges: Self healing, QoS for Smart Grids, long life time of electrical infrastructures, Heterogeneous communications network with several technologies Great number of legacy protocol already deployed and that have to be maintained Security challenges: Secure communications Backwards compatible with devices without any provisions for security 9

10 Some answers to the challenges 1/2 Smart Grids need a spread communications network, with a low cost of deployment, and a distributed ICT system, capable of dynamically adapting to the medium High availability: Meshing the communications system Dual homing Locate storage and computing platforms on the distribution network Replicating data in different platforms Distributing applications Low delay: Networking at MAC layer Spreading the MAC layer all over the distribution segment (MV+LV) Fast recovery/self healing 10

11 Some answers to the challenges 2/2 Smart Grids need a spread communications network, with a low cost of deployment, and a distributed ICT system, capable of dynamically adapting to the medium Buried or underground infrastructure: PLC/BPL necessary Services difficult to be amenable to flows: New concepts to control QoS Always connected services: Need to share the resources fairly No connections: Planning/dimensioning for some services plus prioritization or class based service for the rest Goose messages: Provide a way to create a L2 network over the distribution segments. Interconnecting different L2 technologies to provide a L2 network capable of spanning the distribution segment under consideration. Complex networks: PLC/BPL ( follow closely the electrical infrastructure) Different communication technologies (PLC, Wireless, Fiber Optics,..) An ICT system covering a distribution segment can be considered a new kind of ITU T NGAN 11

12 Networking and QoS Lines being explored: Forming a super L2 heterogeneous network BPL+Wireless TCP/IP INTEGRIS Device I-NMS TCP/IP Internet BPL network TCP/IP DMS Wireless network LR WSN TCP/IP BPL network LR WSN NB PLC RFID 12

13 Example of a network MV I Dev PLC HE MV MV PLC R MV/LV PLC CPE PLC HE I Dev STA MV MV MV/LV I Dev PLC R PLC R STA PLC CPE PLC HE MV PLC R Possibility Wireless AP PLC R LV LV LV I Dev LV I Dev PLC R STA STA PLC R Wireless It can be a mesh or direct link 13

14 Security 1/3 Security for standard IEC is addressed by standard IEC IEC mandates the use of TLS to assure end to end security Challenge 1: The Smart Grid needs intermediate storage and computing platforms. IEC currently does not offer application layer end to end security if multiple transport layer connections are used. Trusted TLS proxies may be a solution for the time being but not the solution. This way may be a weakness. For some Other challenges in Smart Grid security are: NISTIR 7628 Guidelines for Smart Grid Cyber Security (three Volumes) Finally, it is interesting to realize that currently most electrical devices do not handle any level of security. 14

15 Security 2/3 Just a reflection: Security in the Smart Grid is quite different from Office cases Source: Steffen Fries*, Hans Joachim Hof*, Maik Seewald; Enhancing IEC to Improve Security for Energy Automation in Smart Grid Environments, 2010 Fifth International Conference on Internet and Web Applications and Services, Barcelona, Spain 15

16 Security 3/3 16

17 Distributed System Techniques Lines being explored: Data replication in different platforms to improve data reliability and fault tolerance. Replication in circles around the primary repository with decreasing consistency. However, replication increases the number of messages transmitted and could reduce the system throughput if not properly engineered. 17

18 MV cubicle Global INTEGRIS architecture Operator eyes & hands console HMI Control System (MV view) DMS Others EMS OMS AMI NMS As it is defined now, it seems unovidable that each Use Case message has to be translated between pairs of protocols at each integris device (DLMS; 61850,104, modbus and so on). These add complexity and delay to INTEGRIS, affecting the performance of applications. To mitigate the problem, the integris device will have to be defined in a centric way. This means that the rest of protocols will suffer translation delays but the native equipment will suffer less delay because they don't require the translation. Includes All MV UseCases MV App LV App TCP/IP Includes All LV UseCases WAN IEC TCP/IP INTEGRIS NMS TCP/IP INTEGRIS MV SYSTEM DB router (off-the-shelf) ICT app n ICT app 1 DB manager configuration (mesh + QoS) cognitive agent security agent middleware The PLC connection can be also wirelss because both support ethernet as Layer 2 PLC (modem) HE The different IDEV can make a mesh using PLC or Wireless. The PLC network will be formed by CPE, HE and Repeaters and the Wifi network will be formed by AP and so on. Nevertheless, in the diafram is only represented one element of the PLC networks. PLC HR Wireless HR Wireless PLC MV_01.a Topology Changes Propagation MV_01.b Automatic Selectivity Reconfiguration MV_04 Voltage regulator MV-IDev middleware middleware MV-IDev MV_01.a Topology Changes Propagation MV_02 Faul MV_03 Early location, isolation abnormal and restoration conditions detector MV_02 Faul location, isolation and restoration MV_03 Early abnormal conditions detector configuration (mesh + QoS) cognitive agent security agent DB manager DB DB manager security agent LV App cognitive agent DB configuration (mesh + QoS) MV_01.b Automatic Selectivity Reconfiguration MV_04 Voltage regulator LR WSN RFID LR WSN RFID This port has to understand all these protocols?, make a translator is not an objective. Is there in the market some translator that can suite, for example, DLMS to 61850? LR WSN DLMS LR WSN FTP MV Smart <<includes>> meter <<includes>> IEC HE INTEGRIS LV SYSTEM PLC (modem) MV PQ monitor STATCOM Controller PLC IEC Analog Input Schneider modbus ->61850 gateway (G3200) serial MODBUS serial MV cubicle controller Position Measures Overcurrent Fault Location and Protection Alarm sensors can bus CAN Open can bus LV_01 LV PQVI LV_02 LV PQ Monitoring Monitoring LR/HR PLC LV_04 LV Voltage regulator LV_05 Customer Isolation LV IDev DB DB manager middleware configuration cognitive agent (mesh + QoS) security agent Digital Input/Output STATCOM <<send>> <<use>> IEC data model based XML The MV cubicle is considered an unique device CT VT Breaker Switch This device is fromed by the gateway and the MV cubicle controller wich is formed by the Protection devices and the Fault locator LR/HR PLC LV Customer gateway LV Customer gateway may be a smart meter (i.e. same device as LV Smart Meter) or other type of gateway device like Endesa's Proxima. Reading agent Writing agent DLMS/COSEM Leyend <<trace>> HTTP <<use>> HE: Head End HR Wireless: High Rate Wireless LR WSN: Low Rate Wireless Sensor Network RFID: Radio Frequency IDentifier PLC: Power Line Communications LR PLC LV Smart meter LV App1 LV App5 HOME ENERGY MANAGEMENT SYSTEM LV Home automation LV App4 Home energy management system is an independent system. It should have an interface towards Integris LV system. <<use>> LV App1 LV Modbus to IEC gateway (G3200) Modbus LR WSN <<includes>> VT CT Breaker LR WSN <<send>> z-wave <<send>> LR PLC <<use>> PV Inverter Modbus Controller Modbus LV PQ monitor LV Smart meter LR WSN LV App2 LV App1 LV App5 Load control switch PV Inverter (power electronics) <<use>> <<includes>> WSN PV LV DER VT CT Breaker

19 European FP7 project INTEGRIS - INTElligent GRId Sensor communications 19