The Standardization of Distribution Grid Communication Networks

Zhao Li, Member IEEE, Fang Yang, Member IEEE, Dmitry Ishchenko, Member IEEE

meters, sensors, and intelligent electronic device (IED)) and Abstract--This paper reviews the recent development in the system control through sending control commands to distribution utility communication networks including the controllable equipment. As transported messages carry advanced metering infrastructure (AMI) and the supervisory information of system status and/or control that are important control and data acquisition (SCADA) system. The to maintain the normal operation of the system, particularly standardizations of application level communication protocols in when the grid is hit by a large disturbance, the communication both AMI and SCADA systems are the major focus of this paper. network should ensure the instantaneous deliverability of In addition, some smart grid applications facilitated by the advances in the distribution communication networks are also important messages. If a control center fails to deliver or discussed. allows the delay of delivering such a message, it may lose the opportunity to govern a disturbance and potentially render the Index Terms-- Smart Grid, AMI, SCADA, Communication grid system unstable and unreliable. Hence, building a large Protocol. scale real-time communication network becomes one of important pre-requests for Smart Grid [4]. I. INTRODUCTION Nowadays, distribution utilities mainly adopt two types of communication networks: the supervisory control and data Promising solution for challenges presented by the acquisition (SCADA) system [5], [6] and the advanced globally rising energy demand, aging power system A metering infrastructure (AMI) [7]. In the past several decades, infrastructures, increasing fuel costs, and emerging renewable SCADA has played a key role in the online monitoring and resource portfolios is the smart grid [1]. Technically, smart control of the grid system. However, constrained by high costs grid applies the advanced monitoring and control technologies on constructing a real time communication system, SCADA to the electric grid, which enables sustainable options to has been mainly deployed in transmission system and a small customers and improves security, reliability, and efficiency for part of distribution system (e.g., distribution substations) for utility grid operation [2], [3]. real-time monitoring and control. In contrast, because of Since one primary goal of the smart grid is to improve relatively low cost on building a non-real-time communication energy efficiency, it is often referred to as the “green grid”. system and the stimulus from US government, AMI has been Two green features of the smart grid are that it maximizes the widely deployed in the distribution system recently, reaching utilization of renewable resources (e.g., solar and wind) and the feeder and residential levels. Rather than functioning as a improves the efficiency of system operation through loss monitoring and control network, AMI implements an reduction, dynamic pricing, and other applications. On the infrastructure to automatically collect energy usages from other hand, different from a traditional system in which power residential smart meters and transports them back to the generation and consumption can be well planned and control center on a monthly basis. It improves the efficiency estimated, in smart grid, the power generation and of the process of collecting residential energy usages and the consumption become more dynamic and unpredictable due to quality of collected measurements, and eventually enhances the intermittence of renewable resources and dynamic price the quality of customer service. driven power consumption. The unpredictable power In the past several years, efforts have been invested into building the large scale smart grid communication networks production and consumption complicates the grid reaching the feeder level or the residential level. These efforts management. To efficiently manage a grid system with these include extending the scope of SCADA and/or enhancing the dynamic features, utility control centers should be able to real-time capability of AMI. With the scope of the monitored monitor and control the grid in real-time. network reaching the feeder and residential level, the number Utility control centers implement system monitoring by of monitored data points is significantly increased, reaching collecting measurements from field devices (e.g., smart millions. Hence, improving efficiency and quality of the smart grid communication network and increasing interoperability Financial support from the ABB Corp. Research Center is gratefully between devices produced by various vendors become acknowledged. The authors appreciate the discussions and help from important topics and lead to the standardization of the colleagues Zhenyuan Wang and Xiaoming Feng. The authors are with the ABB US Corp. Research Center communication protocols and information models. 940 Main Campus Dr. Suite 200, Raleigh NC, 27606, United States This paper primarily reviews the efforts on standardization Email: {zhao.li, fang.yang, Dmitry.ishchenko} @us.abb.com of application level communication protocols in both SCADA

[9] and AMI [10] recently. The rest of this paper is structured support new requirements (i.e., demands responses) from the as follows: focusing on AMI infrastructure, the second section smart grid [11]. first introduces its components, and then discusses the The standardization of the AMI infrastructure includes the standardization of the AMI protocols. Similarly, the third standardization of both AMI communication protocols and section discusses SCADA and recent standardization efforts. AMI information models. Section four discusses some potential advanced smart grid Table 1 Communication modules supported by AMI vendors applications triggered by real-time and/or near real-time smart-grid communication networks. Section five concludes AMI Vendors Communication Modules this paper. Landis + Gyr Unlicensed RF, PLC Zigbee, unlicensed RF, Public Itron II. ADVANCED METERING INFRASTRUCTURE carrier network (OpenWay) Unlicensed RF, public carrier Elster network A. Background Echelon PLC, RF, Ethernet AMI consists of metering, communication, and data GE PLC, public carrier network, RF management functions, offering the two-way transportation of Sensus Licensed RF (FlexNet) customer energy usage data and meter control signals between Eka Unlicensed RF (EkaNet) customers and utility control centers. The AMI was originally developed from advanced meter reading (AMR) [12], [13], SmartSynch Public carrier network [14], [15], and [16], a one-way communication infrastructure Tantalus RF(TUNet) that implements automatic collection of meter measurements Triiliant ZigBee, public wireless network from residential smart meters to utility control centers for calculating monthly bills and fulfilling other related activities. Partially as the next generation of “AMR”, AMI not only The Standardization of AMI Communication Protocols enhances the traditional data collection functionality (i.e., In the past two years, the focus of the standardization of improving monthly meter data collection to real time or near AMI communication protocols has gradually shifted from the real-time meter data collection) but also develops the remote physical level (e.g., ANSI C12.18 [18]) and the device level control capability from the control center to smart meters. (e.g., ANSI C12.21 [20]) to the application level (e.g., ANSI Motivated by the economic stimulus plan of the U.S. C12.22 [21]) because application level communication government, most U.S. states have begun the process of protocols effectively isolate the details of underlying physical deploying smart meters within AMI infrastructures. At the network configurations and implementation. This section beginning of 2009, for example, Texas initiated a project of introduces two application level communication protocols that deploying six million smart meters and expected to complete are popular in both the United States (i.e., C12.19 and C12.22) it by 2012; and California plans to install 10 million smart and European markets (i.e., IEC 62056-53 and IEC 62056- meters by the end of 2012. The deployment of smart meters is 62). taking place not only in the United States but throughout the world. Based on current estimates, by 2015, smart meter ANSI C12.22 installations are expected to reach 250 million worldwide [17]. Hence, for most utilities, AMI will be a well-deployed Historically, after a set of standard table contents and feeder or residential level communication network, delivering formats were defined in ANSI C12.19, a point-to-point a large amount of data in real time or near real time (e.g., standard protocol (ANSI C12.18) that transported the table every 15 minutes). data over an optical connection was developed. Afterwards, the “Protocol Specification for Telephone Modem B. The Standardization of the AMI Infrastructure Communication” (ANSI C12.21), which allowed devices to transport tables over telephone modems was defined. The In the current market, smart meters from different vendors C12.22 standard, expanding on the concepts of both the ANSI are using proprietary communication protocols that are C12.18, and C12.21 standards, allows the transport of table generally non-interoperable (Table 1). For utilities, deploying data over any reliable networking communications system. millions of smart meters is a long-term investment indeed. The goal of the ANSI C12.22 standard is to define a Therefore, once a utility adopts smart meters from a certain meshed network infrastructure customized for AMI AMI vendor, it must follow up with related products from the applications. The standard contains the following same vendor for the sake of compatibility. functionalities: As utilities are reluctant to commit to a certain meter 1) Defining a datagram that may convey ANSI C12.19 data vendor, enabling interoperability between AMI products from tables through any network, including the AMI specific different vendors by following the same standards becomes an network and general purpose network (e.g., Internet) effective way to protect utilities’ investment. Table 2 lists 2) Providing a seven-layer communication infrastructure for popular standard communication protocols and meter interfacing a C12.22 device information models in the current market, defined by ANSI, 3) Providing an infrastructure for point-to-point IEC and IEEE. Most of them have recently been revised (e.g., communication that will be used over local ports (e.g., C12.18 and C12.19) or newly defined (e.g., C12.22) to optical ports and moderns) 3

4) Providing an infrastructure for efficient one-way messaging model (layer 7): an identification service, a read service, a The ANSI C12.22 mesh network consists of the C12.22 write service, a security service, a trace service, and others. nodes and network. A C12.22 node, a point on the network that attaches to a C12.22 network, is a combination of both a IEC62056 C12.22 device and communication module. The C12.22 communication module is a hardware module that attaches a IEC62056 defines the meter interface classes for the C12.22 device to a C12.22 network. The C12.22 device Companion Specification for the Energy Metering (COSEM) contains meter data in the forms of tables defined by C12.19. model through a series of standards on data exchange for The interface between the communication module and the meter reading, tariffs, and load control. Similar to ANSI device is completely defined by the C12.22 standard. C12.22, IEC62056-53, the application layer communication

Table 2 Popular standard information model and communication protocols protocol in the COSEM model, is defined based on several other IEC62056 series protocols, including IEC62056-21, 42, Time Name to Category Content Domain 46, and 47 [23]. Except for the IEC62056-21, which is used in Market hand-held devices for locally exchanging data with meters, the Gas, C12.19 Information Utility industry end device 2005 Water, remaining protocols define the various layers of the [19] model data tables Electricity communication network that support application level Protocol specification for ANSI Gas, C12.22 Communication Interfacing communication: the physical layer (IEC62056-42), the data 2008 Water, [21] Protocol to data communication Electricity link layer (IEC62056-46), and the transport layer (IEC62056- networks Protocol specification for Gas, 47). Similar to ANSI C12.22, the meter data carried by C12.18 Communication 2005 ANSI Water, [18] Protocol IEC62056-53 are defined by IEC62056-61 and IEC62056-62, Type 2 optical port Electricity which are dedicated meter data models in the IEC62056 Protocol for telephone Gas, C12.21 Communication 2005 modern Water, series. [21] Protocol communication Electricity C12.22 Device C12.22 Communication Model 61968-9 Information Meter data model in power Electricity 2009 [22] Model distribution system only

Gas, C12.19 Tables C12.19 Tables 62065-53 Communication COSEM application layer C12.22 EPSEM C12.22 EPSEM 2007 Water, C12.22 ACSE C12.22 ACSE

IEC [24] protocol protocol Electricity C12.22 Layer 7 C12.22 Layer 7 Gas, 62056-61 Information Meter object identification 2007 Water, [25] protocol system Electricity C12.22 C12.22 C12.22 Gas, 62056-62 Information Interface for data Layer 6 to 1 Layer 6 to 1 Layer 6 to 1 2007 Water, [26] protocol exchanging Electricity

Optical Port Gas, Communication 1701 [27] 2011 Communication Protocol Water, protocol (compatible with C12.18) Electricity To LAN/WAN/MAN Key: Telephone modern Gas, LAN – Local Area Network Communication communication WAN - Wide Area Network

IEEE 1702 [28] 2011 Water, protocol protocol (compatible with Electricity MAN – Metropolitan Area Network C12.21) Gas, P1377/D9 Information End device data tables 2011 Water, [29] Model (Compatible with C12.19) Electricity Figure 1 The open system interconnection (OSI) model defined by C12.22 LAN/WAN Node Gas, P1703/D8 Communication communication 2011 Water, As an application layer communication protocol, [30] protocol protocol Electricity (Compatible with C12.22) IEC62056-53 primarily offers three types of services in its application-level: the GET service (.request, .confirm), the The C12.22 network defines an AMI specific mesh SET service (.request, .confirm), and the ACTION service communication infrastructure that consists of one or more (.request, .confirm). C12.22 network segments (a sub-network) or a C12.22 LAN. Although both IEC62056-53 and ANSI C12.22 provide a Similar to the open system interconnection (OSI) model, the similar way to construct the advanced mesh AMI network, C12.22 communication protocol consists of seven layers each has a unique market focus: IEC62056 primarily focuses (Figure 1): an application layer (layer 7), a presentation layer on the European market while ANSI C12.22 focuses on the (layer 6), a session layer (layer 5), a transport layer (layer 4), a North American market. In the current North American network layer (layer 3), a data link layer (layer 2), and a market, most AMI vendors support C12.18 and C12.21, but physical layer (layer 1). Unlike OSI, C12.22 is customized few support C12.22 since it has only recently been defined. only for meter data transportation. For example, the Because of the advantages of C12.22 and/or IEC62056-53, we application layer of C12.22 supports only ANSI C12.19 predict that most major meter vendors will support either of tables, described by EPSEM and ACSE, the languages that them worldwide in the near future. encapsulate C12.19 meter data. The Standardization of AMI Information Model Supported by layers 1 through 6, which consist of various physical network connections in the meter industry as well as An information model is a representation of concepts, the standard Internet connection, the C12.22 standard defines relationships, constraints, rules, and operations that specify the following services in the application layer of the OSI data semantics for a chosen domain of discourse [32]. In the AMI communication infrastructure, it is necessary to have an 4 information model, by which all communication participants III. SCADA –THE UTILITY MONITORING AND CONTROL can semantically reach a certain level of understanding. NETWORK This section discusses major standard information models in today’s market: ANSI C12.19 and IEC62056-62. The A. Background former is supported by ANSI C12.22 and widely used in the Historically, the most matured utility applications having North American market; the latter is support by IEC 62056-53 close relationships with communication technologies is and widely deployed in the European market. telemetry and telecontrol, which are commonly fulfilled by a supervisory control and data acquisition (SCADA) system, a ANSI C12.19-2008 real-time communication network bridging end devices and utility control centers. Constrained by the high cost of ANSI C12.19 resulted from comprehensive cooperative construction, SCADA has traditionally been limited at the effort among utilities, meter manufacturers, automated meter transmission level and distribution system substations, only reading service companies, the ANSI, Measurement Canada monitoring and control important power system devices. (for Industry Canada), NEMA, the IEEE, Utilimetrics, and As SCADA has been in place since the late 1960s, it has other interested parties. Currently, it has two versions: ANSI transitioned through several generations of communication C12.19-1997 and ANSI C12.19-2008. As the latter is intended technologies and protocols. Initially, RS-232 and RS-485 to accommodate the concepts of the most recently identified were used as physical interfaces with Modbus, DNP, or IEC advanced metering infrastructure, it is the primarily focus of 60870-5-101 [36] as communication protocols in SCADA. this section. Since the late 1990s, Ethernet interfaces have become the de- The heart of ANSI C12.19 is a set of defined standard facto standard for most of the IEDs, and traditional serial- tables and procedures: The former are methods of storing the based protocols have begun to support Ethernet interfaces in collected meter data and controlling parameters, and the latter their physical layers, but the application layer of above are methods of invoking certain actions against the above data protocols is still serial fashion; for example, Modbus and IEC and parameters [19]. The standard tables in C12.19 are 60870 evolved into Modbus over TCP and IEC 60870-5-104 typically classified into sections, referred to as “decades”. [37]. During the transition, DNP3 has become the most Each decade pertains to a particular feature set and a related important utility communication protocol, particularly in US, function. Transferring data from or to an end device that and has been developed into an IEEE Standard, named IEEE adheres to the C12.19 standard entails reading or writing a 1815, [38] in 2010. particular table or a portion of a table. Even though the Unlike DNP3, evolved from traditional communication C12.19 standard covers a broader range of tables and protocols, IEC61850 is a newly defined network based procedures, it is highly unlikely that any smart meter will be communication protocol optimized for the mesh network. able to embed all of the tables or even a majority of those Beyond defining communication specifications, IEC61850 defined in ANSI C12.19. Hence, implementers are encouraged primarily delimits an information model that is compatible to to choose an appropriate subset that suits their needs. the common information model (CIM) [31], the widely The C12.19 standard is a general meter information model deployed information model in the power system domain. that serves various domains, including electricity, water, and Hence applications based on IEC61850 have strong gas using a series of tables, which can be customized through interoperability with most power system applications in state several standard operations. of the art. Nowadays, IEC61850 has been identified as one of the key IEC 62056-62 standards that lay the foundation of the future Smart Grid framework by the US Federal Energy Regulatory Commission Unlike ANSI C12-19, which uses tables to package meter (FERC) in 2010. In this section, DNP3 and IEC61850 are measurements, IEC 62056-62 models meter information briefly discussed. through a series of interface classes. As the information modeled by C12.19 and IEC 62056-62 are nearly identical, we B. Distributed Network Protocol 3 (DNP3) do not duplicate our efforts to further introduce the content of Developed and maintained by the DNP3 Users Group, an IEC 62056-62. Similar to ANSI C12.19, as a general meter open community for vendors and utilities, DNP3 is recognized data model, IEC 62056 supports not only electricity meters as an important utility communication protocol, particularly in but also gas and water meters. the US market. Structurally, even though DNP3 supports For AMI vendors, the preference to support certain Ethernet in its physical layer, it still keeps the serial nature in standards reveals a strong geographical bias. For example, its application layer. Hence, from application’s point of view, most smart meter vendors in the North American market are DNP3 is still a serial communication protocol. more likely to choose ANSI series standards (i.e., C12.19 and As saving communication bandwidth was one of the most C12.22) while those in the European market are more likely to critical goals that the DNP series protocols were trying to select IEC standards. achieve, DNP3 is suitable for situations where communication Triggered by the rapid development of the smart grid, resources (e.g., bandwidth) are limited. Structurally, DNP3 is beyond supporting proprietary communication protocols, most constructed based on a simplified OSI model, including only AMI vendors have begun to support the standard three layers: the application layer, the data link layer, and the communication protocols and meter data models. physical layer. 5

Figure 2 demonstrates the master-outstation communication the application layer confirmation from the master is received model of DNP3 [38], which includes a master or the final confirmation timeout is expired. device/information collectors (e.g., an RTU or SCADA C. IEC 61850 master) and one or more outstation devices/end devices (e.g., IEDs) connected to each other using serial or Ethernet links. IEC61850 is maintained by the International Electro- Here, inputs are values collected by outstation and outputs are technical Commission Technical Committee 57. Technically, commands sent by SCADA masters. Generally, an outstation IEC61850 has been developed based on utility contains several arrays of data points in different types (e.g., communications architecture (UCA), an open communications binary and analog), mapping to various internal device standard for the electric power utilities defined by Electrical parameters in the outstation firmware. Power Research Institute’s (EPRI) in the early 1990s. Rather than having a formal information model, DNP3 IEC61850 was originally designed for exchanging defines the following classes to organize data: Class 0: Static information within substations. Afterwards, it expands data – the snapshot of the outstation current point values; concepts and principles beyond the substation fence. Classes 1, 2, 3 – the historical point data from the outstation Unlike DNP3, IEC 61850 consists of a series specifications event buffer. Semantically, the above classes are mapped to a [39], primarily focusing on the following aspects: the range of data points in outstations. information model and communication structure (IEC61850- 7), the mapping of communication service (IEC61850-8 and IEC61850-9), and configuration description (IEC61850-6). Located in the heart of IEC61850 is information modeling part (IEC 61850-7). To some extent, IEC61850 is more like an information model, modeling the information exchanged between devices in a substation. Representing a logical function performed by a physical device, a Logical Node (LN) is a basic information unit in IEC 61850, the semantic of which is represented by data and data attributes (Figure 3: the LN for a circuit breaker [39]). A group of LNs constructs a Logical Device (LD), representing various functions performed by a physical device (e.g., protection, control, and monitoring). Generally, a physical device consists of one or more LDs. In 61850, over 100 LNs are defined, covering the functionalities across protection, monitoring, and control of a grid system. Logical node

XCBR

Figure 2 The communication model of DNP3 [38] Data Data- Controls Attributes

- As reducing the network traffic is one major focus of Pos controllable Control value ‘ctlVal’ Operate time control DNP3, it introduces the concept of deadband, allowing Originator Control number status value generation of events based on monitoring the analog inputs Status value ‘stVal’ Quality against thresholds (deadbands). An analog input only Time stamp status ... generates an event when changes of its absolute value exceed Subst. enable Subst. value substitution a predefined threshold (deadband) so that the small ... fluctuations of the analog value mapped to the analog input Pulse configuration Control model configuration, are filtered out and will not be reported. SBO timeout description SBO class and extension In DNP3, SCADA masters can poll data from outstations ... by sending a request to the outstations. Meanwhile, the outstations can push data back to a master through a Figure 3 Hierarchical structure of IEC61850 object model mechanism of spontaneous unsolicited reports. During the [39] pushing process, the outstation initiates the reporting process when it accumulates a pre-determined number of class events. In IEC61850, the object models (e.g., LN and LD) can be In addition, to save the bandwidth, DNP3 supports the report- accessed through abstract communication service interfaces by-exception (RBE) function, by which the outstation only (ACSI). Some typical services provided by ACSI include the reports the data that has changed since the last poll. The RBE following: function significantly reduces the amount of information that (1) The SETTING-GROUP-CONTROL BLOCK, switching must be transferred, saving the bandwidth. the field device from one set of preconfigured control To guarantee the information delivery, the following parameters to another set strategy is adopted by DNP3: if the original unsolicited (2) The REPORT-CONTROL-BLOCK, LOG-CONTROL- reporting attempt failed, the outstation initiates a retry until BLOCK, reporting and logging the event data, including 6 immediate real-time reporting, polled reports, and integrity discusses. scans A. Smart grid communication system (3) A generic substation event (GSE) and a generic object oriented substation event (GOOSE), enabling fast peer-to-peer From information point of view, CIM, including IEC61970 communications among the field devices and allowing the and IEC61968, is the widely applied information model in the implementation of de-centralized control schemes based on grid management territory. Most power system applications local intelligence are built on top of it. Therefore, the information collected by (4) Control class model, performing the control actions, the AMI and SCADA has to be translated to CIM before supports both direct control and select-before-operate control being consumed by these applications. For the grid with enhanced security management system, information sent from AMI to the grid The network interface maps the services defined by ACSI management system has to be translated from C12.19 to to actual network protocols. The protocol profiles of IEC IEC61968-9, the newly defined meter data model in CIM; 61850 are defined by the Manufacturing Message while the information sent from SCADA has to be converted Specification (MMS) protocol, the Internet, and OSI protocol into IEC61970 and/or IEC61968. On the other hand, stacks. In addition, the Ethernet (IEEE 802.3x) LAN is also information sent from the grid management system to AMI specified for high-speed protection functions, such as GOOSE and/or SCADA has to be changed from CIM to C12.19 and/or and the processing of digitized waveforms of high sampling IEC61850 separately. rates (Sampled Values). The infrastructure of 61850 is Beyond accommodating different types of information summarized in Figure 4. models, we have to consider adapting the communication protocols when integrating AMI and SCADA with the grid Information Model IEC 61850-7-4/-7-3 Logical Device, Logical Nodes, Attributes, and Data management system. For example, to receive/send information from/to AMI or SCADA, the grid management system needs Abstract Communication Service Interface to have a C12.22 interface or IEC61850-8 interface. To (i.e., SETTING_GROUP_CONTROL_BLOCK, GOOSE) IEC 61850-7-2 illustrate the above integration process, the information integration layer is proposed in Figure 5. Network Interface IEC 61850-8-1 Mapping to MMS, Ethernet (TCP/IP) Smart Grid Communication System Grid Management System Physical Link Key: SM1 MMS Manufacturing Message Specification DR OMS SM2 Figure 4 The overall structure of IEC61850 Information Integration AMI C12.22 C12.19 In summary, DNP3 is a serial communication protocol: DSE even though supporting Ethernet in its physical level, DNP3 SMn still maintains the serial manner in its logical level. Compared IED1 to IEC61850, DNP3 has less communication overhead and is SCADA CIM more suitable for low bandwidth network in practice. In IEC61850-7 IEC61850-8 Common Information Model IED2 CIM contrast, IEC61850 has a well-defined information model that IEC61968-9 is compatible with CIM, the most popular information model IEC61970 Key: used in the power system domain. Hence, IEC61850-friendly IEDn SM: Smart Meter IED: Intelligent Electronic Device devices have less overhead in semantic translation when DR: Demand Respond OMS: Outage Management System integrated with most power system applications. However, DSE: Distribution State Estimation transporting the complex semantics modeled by IEC61850 consumes a large amount of network resource, requiring a Figure 5 Information models and communication well-constructed communication network in the physical protocols in landscape of smart grid layer. Different from CIM, either C12.19 or IEC61850 is an information exchange model, which only models exchanged IV. SMART GRID COMMUNICATION NETWORK AND SMART information. Theoretically, information contained in C12.19 APPLICATIONS and IEC61850 is only a subset of the information contained in CIM, primarily focusing on supervision and control of a grid Figure 5 illustrates the landscape of the standard-based system. As discussed, the bright side for using the standard smart grid system, including the smart grid communication information model is to reach a certain level understanding system and the grid management system. The communication between participants. However, compared to the traditional system includes AMI and SCADA; while the grid applications modeless transportation, a XML-based information package includes some advanced distribution management applications of C12.19 and IEC61850 leads to heavy burden for the enhanced by real-time or near real-time communication communication system because the XML-based transportation network. These two parts are discussed in this section package includes not only values but also tags associated with these values. As current grid systems still possess a lot of 7 legacy equipment and communication facilities, most of them network in utilities, SCADA cannot easily be extended further are incompatible with the newly defined standards. For to the feeder and the customer levels because of its limited example, some smart meters from certain vendors only bandwidth and the serial-based communication infrastructures support proprietary information models and communication which cannot support increased data points caused by such an protocols. Hence, even though the discussed information extension. As a complementary to each other, SCADA and models have many advantages, the deployment of these AMI systems will coexist for many years to come. standards does not occur immediately. To protect the utility investments and enable the interoperability among devices from various vendors, many B. Advanced Grid Management System Applications new standards of communication protocols have been defined The online monitoring and control functions enabled by the for both AMI and SCADA lately to replace legacy proprietary real-time or near real-time smart grid communication communication protocols. The new standards outlined not networks can further enhance many grid applications. Some of only the standard communication infrastructure (e.g., mesh these applications are discussed as follows: network infrastructure defined by ANSI C12.12) but also the (1) DSSE: the AMI system with millions of meters located at standard information models (e.g., ANSI C12.19 and customer sites can provide a large number of customer level IEC61850). Even though the newly defined communication measurements. Based on AMI as well as SCADA data, DSSE protocols have more advantages than legacy protocols, can calculate the best estimation of system operating states because of a large number of legacy facilities currently that servers the foundation of many other applications. The running in grid systems, the legacy communication protocols high performance of the AMI communication system is (e.g., the serial communication protocol) that are supported by critical to provide real-time or near real-time measurements these facilities will continue to exist in a relatively long period for DSSE to facilitate the timely and accurate system state of time. estimation. (2)OMS: when a fault occurs in the distribution system, the VI. REFERENCES smart meters in the outage area send out the last gasp message [1] Cupp, J. and Beehler, M.E., “Implementing Smart Grid Communications”, that carries precise outage information to the utility control TECHBriefs, No. 4, 2008. center through the AMI communication network which is [2] Apel, R., “Energy Automation, enabling the intelligent grid,” Next Generation Utilities: Technology Summit, Evian, France, 10/02/2008. generally faster than the customer trouble calls. In addition, [3] Dugan, R. C. and McGranaghan, M., “Sim City”, IEEE Power & Energy the two-way communication channels of the AMI system also Managine, Vol. 9, Num. 5, Sept./Oct. 2011 allow utility control centers to conduct on-demand polling of a [4] Roy, S, Nordell, D., and Venkata, M. “Lines of Communication,” IEEE Power & Energy magazine, Vol. 9, Num. 5, Sept./Oct. 2011 meter status for the purpose of outage confirmation. [5] Girotti, T.B. Tweed, N.B. Houser, N.R. “Real-time VAr control by (3) DR: during a peak load interval, demand response is one SCADA”, IEEE Transactions on Power Systems, Vol. 5, Num. 1, Feb effective means that utility can adopt to avoid system 1990 [6] Marihart, D.J. “Communications technology guidelines for EMS/SCADA overloading. In a DR event, a utility usually requires some systems”, IEEE Transactions on Power Delivery, Vol. 16, Apr. 2001 load shedding or shifting to reduce the total demand at peak [7] Bennett, C., Wicker, S.B. “Decreased time delay and security enhancement load. The successful execution of a DR event depends on an recommendations for AMI smart meter networks”, Innovative Smart efficient two-way AMI communication system between the Grid Technologies (ISGT), Jan. 2010. [8] Cassel, W.R., “Distribution management systems: functions and payback”, utility and DR customers, based on which the utility can poll IEEE Transactions on Power Systems, Vol. 8, Issue 3, 1993 meter data from the customers and transmit DR signals to the [9] “Supervisory Control and data Acquisition (SCADA) systems”, Office of DR customer meters. the manager, National communication system, October 2004 [10] Yang, F. “Advanced Metering Infrastructure Technology,” Technical The example advanced applications listed above have one Report, No. PT-07045, ABB US Corp. Research Center, 2007. common feature: their implementation largely depends on [11] “NIST Framework and Roadmap for Smart Grid Interoperability real-time or near real-time information transmission provided Standards”, Release 1.0 (Draft), Office of the National Coordinator for by communication networks. In general, the high performance Smart Grid Interoperability, NIST, 2009. [12] Fischer, R. A., Laakonen, A. S. and Schulz, N., N., “A Generation of AMI and SCADA system plays a critical role in facilitating Polling Algorithm Using a Wireless AMR System for Restoration the advanced smart grid applications. confirmation”, IEEE Transactions on Power Systems. Vol. 16, No. 2, pp. 312-316, 2001 [13] Dorey, H., “Advanced Metering in Old and New Worlds,” Power V. CONCLUSIONS Engineering Journal, Vol. 10, No. 4, pp.146-148, August, 1996. This paper briefly reviews the recent development of [14] Y. Jin, and M. D. Cox, “A Pipelined Automatic Meter Reading utility communication networks, including AMI and SCADA. Scheme,” Instrumentation and Measurement Technology Conference, pp. 715-720, 1993. As a large-scale communication network in the utility, the [15] Mak, S. and Radford, D. “Design Considerations for Implementation of AMI system has recently received considerable attention. Large Scale Automatic Meter Reading Systems,” IEEE Transactions on Nowadays, for many utilities, the deployment of the AMI will Power Delivery, Vol. 10, No. 1, pp. 97-103, 1995. enable real-time or near real-time monitoring and control [16] Clay, M. R. J. and McEntee, A. J. “Advanced Meter Reading Tokenless Prepayment,” Power Engineering Journal, Vol. 10, No.4, pp. 149-153, capabilities in the customer level and accelerates the August 1996. emergence of advanced grid management applications which [17] “SmartMeter™ Installation Progress,” PG&E, April 2010, improve the efficiency of the grid system operation. On the http://www.pge.com/myhome/customerservice/meter/smartmeter/deploy other hand, as the legacy real-time monitoring and control ment/ 8

[18] ANSI C12.18-2006, “American National Standard Protocol Specification for ANSI Type 2 Optical Port”, ANSI Inc. 2006. [19] ANSI C12.19-2008, “American National Standard - Utility Industry End Device Data Tables”, ANSI Inc. 2009. [20] ANSI C12.21-2006, “American National Standard Protocol Specification for Telephone Mode”, ANSI Inc. 2006. [21] ANSI C12.22-2008, “American National Standard - Protocol Specification for Interfacing to Data Communication Networks”, ANSI Inc. 2009 [22] IEC61968-9: “Application integration at electric utilities - System interfaces for distribution management - Part 9: Interfaces for meter reading and control”, IEC, 2009 [23] IEC62056: Electricity metering – Data exchange for meter reading, tariff and load control, IEC62056 [24] IEC 62056-53: Electricity metering – Data exchange for meter reading, tariff and load control – Part 53: COSEM application layer, IEC 2006. [25] IEC 62056-61: Electricity metering – Data exchange for meter reading, tariff and load control – Part 61: Object Identification System, IEC 2006. [26] IEC 62056-62: Electricity metering – Data exchange for meter reading, tariff and load control – Part 62: Interface classes, IEC 2006. [27] IEEE 1701-2011: IEEE Standard for Optical Port Communication Protocol to Complement the Utility Industry End Device Data Tables, IEEE, 2011 [28] IEEE 1702-2011: IEEE Standard for Telephone Modem Communication Protocol to Complement the Utility Industry End Device Data Tables, IEEE, 2011 [29] IEEE P1377/D9, IEEE Draft Standard for Utility Industry Metering Communication Protocol Application Layer (End Device Data Tables), IEEE, 2011 [30] IEEE P1703/D8, IEEE Draft Standard for Local Area Network/Wide Area Network (LAN/WAN) Node Communication Protocol to complement the Utility Industry End Device Data Tables, IEEE, 2011 [31] IEC61970, Energy management system application program interface, part 301, common information model (CIM) base, IEC, 2011 [32] Lee, Y. T. "Information modeling from design to implementation" National Institute of Standards and Technology, 1999. [33] Venkata, S.S., Uluski, R.W. and McGranahan, M. “Critical Elements – Distribution Management Systems”, IEEE Power and Energy Magazine, Vol. 9, No. 5, September/ October 2011. [34] Clarke, G., Reynders, D. “Practical Modern SCADA Protocols: DNP3, 60870.5 and Related Systems”, Elsevier, 2004. [35] Olovson, H.-E., Werner, T., and Rietman, P. “Next Generation Substations. Impact of the process bus”, ABB Review, Special Report IEC 61850, 2010. [36] IEC 60870-5-101: Telecontrol equipment and systems - Part 5-101: Transmission protocols - Companion standard for basic telecontrol tasks, Second edition, IEC, 2003-02. [37] IEC 60870-5-104: Telecontrol equipment and systems – Part 5-104: Transmission protocols – Network access for IEC 60870-5-101 using standard transport profiles, Second Edition, IEC, 2006-06. [38] Distributed Network Protocol (DNP3). IEEE Standard 1815-2010, July 2010. [39] IEC61850, Part 1~10, International Standard, Communication networks and systems for power utility automation, 2003-04.