By Margaret Goodrich, Jay Britton, Terry Saxton and Pat Brown
While the need for interoperability between utility systems has always been with us, what is it about today’s environment that makes it such a hot topic? One factor is the move from centralized to distributed generation. Another is the aging of electric utility assets past their planned lifetimes.
Deregulation of the power industry, coupled with the smart grid as a transformative movement, has also significantly contributed to the overall complexity by expanding (1) the number and variety of active participants; (2) the types of energy sources; and (3) the kinds of active business processes. As a result, power grid operations have become dramatically more dependent on complex computer-based analytically-intensive operating practices.
All of these changes have led to the need for massive amounts of data exchange, from systems/applications supplied by many different vendors within an enterprise (horizontal) to distributed energy resources (DERs) and microgrids and intelligent devices at the edge of the grid (vertical). Add to that the peer-to-peer and hierarchical information exchanges between distribution and transmission system operators and regional transmission operators/independent system operators (RTO/ISOs).
One can only wonder if there is any solution that can scale to deal with these massive numbers of interconnections between utility entities.
The Canonical Strategy
Many utilities have recognized that if they are going to successfully tackle this level of operational complexity the first step is to foster a data-driven culture. A future-looking system framework based on standards and a comprehensive enterprise semantic model can act as a common language for data exchange across the utility enterprise, facilitating system interactions. This is known as a canonical data model strategy for defining interfaces.
The International Electrotechnical Commission (IEC) Common Information Model (CIM) standards, specifically the IEC 61968, 61970 and 62325 series, were designed to provide such a solution through a three-layer framework comprising (1) a UML-based information model; (2) a set of business context-oriented interface profiles for defining specific information exchange contents based on a subset of the overall information model; and (3) message syntax and implementation technologies for serializing data messages and files. Following are sections that illustrate how these CIM standards can be successfully applied in practice to achieve interoperability.
Each series of CIM standards provides interoperability support using the three-layer framework for a specific area of the utility industry: IEC 61968 covers the functions related to the support operations of the electric grid; IEC 61970 covers the functions related to the operation and planning of the electric grid; and IEC 62325 covers the functions related to the energy markets. The primary functions discussed in this article surround the operation and planning as well as the support operations of the electric grid.
The initial CIM canonical model was created to allow data exchange and interoperability between different energy management systems (EMS). It was discovered, however, that this model could be used to exchange many different kinds of data for input into different network analysis systems. Each of these exchanges, once defined, became known as a profile and the CIM became a mechanism to exchange data to execute power flows, topology processing and state estimation.
CIM for Network Analysis
Efficient, reliable operation and planning of today’s grid depends heavily on computer-based network analysis. Modern network analysis algorithms are used in the planning, protection and operations domains of today’s utilities and they demand accurate models of the grid. At present, there are too many manual steps in the network model maintenance process, requiring too much valuable engineering time. CIM standards for power grid modeling can automate most of these data processing steps, improving both efficiency and accuracy.
A typical analytical model requires datasets that originate from many different sources. As illustrated in Figure 1, each source must make the required data available in a form that is not only internally correct, but which “fits together” with the data supplied by each of the other sources-forming a sort of building block approach to model assembly.
|FIGURE 1: CIM model part building blocks are assembled into cases|
Many different kinds of models are also required (by different analytic participants), but at the same time, much of the data in these models is the same from one model to the next. For example, a particular transformer should be represented in the same way in any model that requires that transformer. As Figure 2 shows (on page 18), CIM standards provide a building block approach that modularizes data to maximize sharing of data building blocks and minimize duplication of data preparation effort.
There are various departmental sources of data within the utility (illustrated on left side of Figure 3) that are focused on design, construction and maintenance of owned facilities and are the sources of record for the data about the physical system. The electrical grid scope, on the other hand, is focused on assembling all the data in the form required for algorithmic analysis of the electrical grid. And, because electricity does not respect ownership, this function has to include all data that is important to the particular kind of analysis being performed.
|FIGURE 2: Cases assembled from basic set of model part building blocks are supplied to multiple analytics|
A network model management (NMM) function maintains CIM building block datasets that can be used to assemble analytical models of all types (Figure 3). It develops building block datasets representing the physical grid by importing information from engineering design repositories and from external entities such as regional authorities, generators or DER providers. It develops nonphysical building block datasets from other sources, such as energy forecasting, scheduling and markets. It is a hub of activity supporting imports from many sources and outputs to many consumers, all governed and automated through CIM interactivity standards.
|FIGURE 3: Overview of an NMM-based CIM information architecture for power grid modeling|
CIM for Support Functions
While the base canonical CIM addresses the needs of network analysis, the utility industry has broader data exchange requirements, specifically in the realm of operations support. These areas are less dependent on network analysis but equally dependent on interoperability with various enterprise systems needing to cooperate, such as distribution automation, outage management, customer management, asset management, work management and others.
When integrating these systems, the data to be exchanged is still called a profile but at this level it is often a message rather than a model. These messages are much smaller but they require more frequent interaction with the sending and receiving systems. It is important, therefore, not only to define what data is exchanged but how it will be exchanged. The sequence of messages requires a standard envelope for the payload and an enterprise messaging infrastructure to carry and exchange the messages between the systems.
The payload is the information contained in the message that is of interest to the receiving application. These payloads are the profiles that are derived from the CIM canonical model in the form of XML schemas (XSD Message). While these payloads can use a variety of transport systems, there needs to be a standard to define:
• A common message envelope,
• A standard header to describe the nature of the message,
• A basic set of integration patterns that will define how the messages are to be conveyed, and
• Recommendations regarding the transport technologies that could be used.
|FIGURE 4: The IEC 61968-100 Message Envelope|
The above elements are all defined in the IEC 61968-100 standard. Figure 4 depicts the IEC 61968-100 message envelope. The header contains information such as the verb, noun, a time stamp, the ID of the source and other routing and processing information that will be needed to ensure interoperability during the interactive communications and is always required. The integration pattern defines which other sections in the message are to be used.
Using the basic components and building blocks of the IEC CIM series of standards, a complete framework for integrated grid interoperability is provided, whether the need is for full model exchange to accomplish network analysis or for integration of disparate support systems within the utility.
Editor’s Note: These four authors will be presenting more on this topic during Utility University at DistribuTECH Conference and Exhibition. Their course will run from 8 a.m. to 5 p.m. on Monday, Jan. 30, in Room 24C at the San Diego Convention Center. DistribuTECH runs through Feb. 2. See distributech.com for more information.
Margaret Goodrich is president and principal consultant for Project Consultants LLC. Jay Britton is principal consultant for Britton Consulting LLC. Pat Brown is principal technical leader for ICT/Electric Power Research Institute. Terry Saxton is vice president and co-founder of Xtensible Solutions and co-chair for CIMug.