Data - The Power Behind the Smart Grid
- Written by Jeffrey Katz
If it's to be a boon rather than blight, the flood of data generated in the smart grid will need to be handled in standardized frameworks, providing not only for interoperability but for ongoing evaluation of possible conflicts between different component systems. Done right, the results will be more effective maintenance, higher reliability, better options for consumers, and ultimately a higher national good.
How a business monitors, interprets and utilizes data is critical to its success. In the energy and utilities industry, data collected from smart meters and intelligent electrical devices (IEDs) give companies extended insight so that they can achieve, on the operational side, far fewer unexpected equipment failures, greater overall reliability, and market flexibility; a variety of new applications will follow from faster fault diagnostics and increased immersion of technicians in new data-enriched processes. As for consumers, more is known in near real time about energy volume and time of use, enabling utilities to monitor applications that track and adjust electricity rates based on availability and demand, as well as notify consumers of usage spikes, or outages.
Yet the volume of new data can also be a curse. Data modeling is essential, or else the flood of new data may be more trouble than benefit. Modeling allows utilities to better define and analyze data requirements and implement new business processes that improve overall management and operations, as well as construct cross-departmental analytics for overall utility optimization.
Thus, recent smart grid implementations have introduced standard models, such as The International Electrotechnical Commission’s Common Information Model (CIM), which defines architecture for information about electrical network configuration and status. The CIM allows for analytics that are more portable, less tied to the underlying operational systems (such as a particular Distribution Management System, Outage Management System, or Geographic Information System), and more open to innovative application of the data by senior power system engineers at the utility.
As smart grid deployments increase, the standards will continue to evolve with new energy devices, requirements, and conditions. Today the focus of standards organizations, such as NIST, IEEE, IEC and ISO, is on interoperability at several levels of the new smart grid devices and systems. Synonymous with this evolution is the need for improved data communications and security, for if the sensor data cannot be reliably and securely transported, it cannot support the overall mission of a smarter grid. Security of grid control functions is critical, and design of smart grid systems incorporate this is the initial design. Typically there are multiple components and devices from various manufactures and no single communications technology or standard that is suitable for all utilities or even for all operational areas across any individual utility. Thus, we see an evolutionary scenario where the interoperability is paramount to the equation.
Another important aspect of effective use of smart grid data is to have an architecture that considers an environment of distributed intelligence; this is essential due to split-second timing and non-ubiquitous high-speed communications. Coordination of solutions must be evaluated. This endeavor falls under the rubric of what's known as system of systems research. A particularly important aspect is emergent behavior–unforeseen correct actions of individual systems that work to counter the overall goal. For example, one can imagine a demand response system taking different action than a transient stability control system in reaction to the same renewable generation problem. To support research in to this area, high fidelity grid simulation is being used as a vehicle to understand these issues before projects progress beyond the pilot phase..
Another important dimension of handling smart grid data involves real time optimization and predictive analytics. Together, they provide utility companies with the tools to improve operational efficiency, system operations, maintenance and customer satisfaction. Recall that in many utilities, two-thirds of the employees are field technicians, so the more empowered they are by better-automated situational awareness, the faster and more safely grid faults are detected, located, restored or even prevented. Much of the value of the analytics and optimization aspect of smart grid is with new integrated applications that leverage this new data and span the gap between existing, specialized systems.
Typical advanced solutions include predictive maintenance via signal analysis, engineering digital workbenches, which allow power engineers to explore and develop their own views of the data and assemble their own analytics from a tool box; and predictive simulation to enable dynamic integration of solar, wind and energy storage in face of uncertainty from Mother Nature. Predictive analytics, besides enabling utilities to better foresee and prepare for system failures that could lead to large outages, will also provide them with a more comprehensive understanding of consumer behavior in an era of green thinking and rising percentage of electric vehicles. They will be able to design and implement new business models that are adaptive and tailored to their customers.
Looking just a little further ahead, the next wave of sensor technology and monitoring solutions is already approaching. New sensors both for operational and non-operational data, some embedded in IEDs will be available for improved real-time operation. The Phasor Measurement Unit is one such advanced IED helping transmission coordination. This will aid with infrastructure operations, a major function of utilities given their huge base of infrastructure assets.
IBM’s Watson, a Deep Question Answering technology is capable of amplifying the expertise of technicians, can yield quicker diagnostics and, on occasion, prompt operator advisories. Developed by IBM researchers, Watson is a computer system designed to rival a human’s ability to understand human language and answer questions with speed, accuracy, and confidence. DeepQA systems of this magnitude could help address challenges and add a layer of intelligence to power grids.
Another IBM project, a nationwide installation of smart electricity and water meters in Malta, is a good example of how comprehensive some smart grid projects have become. The Maltese National Electricity and Water Utilities and IBM are together working to replace 250,000 analog meters with smart meters and equipping the 250,000 water meters with communication modules to monitor usage close to real time, identify water leaks, curb electricity theft and spot other electricity issues, as well as help customers to better conserve energy. IBM alongside a host of strategic partners is working to rapidly deploy smart meters across the nation, on the three inhabited islands of the archipelago: Malta, Gozo and Comino.
Completion of the project will to position Malta to boast of being the world’s first smart grid nation, with an integrated electricity and water smart grid. But it also will help the country address dire energy and resource problems, as an IEEE Spectrum article reported last year. At present, Malta relies heavily on expensive and energy-intensive seawater desalination to provide the greater part of its fresh water, and it obtains almost all its power from oil-fired power plants, which are subject to the vagaries of the world petroleum market. With the vast amounts of data its smart grid will generate, government officials, utilities and citizens will be able to make more informed decisions. "You will have thousands of times more information than you have now," says Paul Micallef, who oversees the development of smart meters for IBM Global Services. "Unless you're in the position of the utilities, you can't even fathom the change that's going to happen."