Virginia Tech professor Y.A. Liu introduces new technology

FNET offers a bird's-eye view of the power grid

Virginia Tech Power Information Technology Lab

Dr. Yilu Liu's homepage

The Bradley Department of Electrical and Computer Engineering (ECE) at Virginia Tech

Frequency Monitoring Network

Fixing America's power grid

Tidal energy

Virginia Tech solar house

Briefs: Energy-related research

Editorial: Energy and the Whole Community




When electrical power service was restored on August 15, 2003, to most of New York and other parts of the United States and Canada after some 30 hours of the largest blackout in U.S. history, the exact cause remained a mystery. In fact, the official report was not released until some three months later, and the astounding explanation for the massive blackout was the failure by a small Ohio company to trim trees along a power line.

Amazingly, a generating plant in the small town of Parma, Ohio, hundreds of miles away from New York City, went off-line amid high electrical demand and strained high-voltage power lines that went out of service when they came in contact with the overgrown trees. The U.S.-Canada Power System Outage Task Force, charged with the investigation of the incident, also found that the company, FirstEnergy, was slow in its remedial action because, in part, of a software glitch that prevented alarms from showing on their control system. The result was a colossal cascading effect that forced the shutdown of more than 100 power plants, affecting an estimated 50 million people.

In the aftermath, President George Bush’s administration emphasized the need for changes to the U.S. national energy policy, critical infrastructure protection, and to Homeland Security efforts. The blackout caused the failure of most systems that would ordinarily detect unauthorized border crossings, port landings, or attempts to gain access to secure locations.

Virginia Tech professor introduces FNET

A relatively new technology, called Frequency Monitoring Network (FNET), developed at Virginia Tech by Yilu Liu, a professor of electrical and computer engineering (ECE), and several ECE students, addresses the problem of immediately pinpointing the location of a power grid problem before a cascading effect can again cripple large parts of the nation. Liu’s work is based on some earlier work of her ECE colleagues, Arun Phadke, University Distinguished Professor, and James Thorp, department head and the Hugh P. and Ethel C. Kelly Professor.

Liu first proposed FNET in 2000 as an Internet-based, real-time, global positioning system (GPS) synchronized monitoring network. Today, FNET consists of more than 30 Frequency Disturbance Recorders (FDRs) positioned around the United States and managed centrally at Virginia Tech.

FNET allows “for the first time, high-precision dynamic frequency measurements of the entire U.S. power grid. Before, only spotted regional measurements existed due to the high cost of installing expensive devices in the high-voltage substations,” says Liu, a Fellow of the Institute of Electrical and Electronic Engineers. “Just like the human pulse, frequency is one of the most critical pieces of information in an electric power grid.”

A power grid consists of a set of large power plants, including hydropower, nuclear, and other types. Wires connect the grid, and the linked maze allows sharing between companies. If a power plant is removed from the system for repair, the other parts of the grid can pick up the slack.

But if something, such as lightning, causes a power plant to unexpectedly trip off line, the other plants on the grid must meet the increased demand in a timely manner. A problem occurs if they don’t realize what has happened and don’t react correctly. In that case, some plants might disconnect from the grid for self-protection. As each additional power plant disconnects, the problem balloons and the cascading effect results in a situation similar to the August 2003 blackout.

Liu explains that deregulation of the electric utility system is also responsible for problems with power transmissions over a grid. “Because of deregulation, there is reluctance to share information among utility companies. So in order to see a complete picture, we have to be able to pull in the measurements of the electromechanical waves traveling in all of the systems.”

A “bird's-eye view” of the power grid

FNET “gives us a bird’s-eye view if we can monitor from enough locations and use information from the entire power grid to make intelligent decisions,” Liu says. If the Parma, Ohio, incident were to occur with the FNET monitoring network in place, FNET would be able to triangulate the location of the early event and the amount of lost power generation, and then suggest actions to immediately ramp up power in other units to prevent the cascading effect. Triangulation is a navigation technique that uses the trigonometric properties of triangles to determine a location by means of compass bearing from two points a known distance apart. In FNET, one can picture the power grid disturbance as waves on a pond struck by a stone.

GPS provides precise timing to the microsecond and each FDR records the time when it sees a “wave front” as it passes. Multiple FDRs at different locations allow the system to piece together the origin of the wave. Virginia Tech has received a U.S. patent on using wide area frequency measurements to locate a power system event source.

Liu’s triangulation research has received support from the Electric Power Research Institute, whose member companies represent more than 90 percent of the electricity generated in the United States. Major events on a power grid, such as generator loss, a line trip, or a load drop, can be triangulated using automated online mathematical algorithms. In a three-month period in the summer of 2005, FNET captured more than 70 events in U.S. power grids. “This proves the system will be of great value in critical infrastructure monitoring of the power grid. And FNET can also be used in post-disturbance scenario reconstruction and tracking the sequence of events leading to a blackout,” Liu says.

She is looking to develop automated online algorithms for triangulating event location and predicting severity. Current triangulation algorithms are done off-line. “It would be of tremendous value to have event information displayed on a U.S. map with the location and severity level indicated in real-time,” she says.

The simplicity of the technology from a user’s point of view is rather astounding. FDRs have no installation costs; the user just plugs a unit into a standard electrical outlet. All of its components are inside a rectangular box that’s a little bigger than the standard laptop personal computer. “It operates by taking voltage signals from the electric outlet, computing the frequency from the voltage signal, time stamping it with the GPS clock, and sending the synchronized measurement data to the server at Virginia Tech,” Liu explains. “The FDRs do not need to be connected to high-voltage lines in the substations.”

The current generation technology costs only $2,500 per box, with forecasts of it dropping to $1,000 or $500 as the second- and third-generation technologies are developed.

By comparison, the technology developed in the 1980s by Phadke was a synchronized phasor measurement unit (PMU) that allowed additional current measurements. Since PMUs are placed in the high-voltage substations, with installation costs plus the need for dedicated communication channels, the total figure could reach $80,000 per unit, according to Michael Ingram, senior manager of research and technology applications at the Tennessee Valley Authority (TVA), a long-time supporter of FNET work. “Conceptually, you could say FDR is simply a stripped-down PMU, but suddenly it becomes very powerful because it allows monitoring to take place at low voltage levels,” Liu says.

Both Phadke and Virgilio Centeno, who built the world’s first PMUs, are advising the FNET team, while ECE Professor Richard Conners, who Liu calls the “computer guru,” provided the technical backbone for the development of the monitor device. At the time he became interested in the project, “we had no funding, but we had honor students willing to work on it. Rich simply said, ‘Let’s do it.’ And we did,” Liu says.

Later, the National Science Foundation provided an equipment grant to help with the development of the instrumentation, and the State Council of Higher Education for Virginia provided an equipment grant. The TVA and ABB, a global power and automation technology company, supported Liu’s work and assisted with concepts.

Today, TVA, the nation’s largest public power company with 158 locally owned distributors, has 19 of the FDRs. Midwest ISO, an independent transmission system operator that serves the electrical transmission needs for much of the U.S. Midwest, also supports FNET technology. Another believer who has deployed Virginia Tech FDRs is Genscape, a provider of critical, real-time energy generation and transmission data, with operations in the United States and in Europe.

A total of 200 FDRs located around the United States “will be sufficient” to monitor the entire country, Liu says.

Liu joined the ECE department in 1990 upon receiving her doctorate from Ohio State. While a Ph.D. student at Ohio State, she developed a technique in transformer modeling that covered a broader frequency range than previously available. Her work in this area was implemented by the electric power industry. In 1993, she was awarded an NSF Young Investigator Award. In 1994, she received the prestigious NSF Presidential Faculty Fellow Award.

— Lynn Nystrom, College of Engineering