Major power outages are fairly infrequent, but when they happen they can
result in billions of dollars in costs. Now research from North Carolina State
Univ. has led to the development of an approach by which high-resolution
power-system measurements, also referred to as Synchrophasors, can be
efficiently used to develop reliable models of large power systems, which would
help us keep an eye on their health.
Synchrophasors are real-time measurements of voltages and currents that
provide a very high-resolution view of various complex events occurring in a
power system. They are measured by sophisticated digital recording devices
called phasor measurement units (PMUs). “PMUs are comparable to surveillance
cameras that continuously monitor the complex dynamics of groups of people in
busy places, and indicate how different people respond and interact with each
other,” says Dr. Aranya Chakrabortty, lead author of a paper describing the
research and an assistant professor of electrical and computer engineering at
NC State.
“This research is a major step toward helping us understand how
Synchrophasor technology can be used to model the complex behavior of any
large, geographically distributed power system, especially taking into account
the system’s interconnected nature,” Chakrabortty says.
“We need to have a better understanding of how a disturbance entering one generation
cluster—or localized group of nodes—may spread across the entire system,
creating havoc in its neighboring clusters as well,” Chakrabortty says. “More
importantly, we need to investigate if the speed of this spread is dictated by
the way these clusters are connected to each other.” The North American power
grid is divided into four distinct operating zones, each of which has several
such generation pockets, Chakrabortty explains, across which a disturbance can
disseminate very easily.
For example, during the 2003 Northeast Blackout, generating units in Ohio and New England
appeared to be functioning smoothly. However, there was significant disparity
between the two regions when it came to reactive power. That disparity created
a cascading series of “voltage collapse” events—which cut off power to
approximately 50 million people, was linked to multiple fatalities and cost an
estimated $4 to 10 billion. The event highlighted the need for monitoring the
system globally, rather than focusing on individual nodes in isolation.
“In order to understand how the effects of major disturbances can propagate
across the North American power system, we need highly reliable and rigorous
mathematical models that capture the dynamics of its various clusters, as well
as the way those dynamics will evolve when the clusters are connected to each
other in the overall system,” Chakrabortty says. “Traditional measurement
methods in power systems are too slow and, therefore, incapable of capturing
these dynamics, which can change dramatically in fractions of a second. With
the Synchrophasor technology today such models are possible.”
Chakrabortty and his co-authors from Rensselaer Polytechnic Institute (RPI)
and Southern California Edison have developed an approach for creating cluster
models, which uses Synchrophasors from PMUs located at specific points within a
cluster of nodes. The approach also allows one to identify how the clusters are
connected to each other by comparing PMU measurements at different points in
the system. “Once you have modeled the clusters and determined their
connections,” Chakrabortty says, “our algorithm enables you to model the
interactive behavior of the clusters within the larger system in the face of
large disturbances. We also show how to place PMUs optimally at the nodes so as
to extract maximum amount of useful information for better modeling.
“Our models are informative, yet easy to compute,” Chakrabortty adds. “They
will help power-system operators track and predict the global health of any
distributed power system in real time so that catastrophes such as the 2003
blackout can be prevented in the future. The study will lead to an entirely new
vision of monitoring and controlling the North American grid, which is becoming
more expansive, and, more chaotic day by day.”
The paper, “A Measurement-based Framework for Dynamic Equivalencing of Large
Power Systems using Wide-Area Phasor Measurements,” was co-authored by Dr. Joe
Chow of RPI and Armando Salazar of Southern California Edison. The paper is
published online in IEEE Transactions on Smart Grid. The
research was partly funded by the Power System Research Consortium, and is
currently being extended at NC State under the support of the National Science
Foundation.
NC State’s Department of Electrical and Computer Engineering is part of the
university’s College
of Engineering.