ISO-NE has developed a groundbreaking software solution to locate the source of dangerous power system oscillations

Detection is key to mitigating forced oscillations, which helps advance the reliable operation and stability of the electric grid

ISO New England’s Business Architecture & Technology (BAT) department is responsible for advanced research and development, working closely with many business units within the ISO and collaborating with major universities, labs, and vendors to design and deploy tools for power system analysis and control. BAT has been internationally recognized for the development of important new technologies and innovative solutions.

For years, BAT Technical Manager Slava Maslennikov has studied power system oscillations, which can cause dangerous fluctuations on the electrical grid. Oscillations, caused by equipment failure, human error, or control systems malfunctioning, among other things, inject a forced signal into the high-voltage transmission system, causing disturbances and mechanical vibrations that may lead to equipment fatigue and damage, rare occurrences of cascading outages and, in the worst cases, risk to workers.

How often does forced oscillation occur? “At small magnitude, almost every day,” says Maslennikov. After years of study and work, BAT has designed a groundbreaking solution to locate the source of forced oscillations.

Progress, not perfection

In 2013, the ISO installed new data collection devices on the region’s transmission system called phasor measurement units (referred to as PMUs or synchrophasors). The units measure power grid voltage, current, frequency, and phase angle 30 times per second (compared to the previous one sample every two seconds) and use global positioning system technology to timestamp the measurements. The high-speed synchrophasor data and advanced data analytics provide valuable information on the regional power system by enabling the monitoring of system dynamics that was previously not possible; providing fast and accurate post-event analysis; and validating and improving power system models.

Unexpectedly, since deployment of the PMUs, hundreds of small-scale and a few large-scale instances of “forced oscillations” with a magnitude of more than 100 megawatts (MW) have been detected.

“These results were surprising to me,” says Maslennikov. “We are seeing forced oscillations almost every day—which can potentially represent a serious threat to the system—but we don’t know much about them from textbooks. Textbooks typically describe so-called ‘natural oscillations.’ Power grid control systems are designed to mitigate natural oscillations, and that is why they are typically well-damped and do not impose a threat.”  

Experts agree that the only efficient way to actually mitigate a forced oscillation is to locate its source— the generator or cycling load—and isolate it from the network until the source of the forced signal can be disabled. Because of the complexity of power systems, this can be a challenging task. If the frequency of a forced signal is close to the frequency of the natural system mode, together they resonate, resulting in power swings with magnitudes several times larger than at the source—so powerful, in fact, that they can be observed on the power grid a thousand miles away. It is particularly important for grid operators to know when the source is within their control area so they can take action to mitigate the forced oscillations.

Over the years, numerous vendors and researchers attempted to develop methods for detecting forced oscillation sources that, despite effective implementation in simulations, did not work consistently on actual systems.

Success after determined work

BAT studied all existing published literature on the issue, analyzed pitfalls, and eventually landed on a solution that reliably worked with actual PMU measurements. The team spent nearly two years testing and iteratively modifying algorithms in a real environment before the process was considered reliable enough for implementation.

The result was the Oscillation Source Locating (OSL) application, which monitors PMU oscillatory alarms, estimates the source of oscillations, visualizes the results on a one-line diagram, and names the exact source based on energy flow pattern recognition. The main findings were described in a paper, “Dissipating Energy Flow Method for Locating the Source of Sustained Oscillations,” published in June 2017 in the International Journal of Electrical Power and Energy Systems.

In September 2017, the BAT team launched an online, 24/7 Oscillation Alarm Notification service, which takes the results of the OSL application, collates all relevant details in an email, and sends it to a mailing list of all contacts in the New England control area. The process is fully automated and provides just-in-time actionable information that control center operators can use to mitigate forced oscillations.

Having a real-world impact

Since September 2017, the OSL application and its integrated alarm service have automatically processed and correctly identified the source of forced oscillations for several hundred events. The efficiency and usefulness of the OSL was particularly recognized after January 11, 2019. That day, forced oscillations from an unknown source caused power swings up to 200 MW in all utilities across the entire Eastern Interconnection for about 18 minutes. These oscillations, with magnitudes up to 50 MW, were observed at ISO New England. The OSL application automatically estimated that the source of oscillations was located beyond the ISO New England footprint.