Climate change and public policies to reduce carbon emissions are rapidly transforming Chile’s electricity system. This challenging shift involves a gradual move away from fossil fuel technologies like coal, diesel, and natural gas power plants, towards renewable sources such as wind and solar power.
Unlike conventional power plants, which operators could dispatch based on demand, these new renewable sources depend directly on climatic factors like solar radiation and wind intensity. This introduces a new level of variability and uncertainty into the electricity grid.
For example, a solar plant might generate at maximum capacity at midday, only to drop to almost zero in minutes if clouds appear. Similarly, wind generation can fluctuate sharply with changes in air currents, posing challenges for system responsiveness.
This vulnerability became dramatically clear on February 25, when a failure on the Nueva Pan de Azúcar-Polpaico line left nearly half the country without power for hours. The event served as a stark reminder that even seemingly minor incidents can escalate into a blackout given the right circumstances.
To navigate this challenging landscape, Usach academic Dr. Héctor Chávez is heading a Fondecyt Regular project designed to predict the system’s future behavior. His team is integrating reduced mathematical models, data analytics, and artificial intelligence to simulate various scenarios that could threaten the electrical system’s security, all in faster-than-real-time. This advanced tool will empower network operators to make swift, preventive decisions when unexpected situations arise.
“Our goal is to create simulations that outpace real-time operations, providing operators with crucial insights into potential system threats. This empowers them to take preventive action, such as averting a blackout,” he explains.
This proposal rests on a straightforward yet potent idea: if we simplify current electrical system models without sacrificing physical fidelity, and integrate rapid tools like AI, we can run numerous simulations concurrently to predict future scenarios. This means operators can anticipate and address problems before they materialize, rather than waiting for them to occur. This methodology fundamentally differs from traditional simulations, which, while offering extensive detail, fall short in enabling real-time decisions.
“While the February 25 blackout was triggered by a fault that, in theory, shouldn’t have caused a widespread outage, under specific conditions of uncertainty, such a fault can have significantly amplified consequences. Our project seeks to correct this by identifying when the system is in a vulnerable state, allowing us to know before something happens,” he says.
For Chile, these challenges are substantial, compounded by a confluence of factors that amplify its vulnerability. The country’s electricity system is notably smaller than those of major players like Brazil, the United States, and Europe, and its longitudinal geography inherently hinders regional coordination. Furthermore, Chile stands among the global leaders in the proportion of solar and wind power generation. As the academic explains, “If we were to rank countries worldwide, we would be in the semifinals of those with the most solar and wind power connected to the grid.”
Enhancing Power System Security
Aware of the challenges, Chávez’s team developed a four-year project structured into three key stages. The first involves strengthening previous doctoral research by thoroughly exploring the global state of the art. Next, they will define the most appropriate mathematical tools for constructing reduced models that accurately reflect the system’s physical behavior. Finally, these models will be implemented in real-time simulators, utilizing acquired technology to validate the project’s core hypothesis.
“We’re not aiming for an automaton to make decisions independently. The electrical system is delicate, and for administrative reasons, a person must always be legally accountable. Our goal is to reduce calculation times so we can simulate more scenarios in parallel and cover more conditions, ultimately leading to better human decisions,” the researcher explains.
A successful outcome for this project would not only enhance the security of Chile’s electricity system but could also establish it as a leading example for other countries confronting similar hurdles. With decarbonization an undeniable future, having tools that allow for increased renewable generation without compromising system stability will be vital for reaching climate targets while maintaining a secure electricity supply.
“If we successfully implement this tool and help prevent blackouts or major failures, we’ll be contributing not only to the security of Chile’s electricity system but also to progress toward a more sustainable future,” concludes Dr. Héctor Chávez.