Finding out that the sun, the same one that warms your coffee through the kitchen window, can covertly start destroying the power grid hours before anyone on Earth realizes it’s coming, is a bit disorienting. It turns out that this confusion is precisely the issue.
An incoming geomagnetic disturbance was classified as a possible G4 or greater event in May by NOAA’s Space Weather Prediction Center. That’s serious by most standards. Grid engineers are extremely uncomfortable with the voltage instabilities and protective system misfires that a G4 brings. The G-scale runs from G1 to G5. However, the storm’s final intensity wasn’t the only thing that alarmed some observers. The speed was the problem. In contrast to what NOAA‘s forecast models had predicted, the event intensified toward G2 storm levels more quickly. This discrepancy could have reduced the response window from hours to nearly nothing in a different situation.
Coronal mass ejections, or CMEs, are massive clouds of solar plasma and magnetic field material that are violently and indifferently expelled from the sun, causing geomagnetic storms. It can take them anywhere from 18 hours to several days to travel the approximately 93 million miles between the sun and Earth because they move more slowly than light. In theory, that travel time is a grace period. The real world is more chaotic. Scientists can use solar-facing instruments such as the Parker Solar Probe to estimate the speed and approximate magnitude of a CME once it leaves the sun. However, there is a gap—a lengthy, data-free area where the cloud moves covertly. Until the CME reaches the L1 Lagrange point, a monitoring station about a million miles from Earth, it is frequently not possible to confirm the polarity of its magnetic field, which determines much of the storm’s actual impact on Earth. A fast-moving CME can cross that final million miles in 20 to 30 minutes. That doesn’t really serve as a warning.
This is not abstract, which is what makes it relevant to the grid. A southward-directed CME couples to Earth’s magnetosphere rather than deflecting around it, pumping solar energy and particles into the planet’s magnetic system. The practical consequence is the generation of geomagnetically induced currents — GICs — long, slow DC pulses that travel through anything conductive and extended. Like, say, a high-voltage transmission line running several hundred kilometers in a rough north-south direction. In essence, the electrical grid was built to be an ideal conduit for the precise type of current generated by a strong solar storm.
Transformers are particularly at risk. Alternating current, which changes direction numerous times per second, powers grid transformers. GICs exhibit steady, persistent, and pushy behavior more akin to direct current. An AC transformer’s insulation is attacked by localized heating produced when DC saturates the magnetic core. The transformer is not always immediately destroyed. It is more subtle than that. The damage accumulates, shortening the lifespan of equipment that was already expected to run for decades, and in severe events pushing transformers toward what one grid operator bluntly described as becoming “kaput.” It is not the same as changing a fuse when replacing a high-voltage transformer. These units are custom-made, weigh tens of tonnes, are produced by a small number of suppliers worldwide, and even under normal market conditions, lead times can reach years. It is not typical for a severe worldwide solar event to generate simultaneous demand across dozens of nations.
The relatively narrow miss during this recent escalation may lead to a reevaluation of model validation, especially with regard to the rate of intensification rather than just peak magnitude. Operationally, the G-scale classification is important. Grid operators respond to a G2 event in a particular way, whereas a G1 does not. If the storm reaches G2 levels ninety minutes ahead of schedule, those ninety minutes represent actual choices and preparations that were not made. Because their location places them directly in the path of the strongest GIC activity, grid operators in high-latitude regions—such as Scandinavia, Canada, New Zealand, and the northern United States—tend to operate with a more acute awareness of this type of timing risk.
The industry is beginning to acknowledge that mitigation is a practical solution rather than merely a theoretical one. Critical substations are installing GIC blockers, which are essentially capacitors placed on important transformers to stop DC infiltration. Assets can be safeguarded at the expense of brief, controlled outages by using controlled disconnection protocols, which temporarily remove vulnerable transformers from the network prior to a storm. It appears that preparation is becoming more and more important in the math. A severe but realistic storm could cost billions in lost economic output without mitigation, according to an unpublished economic analysis commissioned by a major grid operator. This amount makes the investment in blockers and response planning seem almost insignificant in comparison.
The forecast window is the more difficult issue. Improved tools are helpful. Better models are more beneficial. However, the accuracy with which a cloud of magnetized plasma can be described prior to its actual arrival may be limited. In the end, the difference between the storm’s actions and what the models predicted was manageable. It’s important to consider the consequences of not doing so.
