The same physical process that produces a beautiful aurora display has a considerably less photogenic side effect: it can induce real electrical currents in long-distance power transmission infrastructure, buried pipelines, and other long conductors, and in at least one well-documented case, it has caused a major blackout.
The physics: how a geomagnetic storm becomes an electrical current
During a geomagnetic storm, Earth’s magnetic field changes rapidly, driven by the same disturbed magnetosphere responsible for auroral activity (including, at the sharpest moments, the substorm current wedge described in substorms explained). By a basic principle of physics (electromagnetic induction — the same phenomenon behind how a generator or transformer works), a rapidly changing magnetic field induces an electric current in any long conductor exposed to it. Long-distance, high-voltage power transmission lines, especially those running roughly north-south across high-latitude regions, act as exactly this kind of conductor, picking up geomagnetically induced currents (GICs) directly from the storm.
Crucially, it’s the rate of change of the magnetic field — not its absolute strength — that determines how large an induced current results. This rate is often written dB/dt and measured in nanotesla per minute. A geomagnetic disturbance that builds up strength slowly, over hours, induces comparatively modest currents even if the total field change is large; a sharp, rapid swing — the kind associated with substorm onset or a sudden shock arriving from a CME — induces much larger currents even over a shorter interval. Rates exceeding roughly 300-500 nT per minute during the most intense moments of a strong storm are generally treated as a significant threshold by grid operators monitoring for GIC risk, though the precise impact also depends heavily on local ground conductivity (rock and soil type affect how currents actually distribute) and the specific geometry of the grid infrastructure in that region.
Unlike the alternating current the grid is designed to carry, GICs are quasi-direct current — slowly varying rather than oscillating at the grid’s normal frequency. This mismatch is the actual problem: transformers, a critical piece of grid infrastructure, are not designed to handle this kind of current, and it can push them into a saturated state that generates excess heat, distorts the normal current waveform, and in severe cases can damage the transformer permanently.
The Quebec 1989 blackout: the textbook case
The best-documented real-world example is the geomagnetic storm of March 1989, driven by an unusually strong CME. Geomagnetically induced currents overwhelmed protective systems on the Hydro-Québec grid, triggering a cascading failure that left roughly six million people without power for about nine hours. The same storm caused voltage irregularities and equipment issues on grids elsewhere in North America and Europe, but Quebec’s grid — situated at a high geomagnetic latitude with long transmission lines well suited to picking up induced currents, and built on the highly resistive granite bedrock of the Canadian Shield, which tends to worsen GIC effects compared with more conductive ground — was hit hardest and became the standard reference case for this entire field of study.
Other documented events
The Quebec blackout isn’t the only well-studied case. The October-November 2003 geomagnetic storms — sometimes called the “Halloween storms,” among the most intense of that solar cycle — caused a transformer failure and an approximately hour-long blackout affecting the Swedish city of Malmö, along with a range of other documented grid disturbances, equipment alarms, and satellite anomalies across Europe and North America. Going back further, the 1859 Carrington Event — discussed for its aurora effects in what is a CME — provides a striking historical preview using the infrastructure of its own era: telegraph systems, the long-distance electrical networks of the 19th century, reportedly sparked, shocked operators, and in some well-documented cases continued transmitting messages using only the geomagnetically induced current itself, after operators had deliberately disconnected the telegraph’s own batteries to test what was happening.
Beyond power grids: pipelines and other long conductors
Power transmission lines aren’t the only infrastructure affected. Long buried metal pipelines, which typically rely on a cathodic protection system (a deliberately applied small electrical current to prevent corrosion) to stay in good condition over decades of service, can have that protection system overwhelmed by geomagnetically induced currents during a strong storm, potentially accelerating corrosion at specific points along the pipeline. Long-haul telecommunications cables and railway signalling systems have historically shown similar susceptibility, essentially for the same reason telegraph lines did in 1859: any sufficiently long conductor is a potential path for these induced currents.
What’s changed since 1989
The 1989 event was a genuine wake-up call for the power industry, and grid operators — particularly those at high geomagnetic latitude, such as in Canada, Scandinavia, and the northern United States — have since implemented real-time space weather monitoring, updated operating procedures for storm conditions (such as adjusting reactive power reserves or temporarily taking sensitive equipment offline), and in some cases hardware mitigations like GIC-blocking devices on transformer neutral grounding. NOAA’s space weather alerts (see the NOAA G-scale explained) are used directly by grid operators for exactly this purpose, not just by aurora chasers.
How seriously should this be taken today?
For the storms most aurora watchers care about — G1 through G3 events, which cover the overwhelming majority of storms and most of what makes flagship and even storm-only destination sightings possible — grid risk to ordinary consumers is minimal, and modern monitoring handles these routinely. The genuine open question in the field concerns a truly extreme, G5-class storm significantly stronger than typical modern events (sometimes discussed alongside the historical 1859 Carrington Event, believed to have been considerably stronger than both 1989’s storm and the 2003 Halloween storms) — researchers and grid operators continue to study and prepare for this scenario, but it remains a low-probability, high-impact risk rather than a near-term expectation.
How grid operators actually model and prepare for GIC risk
Beyond real-time monitoring, grid operators and researchers build detailed models of how a given geomagnetic disturbance would translate into induced currents across a specific regional grid, combining the local geomagnetic field data with a detailed map of ground conductivity (since current flows more readily through some rock and soil types than others) and the grid’s own physical layout — the length, orientation and grounding configuration of individual transmission lines. These models let operators run “what if” scenarios ahead of a forecast storm, identifying which specific transformers or substations face the greatest theoretical risk, and pre-emptively adjusting operations at those specific points rather than treating the whole grid uniformly. This kind of scenario planning has become considerably more sophisticated since 1989, incorporating decades of additional storm data and improved regional ground-conductivity mapping.
The connection to your aurora forecast
None of this changes how to actually watch for the aurora — the same Kp index, G-scale level, and solar wind data described in what is the Kp index and reading a solar wind plot that tell you whether a display is likely also happen to be exactly what grid operators are watching for their own, considerably higher-stakes reasons. It’s a genuinely interesting footnote to keep in mind next time a strong G3+ storm watch is issued: the same event bringing the aurora to destinations near Winnipeg, Edmonton, Oslo or Helsinki is also, quietly, being watched closely by power grid control rooms across those same high-latitude regions.
Frequently asked questions
Has a geomagnetic storm ever actually caused a blackout?
Yes — the best-documented case is the March 1989 storm that caused a roughly nine-hour blackout across the Hydro-Québec grid in Canada, affecting millions of people, triggered by geomagnetically induced currents from an unusually strong geomagnetic storm.
Are power grids today better protected than in 1989?
Significantly, yes. Grid operators, particularly at high-latitude utilities most exposed to this risk, now monitor real-time space weather alerts and have operating procedures, and in some cases hardware protections, specifically designed to reduce the risk of a repeat event, though the risk from a truly extreme storm is not considered fully eliminated.
Should I personally worry about power outages during a big aurora storm?
For an ordinary strong storm (G1-G3), no meaningful risk to typical consumers exists. For a genuinely extreme storm (G4-G5), some regions — particularly at high geomagnetic latitude with long north-south power transmission lines — could experience real grid stress, though modern monitoring and mitigation substantially reduce this risk compared with 1989.
Why do high-latitude power grids face more risk than low-latitude ones?
Geomagnetically induced currents are driven by rapid changes in the local geomagnetic field, which are strongest closer to the auroral oval — high-latitude regions like Quebec, Scandinavia, and parts of the northern US and UK experience larger geomagnetic swings during a storm than equatorial regions do.
Is it the strength of the magnetic field or how fast it changes that actually matters for GICs?
How fast it changes -- specifically the rate of change over time, often written dB/dt and measured in nanotesla per minute. A slowly changing, even quite strong, magnetic disturbance induces comparatively little current; a rapid swing, even a briefer one, induces much more. Rates exceeding roughly 300-500 nT per minute during the sharpest moments of a strong storm are generally considered a significant threshold for grid concern.
Do geomagnetic storms affect anything besides power grids?
Yes -- geomagnetically induced currents also flow through long buried metal pipelines, where they can accelerate corrosion at points where the pipeline's protective cathodic protection system is overwhelmed, and through long-haul telecommunications and rail signalling infrastructure historically, in a similar way to how they affected 19th-century telegraph lines.
Have other storms besides 1989 caused documented grid problems?
Yes -- the October-November 2003 "Halloween storms" caused a transformer failure and roughly hour-long blackout affecting the Swedish city of Malmö, among other documented grid effects across Europe and North America during that period, providing a second well-studied case study alongside the 1989 Quebec event.
How did the 1859 Carrington Event affect the infrastructure of its time?
Telegraph systems were the era's equivalent of a long-distance electrical grid, and operators reported sparking equipment, telegraph paper catching fire in some accounts, and in several documented cases the ability to send messages using only the induced current itself, with the telegraph's own batteries disconnected -- a striking, if narrower, historical preview of the same geomagnetically induced current phenomenon affecting modern infrastructure today.