Two major grid disturbances in 2025 and 2026 exposed a stark reality: systems rich in synchronous inertia ride through faults that collapse grids dominated by inverter-based resources. In April 2025, the Iberian Peninsula suffered a total blackout after frequency oscillations triggered cascading inverter disconnections on a system where 71 percent of generation came from inverter-based resources; the collapse unfolded in under a minute and took roughly ten hours to restore. Fifteen months later, a transmission fault in Virginia caused data centers to shed three gigawatts of load across PJM, yet no generation tripped and operators stabilized the system within minutes, aided by a conventional generation fleet that supplied ample rotational inertia.
The Iberian event began with an unexplained frequency oscillation, followed by a second roughly twenty minutes later. Operators responded by raising voltage, but subsequent generation trips in southern Spain removed approximately 2.2 gigawatts — about seven percent of active output — and the combination of overvoltage and persistent oscillations drove a wave of inverter protective trips that collapsed frequency across the peninsula. With nuclear providing only 11 percent and natural gas just five percent of the mix at the time, the grid lacked the kinetic energy buffer that traditionally arrests frequency excursions and buys operators time to act.
By contrast, the PJM disturbance originated from a transmission line fault serving several large data centers. The resulting voltage and frequency deviation triggered the facilities’ own protection schemes, which islanded onto backup power and instantly removed roughly three percent of system load. The disturbance propagated electrically across the PJM footprint, but the predominantly synchronous generation fleet — coal, gas, and nuclear units still running at high output — injected inertial response automatically, damping the frequency swing before any generator protection operated. Very little inverter-based capacity was online at the moment, a condition that likely prevented the positive-feedback loop seen in Spain.
These cases present a profound challenge for failure-mode analysis and grid planning. Both events began with voltage and frequency instability, yet the failure modes diverged completely: one became a cascading blackout, the other a transient blip. The decisive variable was not the trigger but the system’s physical capacity to absorb the shock. As grids worldwide accelerate the retirement of synchronous plant in favor of wind, solar, and batteries, the inertia margin that turned the PJM event into a non-event is eroding. Engineers and regulators must now treat inertia not as a byproduct of thermal generation but as a service that must be explicitly procured, whether through synchronous condensers, grid-forming inverters with synthetic inertia, or market mechanisms that value fast frequency response.
The Iberian blackout caused fatalities and widespread societal disruption; the PJM event barely registered outside control rooms. That difference is the clearest evidence yet that the physics of rotating mass cannot be wished away by policy targets. Grid reliability in the inverter era will depend on deliberate engineering of stability services, not on the hope that low-inertia systems will somehow behave like their predecessors.
Read the full report at Energy Central.