A diagnostic of voltage infeasibility at low net load and the role of inverter-based grid support

The electric grid is a vast, delicate machine that must balance supply and demand the instant they occur. When the sun shines brightly and the wind blows hard, renewable sources like solar and wind farms can produce so much electricity that they cover most of the day's needs. In these moments of low demand, the traditional power plants that usually keep the system stable are often turned off to save money. This leaves the grid with fewer large machines capable of adjusting the voltage, the electrical pressure that pushes current through the wires. If that pressure gets too high, it can damage equipment and trigger a blackout; if it gets too low, lights flicker and appliances fail. For years, engineers have worried that when renewable energy dominates, the grid would suffer from too much voltage, a condition that seemed to require absorbing excess energy to bring the pressure down.

A team of researchers at the University of Seville in Spain has challenged this long-held assumption. They investigated what happens when the grid is flooded with renewable energy and traditional power plants are switched off, a scenario that recently led to a major blackout in the Iberian Peninsula. Using advanced computer simulations of three different electrical networks, they developed a new way to diagnose why a grid fails. Their work reveals that when the voltage gets too high in these low-demand hours, it is often a sign of the opposite problem: a desperate lack of support, not an excess of power. The researchers found that simply trying to absorb energy, as standard safety checks often suggest, would make the situation worse. Instead, the solution lies in keeping specific power plants online to provide active support and ensuring that battery storage systems do not continue to draw power when their local generators are lost.

The study began by looking at the hours when renewable energy covers nearly all the demand. In these times, a cost-driven system turns off most of the large, synchronous generators that traditionally hold the voltage steady. The researchers asked a simple but critical question: can the remaining equipment keep the voltage within safe limits before and after any single power plant suddenly trips? To answer this, they built a diagnostic tool that acts like a medical scan for the grid. Instead of just looking at where the voltage is highest, the tool simulates thousands of scenarios to determine if the system is truly failing because it has too much energy or because it is starving for it. They tested this on a small nine-bus system, a larger thirty-nine-bus network, and a complex seventy-three-bus model of a real-world grid, running simulations for hours when demand was lowest and renewable output was highest.

The results were surprising and counterintuitive. In nearly every case where the grid failed, the system was actually lacking reactive power, a form of energy needed to maintain voltage, rather than having too much. The researchers discovered a specific mechanism that causes the voltage to spike dangerously high. Imagine a long, lightly loaded cable stretching from a power source to a remote town. Because the cable is not carrying much current, it naturally generates its own voltage, pushing the pressure up at the far end. To keep the voltage at the remote town from dropping too low, the system is forced to raise the voltage at the source even higher. This action pushes the voltage at the far end of the cable over the safety limit. A standard check sees the high voltage at the cable end and assumes there is too much power, suggesting the installation of devices to absorb energy. However, the researchers showed that this would be the wrong move. The real problem was that the remote town needed more support, not less. Adding devices to absorb energy at the cable end actually made the voltage rise even higher, worsening the failure.

The study also examined how modern power plants, which use inverters to connect to the grid, can help. These plants can be programmed to control voltage directly, acting like the traditional generators they replace. The researchers found that when these plants are set to control voltage, the number of safe, operable hours jumps dramatically. In one large system, the share of hours that could be operated safely rose from just nine percent to seventy-three percent when plants used voltage control instead of a fixed setting. This improvement came mostly from the plants' ability to generate or absorb reactive power on demand. Even if the plants only had half the usual capacity to do this, they performed just as well as plants with full capacity that were set to a fixed output. The key was the ability to adjust dynamically, closing the loop to keep the voltage stable as conditions changed.

Another critical finding involved the batteries often co-located with these renewable plants. In many current designs, if a solar or wind plant trips offline, the battery at the same location might keep charging, drawing power from the grid. The researchers found that this behavior can be dangerous. When the plant trips, the area loses a source of power, but if the battery keeps charging, it acts as an additional drain, deepening the local shortage of energy. In their simulations, a battery that continued to charge after its plant tripped removed the ability to operate safely in about a quarter of the cases where the system was otherwise stable. Crucially, the battery causing the problem was not always the one at the site of the biggest failure. It was often a battery at a different location that was feeding a remote, vulnerable part of the grid. This means that safety rules cannot just focus on the site of the largest outage; they must require every battery connected to a plant to stop charging immediately if that plant trips.

The researchers concluded that the standard way of reading grid failures is often misleading. When a safety check flags a high voltage limit, it does not automatically mean the system has too much power. In the low-demand hours of a renewable-heavy future, a high voltage reading is frequently a symptom of a local shortage, where the grid is being stretched too thin to support a remote load. The remedy is not to absorb energy but to supply it where it is missing, either by keeping nearby traditional units online or by using the voltage control capabilities of modern inverters. The diagnostic tool they created can tell the difference between these two situations without needing to know the detailed layout of the network, offering a clear path for operators to keep the lights on. By understanding that a high voltage can signal a lack of support, grid managers can avoid applying the wrong fix and instead strengthen the grid exactly where it is weakest.