This week a power line went down outside Washington, DC. Normally, the grid would only need a few seconds to recover from such an event. But this took more than 10 minutes because more than 3 gigawatts of data centers stopped consuming power almost simultaneously.

The event caused voltage on PJM’s grid to spike from northern Virginia to Chicago, according to data collected by Ting Labs, a startup that runs a network of IoT sensors in people’s electrical outlets.

The event did not cause a blackout, but it did cause lights across the region to flicker. The incident demonstrated the effect that data centers can have on the network, an outcome that experts believe will become more common.

Northern Virginia, which is located in PJM territory, is home to the largest concentration of data centers in the world.

“It’s the canary in the coal mine,” Ricardo de Azevedo, chief technology officer at ON.Energy, told TechCrunch. These types of events involving large loads, such as data centers, “are happening more and more,” he added.

The event echoes one that occurred two years ago, also on PJM’s network, and could foreshadow larger events if data centers are not built to more gracefully handle power outages. PJM Interconnection operates networks from New Jersey to Illinois and serves 67 million customers, making it the largest network operator in the United States.

When the power line went out this week, data centers switched to backup power and about 3.1 gigawatts of load disappeared in about 30 seconds, according to PJM data. The network seemed to recover somewhat, but the additional loads subsided shortly thereafter. At its peak, PJM’s grid had an additional 3.49 gigawatts of electricity. It was another 11 minutes before it stabilized. Disconnected data centers accounted for about 3% of PJM’s total demand at the time, according to Reuters.

A small percentage may not seem like much, but the electrical grid must operate in a state of almost perfect balance, with supply and demand closely related. If they don’t, voltages can drop or rise. The network and the devices connected to it can tolerate small fluctuations, but if those fluctuations grow too large, they will trigger security mechanisms within the network or within individual facilities, causing them to go offline.

When data centers in Northern Virginia detected the fluctuation caused by the power line failure, they switched to backup power, removing their load from the grid. As more data centers made the switch, they removed greater amounts of load from the network. What started as a relatively small drop in supply turned into an even larger drop in demand, causing supply to increase and causing light bulbs to flicker.

Most data centers make split-second decisions, and the ones that went offline this week appear to be no different. When the voltage drop hit them, they all decided to disconnect within a few seconds of each other, Ali Zain Banatwala, senior market modeling specialist at the Independent Power System Operator, told TechCrunch.

“We need to find a way for these loads that are located next to each other to be disconnected or reconnected sequentially,” he said. A more orderly process would allow network operators to develop more robust procedures in advance.

Alternatively, data centers could be built to absorb disruptions and not turn their backs on them. One startup, ON.Energy, has been working on a product to help data centers (and the grid) weather events like the one that occurred this week.

The company has developed an uninterruptible power supply system for an entire data center campus, encompassing not only servers but also chillers and other equipment. Basically, the company hides the data center behind a bank of batteries connected to sophisticated power conversion equipment. All the network “sees” is consistent, well-behaved load rather than the peaks and valleys of each individual part of the data center. ON.Energy’s system allows data centers to ramp up and down computing workloads, including AI training, without disturbing the network.

Perhaps more importantly, it also means data centers can absorb grid power fluctuations. Instead of disconnecting from the grid, the ON.Energy system can use any extra power to charge its batteries, and if the flow drops, the system can send power to the servers. Additionally, it can track network leadership in milliseconds, avoiding dips or surges like those that caused this week’s issue for PJM.

ON.Energy is currently installing a total of 3 gigawatts of its systems across four different data center campuses, de Azevedo said.

Network managers have also realized the problem.

ERCOT, for example, will require large loads, such as data centers, to “ride out” outages, de Azevedo said.

However, time is ticking. This week’s mass shutdown was twice as large as a similar event in 2024, when 60 data centers went offline simultaneously, removing 1.5 gigawatts of load from the grid. Back then, data centers accounted for about 6% of PJM’s load, according to Synapse Energy Economics. By 2040, they are expected to represent 24%. If the problem is not fixed soon, things could get much worse.