The storage industry talks about gigawatt-hours and four-hour durations, but the unglamorous truth is that a battery only earns money if it can actually move power onto and off of the grid without destabilizing it. That interface - the bidirectional inverter that converts the battery's DC into grid-compatible AC and back - is where a surprising amount of engineering and project risk sits. A paper posted to arXiv on June 16, 2026, by Qiang Fu, Wenjuan Du, Siqi Bu, and Haifeng Wang takes that interface seriously and asks a question with direct deployment consequences: as you connect more and larger battery stations, what happens to grid stability, and does it matter whether you connect them via AC or DC?

The authors frame battery storage as part of a broader class of "controlled DC resources" that increasingly attach to AC grids. Their concern is converter-driven stability in the subsynchronous frequency range - oscillations slower than the grid's nominal frequency that can build up when many inverter-based stations interact. The setup is stated up front.

"Massive controlled DC resources (CDCRs), such as battery energy storage systems, are connected to AC power systems through bidirectional inverters for power balance requirements."- Fu et al., arXiv:2606.17565, source

What the study finds

The paper compares the AC and DC connection of inverter-based stations by varying three things: the number of controlled DC resources, the direction of power flow (charging versus discharging), and the control parameters of the inverters. Two findings stand out. First, for AC connections, the authors report that the stations "may induce instability as the number of CDCRs increases, regardless of the power flow direction." That last clause is the important one. It means the destabilizing tendency is not confined to discharge or to charge - it shows up whichever way the batteries are pushing power, which removes the comforting assumption that you only have to worry about stability in one operating mode.

Second, the authors find that switching to DC connections "can reduce these instability risks if the DC line resistance is much less than the AC line reactance," and that tuning the inverter control parameters is more effective at improving critical stability under DC connections. Their conclusion is that the DC-connected configuration is preferable for high-voltage transmission. To keep AC-connected stations stable, they calculate a maximum power amplitude the station can push - effectively a stability-derived cap on how hard you can run the interface.

Why a storage desk should care about a stability paper

This is not, on its face, a markets story. But the economics of large-scale storage are increasingly gated not by the battery cells but by interconnection - the queue, the studies, and the equipment required to attach a project to the grid without causing problems for everyone else on it. A finding that AC-connected battery fleets can court subsynchronous oscillation as they scale is, in effect, a finding about a hidden cost and constraint on deployment. If a station's stable operating envelope is capped by a maximum power amplitude, that cap interacts with how much revenue the asset can earn in fast-response markets where pushing hard is the point.

The DC-versus-AC framing has commercial teeth too. The paper's result that DC connection can lower instability risk and respond better to control tuning is a point in favor of DC-coupled and HVDC-adjacent architectures for the largest storage deployments - precisely the configurations that carry different equipment costs and different interconnection-study burdens. The conditional matters: the benefit holds when DC line resistance is much smaller than AC line reactance, which is a physical condition specific to the line, not a universal license. But for project developers choosing a topology, the study is evidence that the connection decision is a stability decision, and therefore a cost-and-revenue decision, not just a wiring detail.

The scaling finding deserves particular weight because it cuts against an intuitive assumption. It would be natural to assume that adding more battery stations to a region simply adds more flexible capacity and therefore more stability headroom. The paper's result is the opposite for AC connections: instability risk grows with the number of controlled DC resources, and it does so independent of whether those resources are charging or discharging. For a market that is racing to interconnect ever-larger fleets of inverter-based storage in the same congested corridors, that is a caution about a coordination problem. The interconnection studies that already gate storage projects exist precisely to catch system-level interactions like this one, and a finding that more inverter-based stations can erode stability is, in practical terms, a finding that those studies and any required mitigation - filters, control retuning, or a different connection topology - become more demanding as a region's storage density rises. That demand is a cost, and it lands on exactly the marginal projects whose economics are already tightest.

How far to trust it

The appropriate caution applies. This is a preprint, and its conclusions rest on modeling "validated in power systems connected with both AC- and DC-IBSs under various network topologies and system scales" - simulation across cases, not a measured field deployment. Stability behavior in a real grid depends on the specific network, the mix of other resources, and protection and control settings that no general study can fully capture. The maximum-power-amplitude figure is a method, not a single number a developer can lift off the page.

Still, the strategic read is durable. As the storage fleet grows, the inverter interface becomes a binding consideration, and the AC-or-DC connection choice carries stability consequences that show up in deployment cost and operating limits. The work is a reminder that the question "does utility-scale storage pencil?" eventually runs through power-electronics and interconnection engineering, not just cell pricing - and that the answer can differ depending on how the battery is wired into the grid in the first place. For a markets desk that prefers to follow the unit economics, this is the upstream constraint that quietly shapes them.