Reliable Is Not the Same as Resilient: What NERC's 2026 Report Is Really Telling Us

On June 24, NERC released its 2026 State of Reliability report, an assessment of 2025 bulk power system performance. Its opening message is appropriately measured: the North American bulk power system continued to provide reliable electric service under increasingly challenging conditions.

That statement is true. It is also not the full story.

I have spent my career around grid monitoring, asset protection, and condition-based intelligence. At Qualitrol, we see reliability pressure not only as a system-level issue, but at the asset level: transformers expected to operate longer, generators cycling more often, substations serving load patterns they were not originally designed around, and customers asking for earlier warning before an equipment issue becomes an outage.

That is why the NERC report deserves close attention. It does not describe a grid in crisis. It describes a grid that is still performing, but with less structural margin behind that performance.

The Grid Is Performing, but the Margin Is Thinning

There is a difference between a system that is reliable today and one that has enough structural margin for what is coming next.

The North American grid remains reliable in the first sense. NERC's data is clear on that. But the same report identifies several signs that the underlying margin is weakening.

The annual weighted equivalent forced outage rate for conventional generation reached 9.2% in 2025, above the historical range of roughly 7% to 8%. NERC also reported year-over-year increases in unavailable energy of 39.8 TWh from coal units and 19.1 TWh from combined-cycle units.

The report notes that this increase was not driven by one major event. It was distributed across much of the year, suggesting a broader decline in baseline availability for parts of the coal and combined-cycle fleets.

That matters because these assets still provide a large share of dispatchable capacity. Many are older, and NERC notes that large coal and combined-cycle units were not designed for regular cycling. Cycling, thermal stress, maintenance demands, spare-part constraints, and limited technical resources eventually appear in availability.

What concerns me is not one metric. It is the pattern.

Forced outage rates are rising. Demand is growing. Large new loads are connecting faster than traditional planning processes were designed to absorb. Transmission development is not moving at the same speed as load growth. NERC states directly that major transmission development is not advancing at a pace that can adequately support future reliability needs.

This is how structural risk often appears. Not as one dramatic signal, but as more exceptions, more manual judgment, more conservative operating decisions, and less room for error when several stresses arrive together.

Operational Mitigation Is Not the Same as Structural Adequacy

NERC describes meaningful industry responses, including more conservative operating practices, greater use of probabilistic and all-hours assessments, improved model validation, resilience analysis for extreme weather, and additional guidance for large loads and load-pocket risks.

These are important actions. They reflect serious work by experienced professionals.

But we should be clear about what increasing dependence on operational mitigation means. It means operators are being asked to compensate when the physical system, planning assumptions, or asset base no longer provide the margin they once did.

That does not make these actions wrong. It makes them a warning sign.

The grid has historically absorbed stress through reserve margins, fleet diversity, transmission redundancy, and deep operating experience. The 2026 report suggests that several of those buffers are thinning at the same time.

Reserve margins become harder to rely on when forced outage rates increase. Fleet diversity changes as dispatchable thermal assets retire and the replacement mix behaves differently. Transmission redundancy becomes less forgiving when high-density loads cluster in specific corridors.

The workforce challenge is harder to quantify, but it is visible. NERC's review of prolonged generator outages identified supply chain constraints, difficulty locating part specifications, inexperienced project management, and insufficient staff among the contributing issues. The report also recommends increasing the number of trained technical maintenance personnel.

In the field, these pressures appear in practical ways: longer repair cycles, greater dependence on scarce experts, more caution around outage windows, and rising value for any tool that helps operators understand an asset's condition before it fails.

Data Centers Are Exposing a Speed Mismatch

Large computational loads are one of the most important new stress vectors on the grid.

The issue is not simply that data centers consume significant power. The larger issue is their speed, concentration, and operating behavior.

Data centers can move from concept to energization much faster than transmission systems can be planned, permitted, and built. NERC identifies data center and computational loads as a growing reliability concern because of their scale, speed of development, and unique operating characteristics.

In 2025, NERC recorded two data center customer-initiated load reductions that exceeded 1,000 MW, along with many more events above 100 MW. The report describes individual Eastern Interconnection events involving reductions of approximately 1,800 MW and 1,300 MW.

That is not normal commercial load behavior.

A large, voltage-sensitive facility may respond to a grid disturbance exactly as designed from the customer's perspective while still creating a system-level reliability challenge. If several facilities in the same corridor respond in a similar way, the grid can experience a sudden and concentrated load swing.

This is not an argument against data centers. They are essential to the digital economy. It is an argument for treating them as major grid participants rather than passive loads.

The traditional planning model was not built for large, fast-growing computational loads clustering around the same transmission constraints. That gap must close.

BESS Helps, but It Does Not Eliminate the Need for Sustained Capacity

Battery energy storage systems are playing a valuable role. NERC recognizes their contribution to rapid frequency response, frequency recovery, and smoothing variable renewable output.

That contribution is real.

But BESS is not the same as sustained dispatchable capacity. NERC notes that battery systems are energy-limited, typically intended to operate for a few hours. Their role during a frequency disturbance is to respond quickly and support the system until other resources can come online for longer-duration support.

The report is direct that this limitation matters during widespread, long-duration system stress and that BESS is not a complete solution for events such as major winter storms.

Planning frameworks need to preserve this distinction. A resource that can arrest a frequency decline is not necessarily a resource that can carry the system through an extended period of high demand, low renewable output, and reduced conventional availability.

This matters because reliability challenges increasingly result from combinations: higher forced outage rates, weather stress, transmission limitations, data center behavior, and changes in the resource mix occurring at the same time.

No single technology solves that.

Visibility Has to Improve at the Asset Level

The next reliability advantage will come from seeing risk earlier.

Traditional monitoring approaches were developed for a more stable grid, with large synchronous generation, more predictable load shapes, and assets operating closer to their original design assumptions. That is no longer the system we are managing.

Today, operators need more than status information. They need condition intelligence.

They need to know not only whether an asset is energized, but whether it is degrading. They need to understand whether a transformer, bushing, breaker, generator, or substation asset has enough remaining margin for the duty being placed on it. They need earlier warning when thermal, electrical, mechanical, or insulation stress is moving in the wrong direction.

This is where asset intelligence becomes part of reliability, not only maintenance.

When replacement lead times are long, every critical asset becomes more valuable. When demand is growing faster than infrastructure, every avoidable failure matters more. When experienced people are scarce, the system needs better evidence to support faster decisions.

That is the practical meaning of grid visibility.

The Work Ahead

The NERC report should not be read as alarmist. The grid is still delivering reliable service, and the professionals managing it deserve credit for that performance.

But the report should not be read as simply reassuring either.

A grid can be reliable while losing margin. A system can perform well while becoming more fragile. Operational excellence can compensate for structural pressure for a long time, until a combination of conditions exceeds the available margin.

The 2003 Northeast blackout showed how localized failures can cascade when situational awareness and system safeguards are insufficient. The lesson is not that every warning becomes a crisis. It is that visibility and structural margin matter most before the system crosses a point from which recovery becomes difficult.

The industry needs to act on three fronts.

First, we need better visibility into asset condition and system behavior, especially where aging infrastructure is serving new load patterns.

Second, we need planning assumptions that reflect how large computational loads, inverter-based resources, batteries, and conventional assets actually behave during disturbances.

Third, physical infrastructure, especially transmission, needs to advance at a pace closer to the demand being placed on the system.

The honest takeaway is simple: the grid is not failing, but it is being asked to operate with less slack. That should concern anyone responsible for reliability.

The next reliability challenge will not be solved by operational skill alone. It will require rebuilding structural margin and identifying asset risk early enough to act before reliability becomes a recovery exercise.

That is the work that matters now.

References

North American Electric Reliability Corporation (NERC). (2026). 2026 State of Reliability: Assessment Overview of 2025 Bulk Power System Performance. Published June 24, 2026.
https://www.nerc.com/globalassets/programs/rapa/pa/nerc_sor_2026_overview.pdf