Most dam safety conversations start with the wrong number. They count structures — over 90,000 documented constructed dams in the United States — and treat that inventory as the risk metric. It isn't. The number that actually determines whether a dam fails quietly or catastrophically is simpler: how long since the last inspection, and what did that inspection actually reach?
That second part is where the math gets uncomfortable.
Below the Waterline Is Where Problems Hide
A surface inspection of a dam tells you what the surface looks like. It tells you nothing about what's happening at the outlet structure, the gate seating, or the internal drainage paths 80 feet down. The Forest Service learned this directly at Council Bluff Lake in Missouri, where a scheduled underwater inspection — conducted in near-freezing water with near-zero visibility — found a 24-inch manhole cover blocking a flood control gate. The gate couldn't seat properly. Nobody knew. The problem was invisible from the surface and would have remained invisible until a high-water event demanded the gate actually close.
That underwater inspection is scheduled on a five-year cycle. Five years is a long interval for a mechanical system under continuous hydraulic load. The Forest Service engineers who wrote up the Council Bluff work noted that the woody debris found lodged in the level control structure "could have eventually moved further down inside the structure and been a bigger problem if not removed." That's the failure cascade in miniature: debris accumulates, gates bind, flood control capacity degrades, and the degradation is invisible until you send a diver down.
The question the inspection backlog literature never quite answers directly: how many of those 90,000 dams have had a below-waterline inspection in the last five years? The answer, based on available public data, is not reassuring — but the precise figure requires FEMA's National Inventory of Dams data that isn't fully extracted in current sources. What's documented is the structural reality: underwater inspection requires specialized dive teams, confined space certification, and crane equipment. It's expensive, logistically complex, and easy to defer.
Aging Structures, Shifting Loads
The inspection backlog problem is compounded by a design mismatch that's accelerating. The Bulletin of the Atomic Scientists flagged this in May — dam failures are "slowly unfolding" events driven by aging infrastructure meeting climate conditions the original designs didn't anticipate. A dam built in 1960 was sized to a hydrological record that ended in 1960. The spillway capacity was calculated against the probable maximum flood of that era. Both numbers have since changed, and not in the dam's favor.
This isn't a theoretical concern. The 1889 South Fork Dam failure — the event that prompted National Dam Safety Awareness Day — involved a blocked spillway, deferred maintenance, and a structure that had been modified without engineering rigor. The failure mode is not exotic. Debris blocks a gate. Seepage erodes an embankment. A spillway sized for a 100-year storm encounters a 500-year event. The structure was inspected on schedule; the inspection didn't catch what mattered.
Continuous Monitoring Doesn't Replace Inspection — It Changes What Inspection Finds
The U.S. Army Corps of Engineers Nashville District operates ten dams across the Cumberland River Basin, and their approach illustrates what a well-resourced program actually looks like. Brad Long, the Nashville District's Dam Safety Program manager, described the framework as "portfolio risk management" — embedded digital sensors providing continuous structural data, with physical inspection cycles layered on top to catch surface wear, mechanical issues, and environmental vulnerabilities the sensors can't see.
Long's framing is worth holding onto: "In modern engineering, no dam is considered '100% safe.' Instead, our dam safety programs focus on continuous risk reduction — identifying the most likely failure pathways and systematically investing resources to interrupt them." That's the honest engineering position. The sensors tell you the dam is behaving within parameters right now. The inspection tells you whether the parameters themselves are still valid.
The problem is that the Corps' ten-dam Cumberland Basin program represents the well-funded end of the spectrum. The 90,000-dam national inventory includes thousands of structures owned by municipalities, agricultural districts, and private parties who don't have regional dive teams on call. For those owners, the inspection interval isn't a policy choice — it's a budget constraint dressed up as a schedule.
The Metric Worth Tracking
The FEMA National Inventory of Dams tracks inspection dates alongside hazard classifications. High-hazard dams — those where failure would likely cause loss of life — are supposed to be inspected on tighter cycles than low-hazard structures. The gap between "supposed to" and "actually inspected" is where the real backlog lives, and it's a number that state dam safety programs publish inconsistently.
Watch for state-level dam safety program budget cycles this fall. Several states are currently in appropriations processes that will determine whether inspection staffing expands or contracts. The inspection interval is a policy output, not a natural law — and right now, the math on deferred assessments is running ahead of the math on completed ones.
