The Colorado River Basin is one of the most closely studied examples of a water system’s operating envelope eroding under prolonged reduced inflows — not through catastrophic failure, but through routine conditions slowly becoming emergencies.
The Crisis
The numbers are stark:
- The Colorado River’s flow has fallen an estimated 20% this century — research attributes much of that decline to warming-driven evapotranspiration increases across the basin
- Lake Powell and Lake Mead have both sat below 30% capacity in recent years
- In 2023, Lake Powell’s surface level dropped within 30 feet of minimum power pool — the threshold below which hydropower generation must cease
- 25 million people and billions of dollars in agriculture depend on this system
But the infrastructure problem runs deeper than low water levels.
The Engineering Failure Mode
At very low storage levels, the normal operational spillways and outlet works at Glen Canyon Dam cannot be used — they require sufficient hydraulic head. The only remaining water release mechanism becomes the River Outlet Works.
The River Outlet Works has a critical design flaw: it experiences damaging cavitation at the flow velocities required when it’s the sole release pathway. It was designed as a backup system, not a primary one. When water levels dropped to the point where it became the primary release mechanism, operators were caught between releasing water through a system that damages itself in doing so, or not releasing water at all.
It’s a pattern worth recognising: infrastructure designed around a stable operating range can fail well before a reservoir empties, once that range narrows enough.
The Broader Lesson
The 1922 Colorado River Compact — the foundational water allocation agreement across seven US states — was negotiated during an unusually wet period. It created “paper water” entitlements that exceed actual average supply by a significant margin. Legal rights were allocated against a resource that was never as abundant as the negotiators believed.
As Wade Graham summarised in High Country News (Feb 2026): “Lake Powell — and with it, the entire Colorado River system — is perilously close to operational failure.”
The Australian Parallel
We’re not immune to this pattern, and it’s worth asking how much of our own dam infrastructure was designed around historical inflow records that are now being revised downward. Spillway adequacy assessments and operating rules approved against 20th-century hydrology are a reasonable place to start checking.
The questions worth asking:
- Which of our major storages are operating with reduced buffer against inflow variability? Not just current storage levels, but structural reductions in average inflows.
- Are operating rule curves based on inflow assumptions that are no longer valid? Particularly in the Murray-Darling and southern Australia where long-term drying trends are clearest.
- Where are our “River Outlet Works” equivalents? Every major dam has a failure mode that was designed as a last resort. Climate change is steadily shrinking the operating envelope that keeps those last resorts from becoming the primary option.
Worth asking before the answer gets forced on us: how much runway is actually left?
Source: “The coming failure of Glen Canyon Dam” — Wade Graham, High Country News, February 2026. Originally shared on LinkedIn — 1,696 impressions.