How does ASCE treat risk?

Why Earthquake Design Allows Higher Collapse Risk (and Why That’s Okay)

As you know, I am one of the proponents of a proposal that is working its way through the ASCE7 process. We are proposing an adjustment to the risk (of collapse) target for seismic design. So I asked ChatGPT to compare the current risk target of 1%/50yr for seismic hazard to that of other hazards considered in ASCE. As usual, the response was very helpful.  In this post, we'll break down why seismic design allows a higher collapse probability, how it compares to other hazards, and why that balance reflects both engineering reality and societal choice:

When it comes to structural design, not all hazards are treated equally — and for good reason. Earthquake design standards, such as those in ASCE 7, allow for a higher risk of collapse than for hazards like wind, snow, or flood. This isn’t a flaw in the system, but rather a reflection of how different hazards behave, how predictable they are, and the economics of construction.

How ASCE treats collapse risk for earthquake compared to other hazards

Hazard Type	Typical Targeted Risk of Collapse	Basis
Earthquake (Seismic)	~1% probability of collapse at MCER level	Based on nonlinear response, uncertainty, and multiple performance objectives.
Wind	Implicit collapse probability often < 0.1%	Based on strength design, relatively linear behavior up to failure, fewer uncertainty factors.
Snow	Implicit collapse probability very low (~0.1% or less)	Snow loads usually do not induce sudden global collapse; more controlled failure modes.
Flood	Typically governed by hazard recurrence, e.g. 100-yr or 500-yr events	Risk depends on design flood elevation; not typically expressed as collapse probability.
Tornado / Extreme Wind (ASCE 7-22, Ch. 32)	Target risk often 1E-4 to 1E-5 annual collapse	High consequence, but low frequency events, treated probabilistically in design maps.

Why seismic is "allowed" to have higher collapse risk

• Nature of hazard: Earthquake loading is highly uncertain, both in frequency and intensity. The shaking demands are very difficult to predict for a specific site.

• Collapse mechanism: Seismic collapse is nonlinear, ductility-based, and highly sensitive to detailing, configuration, and dynamic interaction — more so than for most other hazards.

• Cost-benefit tradeoffs: Designing to wind or snow failure levels often results in much higher safety margins; replicating that level of conservatism in seismic design would be prohibitively expensive, especially in high seismic zones.

• Multiple limit states: ASCE 7 treats seismic design as collapse prevention at MCER, but damage control at lower return periods (e.g. frequent, design basis earthquakes).

  
  

Rough analogy

For seismic, society accepts a higher collapse risk because the hazard is rarer, more uncertain, and the cost of ultra-conservative design would be enormous.

For wind, snow, and flood, society expects very low collapse risk because these hazards are more predictable, and their design loads typically lead to more linear behavior up to failure.

Summary

“Seismic design allows higher collapse probabilities than other hazards, not because we value life differently, but because the nature of earthquake loading — its unpredictability, complexity, and the cost of fully eliminating collapse potential — makes a small, controlled residual risk tolerable. For more predictable hazards like wind and snow, much lower collapse probabilities are achievable and economically justified.”