When Is a Site-Specific PSHA Worth the Effort?

In the previous article, I discussed what a site-specific probabilistic seismic hazard analysis (PSHA) actually does. The conclusion was that a site-specific PSHA is not a fundamentally different type of analysis than the one used to develop the national hazard maps. It is the same methodology, applied with more project-specific information and evaluated directly at the site of interest.

That naturally leads to the next question:

When is all of that additional effort actually worth it?

Many discussions about site-specific hazard analyses quickly turn into discussions about project types. People will often say that hospitals need them, bridges need them, tall buildings need them, and ordinary buildings probably do not.

While those rules of thumb can be useful, they miss the more fundamental question.

The decision to perform a site-specific PSHA is ultimately a decision about uncertainty.

The question is not:

"Is this project important enough?"

The question is:

"Would a better understanding of the seismic hazard change an engineering decision?"

What Are You Actually Buying?

One of the most common misconceptions about site-specific hazard studies is that they exist primarily to produce a different response spectrum.

Sometimes the resulting spectrum is lower than the mapped spectrum.

Sometimes it is higher.

Sometimes the difference is negligible.

The value of a site-specific study is not the spectrum itself. The value is the information that comes with it. In other words, you are paying to reduce uncertainty, not necessarily to reduce design forces.

A site-specific PSHA can provide a better understanding of:

• Which seismic sources control the hazard.

• Which assumptions have the greatest influence on the results.

• How sensitive the hazard is to alternative interpretations.

• How local site conditions influence the expected shaking.

• How much confidence should be placed in the resulting spectrum.

In many cases, this additional understanding is more valuable than the spectrum itself.

When Additional Information Has Limited Value

Not every project benefits equally from a more refined hazard model.

Some structures are relatively insensitive to modest changes in the seismic hazard. Low-rise buildings, stiff structures, and projects where seismic demands are not governing often fall into this category.

Similarly, if the design objective is simply to satisfy the prescriptive requirements of the building code, the modern USGS Multi-Period Response Spectrum (MPRS) will often provide an excellent basis for design.

In these situations, the additional effort associated with a site-specific study may have little influence on the final engineering decisions.

When Additional Information Becomes Valuable

The value of a site-specific study generally increases as the consequences of uncertainty increase.

Long-Period Structures

Tall buildings, long-span bridges, base-isolated structures, and other flexible systems are often sensitive to the long-period portion of the spectrum.

For these structures, differences in spectral shape, basin effects, source characterization, or ground-motion models can have a meaningful influence on both design demands and predicted performance.

Sites with Complex Geology

Some sites are not adequately described by a single site-class designation.

Deep sedimentary basins, strong impedance contrasts, unusual soil profiles, and complex geologic conditions can all influence the amplitude, frequency content, and duration of ground shaking.

In these situations, a site-specific study may provide insights that are difficult to obtain from the mapped spectrum alone.

Sites Near Major Faults

Hazard does not always change gradually with distance.

Near major seismic sources, relatively small changes in location can sometimes produce noticeable differences in predicted ground motions. While engineers often think about site-specific studies in terms of source characterization or site effects, the location itself can become important. The USGS maps estimate hazard at a site by interpolating values computed at surrounding grid points. In most situations this approach works extremely well and provides an excellent estimate of the seismic hazard. However, when hazard varies rapidly over short distances, there can be value in evaluating the hazard directly at the project coordinates.

This does not imply that the mapped values are incorrect. Rather, it reflects the fact that the closer a site is to the sources controlling the hazard, the more important local details can become.

This issue has received increasing attention in recent years, particularly for sites located near major faults where hazard gradients can be steep. It is a topic that I and, likely, others have raised during discussions of future hazard-map development, and the USGS is actively evaluating improvements to grid spacing in areas where hazard can vary significantly over relatively short distances. These efforts are intended to further improve the representation of local hazard while maintaining the practicality of a national mapping framework.

Performance-Based Design

Performance-based design often involves sophisticated nonlinear analyses and explicit performance objectives.

When significant effort is devoted to understanding structural behavior, it is reasonable to ask whether the hazard representation deserves a similar level of attention.

For these projects, a site-specific PSHA is often viewed not as an additional expense, but as an important component of the overall design process.

Site-Specific PSHA vs. Site-Response Analysis

At this point, it is worth noting that site-specific PSHA and site-response analysis are not the same thing, even though the terms are often used interchangeably.

A site-specific PSHA focuses on estimating the seismic hazard at a particular location by evaluating earthquake sources, recurrence relationships, ground-motion models, and site characteristics. In many cases, the analysis may be performed using the measured VS30 of the site without requiring a detailed evaluation of the entire soil profile.

A site-response analysis addresses a different question.

Rather than characterizing the site through a single parameter such as VS30, it explicitly models how seismic waves propagate through the soil layers beneath the site. This typically requires detailed geotechnical information, including shear-wave velocity profiles, soil layering, dynamic soil properties, and estimates of nonlinear soil behavior.

As a result, site-response analyses are often significantly more labor-intensive than the hazard analysis itself. They are typically performed when local soil conditions are expected to have a substantial influence on the ground motions, when required by code provisions, or when the project warrants a more detailed evaluation of site effects.

Although the two analyses are often performed together, they are independent. A project may use a site-specific PSHA without performing a site-response analysis, and conversely, a site-response analysis may be performed using ground motions derived from the mapped hazard.

The 80% Floor

At this point, some readers may be wondering whether a site-specific study is simply a mechanism for obtaining a lower design spectrum.

The answer is no.

ASCE 7 recognizes that site-specific studies contain uncertainty just as national hazard models do. As a result, the code includes minimum-spectrum requirements relative to the mapped spectrum. Depending on the provisions being applied, these requirements often result in portions of the site-specific design spectrum being required to remain at least 80% of the corresponding mapped values.

The intent is not to suggest that the site-specific analysis is incorrect. Rather, it reflects the recognition that multiple scientifically defensible interpretations of the hazard may exist. The minimum-spectrum provisions provide a safeguard against unconservative results arising from modeling assumptions, limited data, or alternative interpretations of the hazard.

This reflects an important philosophy of modern seismic design: site-specific studies are intended to improve our understanding of the hazard, not simply reduce design forces.

The existence of these minimum-spectrum provisions is another reminder that the objective of a site-specific study is not to obtain a smaller spectrum. The objective is to obtain a more realistic representation of the hazard.

The Wrong Question

Perhaps the most common question asked about site-specific hazard studies is:

"Will it reduce my spectrum?"

That is often the wrong question.

A better question is:

"Would a more detailed understanding of the seismic hazard change the decisions I am making?"

If the answer is no, the mapped spectrum may be entirely sufficient.

If the answer is yes, a site-specific study may provide information that cannot be obtained from the maps alone.

The goal is not to make the analysis more complicated. The goal is to reduce uncertainty where that uncertainty matters most.