The Challenge of Performance-Based Design

Ronald O. Hamburger, Structural Engineer
EQE International, Inc.
Oakland, California

Over the past ten years, leading structural engineers have promoted the development and application of performance-based seismic design concepts. As practiced today, performance-based seismic design initiates with a discussion between the client and engineer about appropriate performance objectives. The engineer then endeavors to deliver a design capable of meeting these objectives. Performance objectives are expressed as an acceptable level of damage, typically categorized as one of several performance levels, such as immediate occupancy, life safe or collapse prevention, given that ground shaking of specified severity is experienced. The development of this approach has been a natural outgrowth of the evaluation and upgrade process for existing buildings. In that process, an engineer would typically be requested by an owner to evaluate the likely performance of a building, the owner would determine the acceptability of that performance, and then were it deemed unacceptable, request that an upgrade be designed to meet a specific performance expectation.

Initially, the practice of meeting performance-based objectives was rather informal, nonstandard, and quite qualitative. Some engineers would characterize performance as life-safe or not; others would assign ratings ranging from poor to good. This qualitative approach to performance prediction was appropriate given the limited capability of seismic-resistant design technology to deliver building designs capable of quantifiable performance. In 1992 the Federal Emergency Management Agency (FEMA) sponsored the development of national consensus guidelines for the seismic retrofit of buildings, the ATC-33 project. This was the first attempt to standardize the performance-based approach. That project standardized the qualitative descriptions of performance previously used into a series of quantifiable performance levels that could be predicted through the use of specific design parameters (element force and displacement demands). This approach was then quickly adopted by SEAOC's (Structural Engineers Association of California) Vision 2000 project and extended to include the design of new buildings. Together, the FEMA-273 NEHRP (National Earthquake Hazard Reduction Program) Guidelines for Seismic Rehabilitation of Buildings, resulting from the ATC-33 project, and the Vision 2000 report have defined the current state of practice in performance-based engineering and have created a significant market for this service within the building community.

Although the technical approach adopted by FEMA-273 and the Vision 2000 report was a necessary step in the development of performance-based design concepts from previous design approaches, I believe it is critically flawed in several aspects. First, the basic performance evaluation technique adopted by these documents is component- rather than system-based. The behavior of individual structural components and the damage they sustain is important to the evaluation of overall building behavior, but the failure of one or more isolated elements to meet specific acceptance criteria for a particular performance level should not necessarily imply that the building as a whole will not perform to a desired level. In the development of the FEMA-273 document, the ATC-33 project team struggled with this issue of relating system performance to component performance but was unable to develop a specific tool other than the judgment of the individual engineer in the designation of primary and secondary elements to bridge this gap. Since engineering judgment is widely variant, it is quite possible that engineers can employ the same evaluation and design methodology yet arrive at very different predictions about what performance will result from a particular design.

Structural performance prediction inherently involves a number of uncertainties and variabilities. It is difficult to predict precisely the character of ground motion a building will experience, the strength of the building materials, the quality of construction, the amount of force and deformation to individual building elements, and the capacity of these elements and the building system to resist these demands. Given these intrinsic uncertainties, performance prediction and performance-based design should be conducted on a probabilistic basis, with the probability of exceeding a certain desired performance and the confidence in characterizing this probability clearly expressed. As currently practiced, performance is expressed in a very deterministic manner. The engineer asks the owner what performance level is desired for a given level of ground shaking, and then proceeds to design to that specification—having little understanding of the likelihood of actually achieving the client's request. Clearly this approach creates significant potential liability for the practitioner if the client is not informed at the outset of a project that, despite the optimum application of current technology, the performance of the building will still be uncertain. I believe that many engineers currently practicing neither understand the existence of this uncertainty nor communicate it to their clients.

The foregoing highlights what I believe are the key challenges facing engineers and researchers working to extend our performance-based design technology today. First, we must find ways to more explicitly evaluate system as opposed to element behavior and performance in a manner that can be standardized in the design process. Second, we must develop a methodology to characterize the uncertainties and variabilities inherent in performance prediction in a way that can be understood by both the designer and the building owner, and implemented in a routine manner by the designer.

References

Federal Emergency Management Agency. NEHRP Guidelines for the Seismic Rehabilitation of Buildings. 1997. Prepared by the Applied Technology Council. [Washington, D.C.: Federal Emergency Management Agency].

FEMA 273. Structural Engineers Assn. of California (SEAOC), Vision 2000 Committee. April 3, 1995. J. Soulages, ed. Performance Based Seismic Engineering of Buildings. [Sacramento, Calif.:]. 2 vols.