A new design strategy based on a deterministic definition of the seismic input to overcome the limits of design procedures based on probabilistic approaches

In this paper, a new seismic Performance Based Design (PBD) process based on a deterministic definition of the seismic input is presented. The proposed procedure aims to address the following considerations, arisen from the analysis of seismic phenomena, which cannot be taken in account using standard probabilistic seismic input (PSHA): a) any structure at a given location, regardless of its importance, is subject to the same shaking as a result of a given earthquake, b) it is impossible to determine when a future earthquake of a given intensity/magnitude will occur, c) insufficient data are available to develop reliable statistics with regards to earthquakes. On the basis of these considerations, the seismic input at a given site - determined on the basis of the seismic history, the seismogenic zones and the seismogenic nodes - is defined using the Neo Deterministic Seismic Hazard Assessment (NDSHA). Two different analysis are carried out at different levels of detail. The first one (RSA) provides the Maximum Deterministic Seismic Input as a response spectra at the bedrock (MDSIBD), similarly to what is proposed by the codes. The second one (SSA) takes the site effects into account, providing a site specific seismic input (MDSISS). A SSA provides realistic site specific seismograms that could be used to run time history analysis even where no registrations are available. Reviewing the standard PBD procedure, MDSISS is always associated with the worst structural performance acceptable for a building, called Target Performance Level (TPL). In this way, the importance of the structure (risk category) is taken into account by changing the structural performance level to check rather than to change the seismic input.

💡 Research Summary

The paper introduces a novel Performance‑Based Design (PBD) framework that replaces the conventional probabilistic seismic hazard assessment (PSHA) with a deterministic definition of the seismic input. The authors argue that PSHA suffers from three fundamental shortcomings: (i) it treats every structure at a given site as receiving the same shaking regardless of its importance, (ii) it cannot predict when a specific magnitude earthquake will occur, and (iii) the available earthquake catalogue is often too sparse to support reliable statistical estimates. To overcome these issues, the authors adopt the Neo‑Deterministic Seismic Hazard Assessment (NDSHA) methodology, which builds a physically based model of the seismogenic environment (zones and nodes), assigns maximum credible moment‑tensor parameters to each source, and propagates the resulting wavefields through realistic Earth models.

Two levels of analysis are defined. The first, Regional Seismic Assessment (RSA), computes the Maximum Deterministic Seismic Input at Bedrock (MDSIBD). This is expressed as a response‑spectrum that can be directly compared with the spectra prescribed by current building codes, thereby preserving regulatory compatibility. The second, Site‑Specific Assessment (SSA), incorporates local site effects (soil non‑linearity, basin geometry, amplification, etc.) to generate the Maximum Deterministic Site‑Specific Input (MDSISS). SSA delivers realistic, site‑specific accelerograms that can be used for time‑history analysis even in the absence of recorded strong‑motion data.

A key conceptual shift concerns how structural importance (risk category) is treated. Traditional codes modify the seismic input by applying importance factors or design spectra with different return periods. In the proposed deterministic framework the seismic input remains unchanged for a given site; instead, the designer selects an appropriate Target Performance Level (TPL) that reflects the acceptable worst‑case performance for the building (e.g., Immediate Occupancy, Life‑Safety, Collapse Prevention). By varying the TPL rather than the input, the same deterministic ground motion can be used to evaluate structures of different importance, ensuring that safety is governed by a physically justified maximum shaking scenario while economic considerations are handled through performance targets.

The authors illustrate the workflow with case studies that demonstrate how RSA provides a code‑compatible baseline, whereas SSA refines the input with site‑specific amplification and produces synthetic accelerograms suitable for nonlinear dynamic analysis. The deterministic spectra are intentionally conservative (maximum credible), guaranteeing that the structure is designed for the most severe shaking that could plausibly affect the site.

Overall, the paper makes several important contributions:

1. Physical Realism – By modeling the source, path, and site physics, NDSHA yields a seismic input that reflects the true maximum credible shaking, bypassing the statistical uncertainties inherent in PSHA.

2. Data‑Independent Hazard – The method does not rely on long‑term earthquake catalogs; it can be applied even in regions with limited historical records, provided that geological and seismotectonic information is available.

3. Unified Treatment of Importance – Importance is expressed through performance objectives rather than altered hazard levels, simplifying code compliance and aligning design philosophy with the core principle of PBD: design for a target performance under a defined hazard.

4. Practical Implementation – The two‑step RSA/SSA approach allows engineers to first meet code requirements with a bedrock spectrum and then refine the analysis with site‑specific effects, facilitating a smooth transition from conventional practice to the deterministic paradigm.

5. Enhanced Safety Margin – Because the input is the “Maximum Deterministic” value, structures are inherently designed with a built‑in safety margin that is transparent and physically justified, reducing the reliance on probabilistic safety factors that may be poorly calibrated.

In conclusion, the deterministic PBD process proposed in this study offers a robust alternative to probabilistic seismic design, addressing the intrinsic uncertainties of PSHA while preserving compatibility with existing codes. By anchoring design on the most severe physically plausible earthquake and by linking structural importance to performance levels, the methodology promises safer, more economical, and scientifically grounded seismic design practices.