As performance based seismic design continues to garner interest in the mainstream construction industry [1], some building owners are expressing a desire to target performance levels above those specified by code requirements. While this shift in interest and motivation towards higher seismic performance might be a welcomed change for some designers, the determination of the quantity of investment, as well as the most efficient application of the investment to achieve the desired performance, involves understanding the complex relationships between the seismic performance of a structure and its nonstructural components. This task can be daunting as determining the most efficient performance target for each component within the entire building system requires the consideration of multiple interconnected factors such as:
In order to determine the most efficient resource application across various structural and nonstructural upgrade options while considering these interconnected factors, an optimization methodology has been developed which considers the minimizing of seismic losses against the capital investments required for any number of defined seismic upgrades. The process is based on a genetic algorithm and provides enough flexibility to be able to consider various upgrade decision variables as well as different target optimization metrics. The results obtained from this optimization process can be used by building owners and non-engineering professionals to better conduct investment planning and risk mitigation analysis throughout an asset’s life cycle.
A hypothetical case study is used as an example in this article to demonstrate the results obtained by the optimization method. In this scenario, an institutional client wishes to determine the optimal seismic retrofit upgrade investment strategy given a desired annual discount rate of 4% and an expected occupancy time of 40 years. To demonstrate the flexibility of the process, a secondary objective seeks to determine the optimal quantity of retrofits with the goal of minimizing future potential downtime due to seismic events. The client’s asset consisted of an existing three-story steel moment resisting frame which has pre-Northridge connections, an office type occupancy, and was located in Seattle, Washington.
The potential retrofit scenarios considered for this frame included 4 different structural retrofits implemented on any number of the building floors: (1) the improvement of the connection detailing using self-centering sliding hinge joints (SCSHJ), (2) the installation of buckling restrained braces (BRB), and the installation of viscous dampers targeting either (3) 10% or (4) 25% damping in the first mode of vibration. The scenarios also included the retrofit of 26 different nonstructural components commonly found in office type occupancies [2]. A summary of the various upgrade options is summarized in its optimization algorithm form in Figure 1 (a), while an elevation and plan of the structure is shown in Figure 1 (b), (c), respectively.
The first objective investigated was the reduction of economic cost and used the net present value to select the upgrade expenditures which provided only a positive return over the buildings life cycle. The second objective was the reduction of downtime, which compared the time required for retrofit against the reduction in expected downtime due to seismic activities over the buildings life cycle. The functions for both these objectives are shown in equation 1 and 2, respectively.
Where EALO and EALU are the estimated annual loss of the original and upgrade scenario, respectively, R is the discount rate, t is the expected occupancy time, UC is the upgrade cost, EATO and EATU are the estimated annual downtime of the original and upgrade scenario and UT is the downtime required to perform the retrofit. Each objective was run multiple times to ensure convergence on similar results as the several layers of uncertainty are modeled in a probabilistic Monte Carlo method [3].
A summary of the results is shown in Figure 2 for both targets. The priority of each available retrofit option is displayed, where a higher value indicates the propensity for a retrofit to be selected. Components with a low priority number indicate that the variability in their retrofit costing implementation results caused the net benefit of the retrofit to approach zero. This variability can be controlled with more accurate retrofit cost estimations, as those provided by contractors. Non-selected components indicates a net negative return on the investment.
When considering structural retrofits, the 25% viscous damping strategy is selected for both target metrics. While the capital cost for this strategy was the highest, the reduction in damage to both the acceleration sensitive nonstructural components and increase in the structural performance outweighed the investment required at this specific owner scenario. When considering the nonstructural components, some variation in the optimal selection occurred between objectives, where retrofits of expensive equipment, such as the mechanical equipment and the piping, are selected for the economic reduction while retrofits of the finishes, such as the glazing and partitions, are selected for the downtime reduction target.
The components selected for either targets consisted of components whose damage states had a high likelihood of occurrence before the collapse of the structure, and whose cost or time required for retrofit far outweigh the cost or time required for repair. Furthermore, ulterior factors linked across individual components caused the optimal identification of retrofit to components which did not necessarily satisfy the above criterions. As in the case with various piping components when retrofitted as a uniform system, the reduction in cost of related flooding damage was only provided by fully upgrading the entire system, and the economies of scale of retrofitting all of the systems at once provided a net positive benefit.
Beyond the optimal identification of individual components for specific business cases, owner scenarios, or performance targets, this optimization methodology can also provide a rapid process of comparing the effects of changes to the dynamic properties of the system on the supported nonstructural components. This is particularly useful during the early stages of design development, where multiple fundamental structural design approaches might be under consideration. This is particularly important when other factors, such as architectural impacts or owner preferences, might impede the implementation of the most optimal structural solution.
While the optimization process identified the 25% viscous damping in the first mode as the optimal structural retrofit solution for the case study, the impact of the other structural systems on the optimal component selection is examined further for the economic reduction target. The discount rate and occupancy time were also modified to measure different owner scenarios, where lower discount rates and longer occupancy times provide a more generous capital investment scenario. Figure 3 summarizes the results obtained for all 4 structural retrofit options, where the graphs provide an indication of the return on investment (ROI) by comparing the net present value of the estimated annual loss versus the cost of retrofit implementation.
A comparison across the different structural modifications indicates that the use of SCSHJ connections increases the drift capacity of the structure before collapse occurs, but requires some significant investment in the drift sensitive nonstructural components to maintain a positive ROI. When examining the BRB strategy, the retrofit of acceleration sensitive nonstructural components is critical to offset the increase in peak floor acceleration caused by the additional BRB stiffness. Finally, the lower quantity of damping requires a larger investment in nonstructural retrofits when compared to the selected 25% viscous damping strategy. The 25% viscous damping strategy, while more expensive structurally, appears to provide a more advantageous return on investment and a lower overall expense regarding the retrofit on nonstructural retrofits.
Overall, this method provides guidance on the optimum seismic design or retrofit of buildings within the Performance-Based Earthquake Engineering (PBEE) framework. This algorithm integrates the consideration of both structural and nonstructural upgrades based on system-level performance and loss estimations. The capabilities of the optimization procedure were illustrated through the seismic retrofit case-study of a three-story steel moment-resisting frame archetype building where the optimal retrofit strategy was determined for either an economic target, where an investment of 14% of the buildings value provided a loss reduction of 25%; or a downtime target, where 21 days of construction reduced the estimated number of downtime days by 240 over a 40 year occupancy. The algorithm can also be easily modified for the determination of any number of target performances desired by building owners in the future. More information can be found in Steneker et al. [4].
References:
1. Moehle, J., and Deierlein, G. (2004). “A Framework Methodology for Performance-Based Earthquake Engineering.” Proceedings of the 13th World Conference on Earthquake Engineering (WCEE), Vancouver, BC, Canada
2. FEMA (2011). “Reducing the Risks of Nonstructural Earthquake Damage-A Practical Guide.” Report E-74 Federal Emergency Management Agency, Washington, D.C.
3. FEMA (2012a). “Seismic Performance Assessment of Buildings - methodology.” Report P-58 Federal Emergency Management Agency, Vol. 1, 1-278, Washington, D.C.
4. Steneker, P., Filiatrault, A., Wiebe, L., Konstantinidis, D. (Forthcoming): “Integrated Structural-Nonstructural Performance-Based Seismic Design and Retrofit Optimization of Buildings.” Journal of Structural Engineering, DOI: 10.1061/(ASCE)ST.1943-541X.0002680.