Seismic Vulnerability of the Highway System

Task B1-2: Enhancements and Programming of REDARS

Subject Area: Loss Estimation Methodologies
Research Year 3

Principal Investigator and Institution

Stuart D. Werner, Seismic Systems & Engineering Consultants


The objective for Research Year 3 of Task B1-2 is to begin the development of REDARS as a public-domain software package for seismic risk analysis (SRA) of highway systems. This will include: (a) software development planning and initial programming, (b) technical improvements to the various REDARS modules, (c) development of guidelines for assessing whether the number of simulations assumed for a user's probabilistic SRA is sufficient, and (d) development of plans and procedures for adapting REDARS for use as a decision guidance tool that will facilitate evaluation of alternative seismic risk reduction strategies for a highway system. These efforts will provide the framework for work during future years of MCEER Highway Project 094, which will focus on programming, documentation, and example application of REDARS.


Research on this task will be conducted via the following subtasks:

Subtask 1 - REDARS Software Development  During Research Year 2, Task B1-2 evaluated various technical and software development options to be considered for the development of the REDARS software package during future years of the project. For each option, evaluations considered scope, costs, assessments of the importance of the option for meeting anticipated user needs, and whether the overall cost of the option could be accommodated within estimated project budget constraints for the remainder of this project. From this, a general plan for developing REDARS during the remaining years of the project was developed.

Under this subtask, work will be initiated to develop the REDARS software in accordance with these plans. This will include more detailed planning of the program structure and specific features to be included in the software package, as well as initial programming efforts. The detailed planning will focus on the basic REDARS platform into which the hazards, component, and system modules will interface, the specifics of these interfaces, input database redesign specifics, and graphic/GIS features and editing options.

Subtask 2 - REDARS Module Development  Enhancements will be made to the systems, component, and hazards modules, as summarized below. For the systems module, enhancements will include post-earthquake traffic demand modeling, improvement of the minimum search path algorithm, and investigation of the MAM matrix training scheme in concert with the effort under Task B1-4. Work on the component module will address two main issues: modeling of the performance of bridges subjected to liquefaction induced ground displacements, and the performance of roadway pavements during earthquakes. Work on the hazards module will address two main issues: liquefaction hazard modeling, and incorporation of earthquake and ground motion models for additional regions of the United States.

It is well known that bridges can be damaged by liquefaction of the backfills or subsurface soils during an earthquake. Although REDARS will include a credible model for estimating site-specific liquefaction hazards, it does not yet include a sufficiently credible model for estimating how bridges and their foundations perform when subjected to these hazards. Procedures for developing fragility curves for bridges subjected to liquefaction-induced permanent ground deformation will be investigated, along with similar procedures for roadway pavements. This will include review of: (a) empirical data and damage reports from past earthquakes, (b) results from experimental and analytical programs, and (c) data on repair procedures, costs, and downtimes for bridges damaged by liquefaction-induced ground deformation. In addition, engineers and highway-roadway maintenance personnel from Caltrans who have observed and repaired bridges subjected to liquefaction damage during recent earthquakes will be interviewed. From this, a strategy will be developed for modeling the performance of bridges subjected to liquefaction.

In recent discussions, several DOT agency representatives also underscored the importance of including the performance of roadway pavements when evaluating seismic risks to highway systems. A first order model of the performance of pavements during earthquakes was prepared under Project 106 (see Section 5.5 of MCEER Technical Report 00-0014). This will be further investigated to determine how this model can be updated, through review of empirical data from past earthquakes, and interviews will be conducted with engineers and highway maintenance personnel from Caltrans who have observed and repaired roads that were damaged during recent earthquakes.

REDARS currently includes a liquefaction hazard analysis procedure developed as part of the earlier Project 106. Although well suited for application at individual bridge sites (where soil exploration and testing to obtain the required detailed soils input data is practical), this procedure is less practical for application to multiple bridge sites throughout a spatially distributed highway system where these detailed soils data are often not readily available nor easily obtained. The applicability of alternative liquefaction hazard models for use in REDARS will therefore be assessed. These models should build on prior databases yet require less extensive site soils data that are more likely to be available at the large numbers of bridge or component sites within a spatially dispersed highway system. One promising candidate procedure that will be considered in this context is the recent liquefaction modeling work of Professor Bardet at USC.

Finally, for REDARS to be applicable to highway systems throughout the United States, it must include a range of earthquake and ground motion models that encompass the range of seismologic and geologic conditions that may be encountered in various regions of moderate to high earthquake activity nationwide. REDARS currently includes such models for the Central U.S., and the inclusion of models for California is being planned for a possible future project by Caltrans. During Year 3, earthquake and ground motion models for one additional region of the country will be incorporated into REDARS (e.g., Charleston, SC). This will serve as a template for including models for other regions of the nation into REDARS during future years of this project.

Subtask 3 - Confidence Levels  MCEER Technical Report 00-0014 describes an approach for estimating nominal confidence levels for the average annualized loss, as a function of the number of simulations used in the risk analysis. Under this subtask, work will be done to clarify how to also determine confidence levels for point estimates of loss (e.g., 100-year loss estimates), and how confidence levels can be used to assess whether the number of simulations used in a REDARS application to a given highway system is sufficient. The subtask will also assess if methods are available for reducing the number of simulations needed for a given REDARS analysis, while still achieving a desired confidence level for the results. If successful, this would simplify future REDARS applications by reducing run-time requirements.

Subtask 4 - Decision Guidance Framework  An important aspect of REDARS is its suitability for use as a decision guidance tool. This will require adapting REDARS to provide cost and risk results for alternative seismic risk reduction strategies, in forms that could be used to guide decision making. This subtask will establish a framework for using REDARS as a decision guidance tool, by showing how it can be adapted to several of the more common decision procedures available. This use of REDARS as a decision guidance tool will be demonstrated in a future example application of REDARS; i.e., one that is planned for Research Years 5 and 6, after REDARS software development has been completed.


  • REDARS Software Development Progress Report 1 - Description of the software development steps carried out during Research Year 3, problems encountered and resolved, and an updated plan for continuing software development during Research Year 4 and beyond.
  • Technical Report - Documentation of results of subtasks 2 through 4, problems encountered and resolved, and discussion of how these results and models will be incorporated into REDARS during Research Year 4.

Technical Challenges

The challenges for Research Year 3 will be: (a) laying the groundwork for eventual development of REDARS as a technically sound, practical, efficient, and useful software package; and (b) coordination of the modeling and analysis results from subtasks 2 through 4, and from Tasks B1-3 and B1-4 into the software planning in subtask 1, so that software planning and subsequent programming proceeds efficiently.

return to Task Statement List