1. Project Goals/Objectives:

The goal is to provide tools for retrofit decisions in metropolitan transportation networks. The objective of Year 7 research activities is to extend and leverage to work to date completed the PEER Highway Demonstration Project by developing a model capable of estimating the economic losses associated with trips eliminated from the transportation network following an earthquake.

2. Role of this project in supporting PEER’s mission (vision):

These results build closely on what has already been done, and will be relevant to both MAE and MCEER researchers pursuing related research. The proposed models would fit usefully into the FHWA/MCEER REDARS 2.0 Project. REDARS 2.0 is expected to be out in Beta form mid 2004. REDARS does not currently include economic loss models, and there are no plans to extend research at MCEER to address this problem. Therefore, it would greatly add to the tri-center initiative to provide the basis for implementing a travel-foregone module into REDARS. Once implemented, such a model would make REDARS a more attractive tool for investigating other issues of interest to PEER, for e.g.., evaluating the economic impact of improved bridge performance and/or the ability to more accurately predict performance through improved fragility models.

3. Methodology Employed:

Year 7 Work: Accounting for the Efficiency Implications of Demand Shifts

The standard version of the network design problem focuses on the total cost of travel on the network, treating travel demand as exogenous. This approach is inadequate in the earthquake context.

3.1 Accounting for Demand Shifts Across Time-of-Day (Fan, UCD)

In an earthquake context, this standard perspective must be extended to accommodate shifts along the travel demand function of the sort accounted for the variable travel demand formulation. The work done to date advances the modeling state of art to accomplish this, but should be further extended to account for demand shifts across time of day. Time-of-day travel demand profiles are usually taken as exogenous in most standard network modeling applications. In the case of an earthquake, some trips would be eliminated, and some will shift merely be shifted to other periods. These shifts are as yet un-modeled, but can be modeled as a straightforward extension of the work completed to date.

3.2 Quantifying Economic Losses Due to Demand Shifts (Moore, USC)

Work to date parameterizes the travel demand function to account for reductions in demand as a result of lower levels of service. In addition, we believe that these results can be further leveraged to provide the means to compute these net changes in benefits for flows occurring between each origin-destination pair, and for freight and passenger flows. The opportunity to estimate changes in the total net benefits provided by travel is perhaps the most useful consequence of accounting for the elasticity of travel demand. Fig. 1 depicts two kinds of changes in the net benefits of travel associated with a reduction in transportation supply. The area of the grey rectangle is the loss of net benefits that would otherwise accrue to continuing users of the system, who now experience a lower level of service than they did before the earthquake. The area of the black triangle is the reduction in net benefits due to trips forgone.

Survey-based transportation origin-destination matrices published by the Metropolitan Transportation Commission (MTC) and the Southern California Association of Governments (SCAG) provide the opportunity to calibrate the estimated travel demand function with observed California urban travel patterns. Previous work investigating Los Angeles travel patterns following the Northridge Earthquake may provide further opportunity to validate the model.

Fig. 1 Reductions in the Net Benefits of Travel on a
Network Damaged by an Earthquake

3.3 Extensions to Other Lifeline Systems (Moore, in collaboration with PEER researchers)

The methodology and tools proposed here can extend to other lifeline systems. This research will provide a plan of how the approach might apply to other lifeline systems currently being considered by PEER – primarily electric power systems. This extension would eventually help in understanding the impacts of earthquakes on electric power systems, and perhaps the inter-relationships between lifeline systems.

4. Brief Description of past year’s accomplishments (Year 6) & more detail on expected Year 7 accomplishments:

The work completed to date in PEER projects A5, 104199, and 3222001 links earthquake damage to transportation structures to transportation network performance at a metropolitan scale. The work to date identifies three likely areas of new inquiry:

Post-earthquake reconstruction decisions are dictated by many factors outside the direct control of transportation planners. Consequently, the proposed work will be directed to the basic questions of economic impact combined with pre-event decisions. Year 7 research will focus on the first question. The economic costs of congestion and foregone trips can be analyzed without double-counting these costs. These two impacts are separately measured. This information is needed to evaluate pre-earthquake retrofit options.

5. Other Similar Work Being Conducted Within and Outside PEER and How This Project Differs:

Prof. Yueyue Fan, UC Davis Department of Civil and Environmental Engineering is submitting an identical statement of work. Profs. Moore and Fan collaborated on PEER projects 104199, and 3222001, when Prof. Fan was a doctoral student a USC, and they will continue to collaborate on the Year 7 research proposed here. The USC and UCD budgets are separate, and total $70,000.

The NSF EERC Year 6 Annual Report, Volume 3: Tri-Center Collaboration identifies a number of the ways in which transportation research at the PEER, MCEER, and MAE centers relates. The work described here strengthens these connections further. The FHWA/MCEER REDARS project now incorporates the variable demand modeling approach developed as part of PEER project 3222001. Further coordination will be achieved as the Caltrans test deployment of REDARS in the Bay Area proceeds. The Year 7 work described here is a logical extension likely to be relevant to the Caltrans REDARS project and any similar efforts.

6. Plans for Year 8 if project is expected to be continued:

Work Subsequent to Year 7

The best way to treat retrofit decisions is as a large-scale transportation network design problem. This is a difficult class of problems that has been the subject of substantial investigation in the literature. Conventional approaches to these problems combine mathematical programming with bi-level control or implicit enumeration techniques. Even these well-investigated techniques will be difficult to apply to in combination with large, metropolitan area models.

6.1 Stochastic Network Design

Pre-earthquake facility decisions are more complicated than post-event decisions. These are perhaps best represented as examples of the stochastic transportation network design problem, which focuses on the performance of degraded networks [Bell and Iida, 1997; 2001]. The objective is to find the transportation network configuration on which user equilibrium flows produce the least expected total congestion, subject to retrofit budget (and possibly other) constraints. Unfortunately, the stochastic version of the problem is an embedded optimization problem with a tri-level structure. The upper level is the decision by the network authority; in this case a pre-event retrofit or reconstruction decision. The intermediate level outcome, a function of the upper level decision, is a random result of nature, i.e., the earthquake. The lower level, a function of the upper level decision and the intermediate outcome, is the decision by the traveler.

Explicit enumeration of options is out of the question. The solution space for the stochastic version of the problem is even larger than for the deterministic problem. A transportation network with M links presents 2^M retrofit options. A random act of nature converts the network to a collection of L < M links. Thus, from a computational perspective, pre-event bridge retrofit decisions are vexingly difficult to optimize in a network context.

6.2 The Role of Heuristics

A way forward, is still required, of course. Public authority has a compelling economic incentive to make rational decisions about the seismic retrofit of transportation structures, regardless of how difficult it is do so optimally. Practical alternatives must be identified and evaluated. Further, transportation authorities need to be able to respond quickly and cogently with plans when presented with special budget opportunities to undertake seismic retrofit and reconstruction program. Since such pre-earthquake decisions cannot yet be treated optimally, they must certainly be handled heuristically.

7. Describe any actual instances where you are aware your results have been used in industry:

 

8. Expected Milestones & Deliverables:

This one-year effort will culminate with a report that will be suitable for immediate stand-alone publication by the PEER Center, or as part of a larger multi-year report. The report will be completed within 90 days of the project end date, or on any other reasonable reporting schedule preferred by PEER. Brief interim reports will be produced at intervals no more frequently than quarterly, and preferably at a six-month interval. In addition, the investigator will honor reasonable requests to travel to meet with PEER, Caltrans, or related personnel to discuss the project.