A mathematical model for post-disaster road restoration: Enabling accessibility and evacuation

https://doi.org/10.1016/j.tre.2013.10.009Get rights and content

Highlights

  • Scheduling of road network restoration efforts under limited equipment.

  • Dynamic path based mathematical model maximizing network accessibility during restoration.

  • Decomposition based solution approach that distributes equipment optimally among regions.

  • Ability to solve realistic size problems according to computational results.

Abstract

This paper focuses on the planning of road restoration efforts during disaster response and recovery. The primary objective is to maximize network accessibility for all locations in the area during the restoration process so that survivors are evacuated and road side debris is removed as soon as possible. We propose a dynamic path based mathematical model that identifies criticality of blockages and clears them with limited resources. This model is more efficient than link based models and can solve restoration problems for realistic size networks within reasonable time. Algorithm performance is demonstrated using two instances based on districts in Istanbul.

Section snippets

Introduction and problem description

Post-disaster road restoration constitutes the first step in disaster response and recovery (FEMA, 2007). In any kind of disaster, whether hurricane or earthquake, the goal is to maximize survival rates. It is essential to be able to reach survivors and offer them relief and a possibility to evacuate the affected region during the first few days after the disaster strikes. Road network disruptions impede timely access to help and delay evacuation to shelters.

This paper addresses the issue of

Literature survey

Phase 1 activities of disaster recovery and response involve clearing roadside debris and restoring the road network in order to open up evacuation routes and other important lifeline paths so that traffic flow is enabled in affected areas. The restoration operation can be conducted efficiently by identifying the optimal order in which critical blocked links in the road network are cleared. The goal is to maximize the overall earliness of path restoration times, which leads to maximizing

The mathematical model

In this section, we formally define the Debris Clearance Scheduling Model (DCSM) that schedules the road restoration work in a region with the goal of maximizing the total weighted earliness of all cleared paths. As mentioned in Section 1, the DCSM is an integer programming model solved to optimality for each zone or district.

Before we can define the DCSM, we should provide the notation used in the model.

Sets and parameters:

PSet of predefined paths, indexed by pP,
ASet of links, indexed by iA,
A

Computational results

We demonstrate our approach on two test instances based on the road networks of two districts in Istanbul, Turkey. The first district (Caddebostan) depicted in Fig. 1a contains 212 road segments out of which 49 are blocked, whereas the larger (Fatih) district has 386 segments where 79 are blocked (Fig. 1b). The restoration times for blocked segments vary between 1 and 10 h. For both districts, the set of predefined paths P has been constructed as explained in Section 3. In Fig. 1a, the seashore

Conclusion

This study addresses a crucial aspect of disaster response: restoration of road networks. Here, the goal is to make all locations in the affected area accessible for receiving help and evacuation. The restoration process is carried out by a limited number of equipment. It is essential to restore the network as much as possible during the first 3 days of response in order to maximize survival rate.

Several approaches have been proposed in the literature to solve this problem. These can be

References (21)

There are more references available in the full text version of this article.

Cited by (0)

View full text