A key element of emergency planning is evacuation planning. Originally formalized as a requirement of the nuclear power industry, and performed for coastal areas prone to hurricane activity, evacuation planning has been extended to other population centers where such planning is deemed prudent and timely. Now, in what is widely regarded as a changed world with much greater emphasis on emergency planning, evacuation planning has taken on a new urgency.
Planning an evacuation trip differs fundamentally from the normal trip-planning process. Normally, the purpose of an everyday trip is to arrive at a selected destination by a specified time. The traveler generally has knowledge and experience to assist trip planning: The location of the trip destination is known; available routes between the trip origin and destination are known; the time of day during the trip is anticipated; traffic conditions can be anticipated; and they depend on the time of day of the trip.
Planning an evacuation trip is an entirely different process. The purpose of an evacuation trip is to depart an area at risk so as to minimize the hazard associated with the causative event. The need for evacuations can be triggered by natural events (hurricanes, floods, earthquakes), technological events (toxic materials release, transportation accidents, industrial accidents) or acts of war or terrorism.
Regardless of the causative event, an evacuee most likely has little knowledge or experience to guide the evacuation-trip planning process.
Evacuation trips may be planned after the event occurs (no-notice evacuation) or after a period of mobilization.
Evacuation routes may be event-dependent. Wind direction, speed and the toxicity of the hazard all determine the extent of the area to be evacuated and the evacuation routes to be utilized.
The immediate goal of an evacuation is to leave the area at risk as soon as possible. Once this goal has been achieved, additional travel may be necessary to arrive at a host facility, reception center or other locations where evacuation services may be found. It should be noted that the “best” route out of the area at risk need not be the best route to a final evacuation destination.
Since evacuations are statistically rare events, few, if any, evacuees have experience driving in evacuation traffic conditions.
Evacuation can be initiated during any period of the day; timing is event-driven.
Evacuating the sims
The transportation aspects of evacuation planning can be studied using computer simulation techniques. Evacuation modeling can be performed with either microscopic models or with macroscopic models.
Micro models move individual vehicles through a network. These vehicles have their own characteristics and their drivers respond to the presence of other vehicles, traffic-control devices and their need to follow a specific route. Macro models operate not on individual vehicles but on overall traffic flow on a given link of the network.
The tradeoffs between micro and macro models are generally that micro models provide a more detailed simulation at the expense of computing speed, whereas macro models are capable of simulating large areas operating under high-traffic conditions very quickly. As the developers of both microscopic and macroscopic models, KLD Associates Inc., Commack, N.Y., has used both classes of models for evacuation purposes.
A comparison between microscopic and macroscopic modeling was recently conducted based upon an evacuation study conducted for the Nine Mile Point Nuclear Power Station in New York state. The network consisted of 964 links. The evacuation time estimates produced by both models were similar (less than a 5% difference). However, the microscopic model took 300 times as long to produce comparable results (35 minutes vs. 0.1 minute). Where large networks are involved or large numbers of evacuation scenarios must be studied, the macroscopic modeling approach provides reasonable accuracy at a significantly higher level of software efficiency. This level of efficiency makes the macroscopic modeling approach a viable candidate for use in real-time emergency planning applications.
Existing planning models require trip tables, which are matrices of origin-destination trips. Often, these tables are disaggregated by trip purpose. Origin-destination traffic demands are based on existing patterns of normal trip-making activities. Existing planning models can perform dynamic assignment, but routing remains inflexible. They accommodate static control, and they tend to be computationally slow. Consequently, most of them are designed for off-line planning purposes.
KLD developed the IDYNEV Evacuation Modeling System under contract to the Federal Emergency Management Agency to address these issues. Use of IDYNEV does not require the development of a trip table. The TRAD Model component integrates trip distribution and traffic assignment and is designed expressly for evacuation planning. The model is computationally extremely fast.
Existing evacuation planning systems have been used to plan for events occurring at fixed facilities in response to well-defined hazards. These include the areas surrounding nuclear power stations and coastal areas prone to hurricane strikes. These planning tools must be expanded so that they can be used interactively in real-time. Such a system should be capable of performing the following functions:
- Create evacuation routing schemes in minutes to respond to any emergency event;
- Support changes to these routing schemes in real-time as circumstances dictate;
- Train emergency-response personnel in an interactive-simulation facility; and
- Interface with other software systems such as: crisis management, hazardous material databases, plume dispersion, meteorological modeling, geographic-information systems, traffic-control plans, census planning databases and animation displays.
Thinking outside the paradigm
Each of the software systems identified currently exists as an off-the-shelf product. Linking these systems with an evacuation-planning model gives emergency planners an opportunity to fundamentally change the evacuation traffic-routing paradigm.
Traditionally, traffic assignment for nonemergency planning purposes is based on the identification of paths between trip origin and trip destination that minimize an objective function consisting of travel time or vehicular delay or both. Combining the emergency planning tools identified would allow the creation of a new objective function for traffic routing. This function—minimizing the population exposure to hazard—can potentially identify evacuation routes that protect the health and safety of the population at risk under the conditions unique to an emergency requiring an evacuation response.
One of the implications of this approach is the need to transmit evacuation route information to the public and to enforce the desired routing during an evacuation. Additionally, in the event of an emergency, efforts also must be made to reduce discretional travel demand in the area to give priority to evacuation traffic flows. Intelligent transportation system technologies can facilitate transmitting route information to the public. Applicable technologies include variable message signs, highway advisory radio broadcasts, television, commercial radio and Internet websites.
Enforcing desired evacuation routing is largely a matter of exercising traffic control along evacuation routes. The implementation of traffic control is vital to the safe and timely evacuation of the public from the high-risk area. Traffic volumes and patterns resulting from an emergency evacuation would generally differ from those characteristic of normal traffic conditions at that time of day. Traffic-signal control is designed to serve normal traffic demand and routing. Consequently, it is advisable to assign traffic-control personnel at (usually signalized) intersections to effectively direct evacuation traffic.
Since the traffic-control tactics at a specific location are dependent on the evacuation routes passing through that site, police and other traffic-control personnel must be provided with materials instructing them on the physical layout of the site (for instance, where to place traffic cones, barricades and other support equipment) and on the traffic movements to facilitate or discourage. It is important to note that the terms “facilitate” and “discourage” imply that the function of the traffic-control personnel is not to force evacuees to take a specific route, but rather to direct them to the most appropriate path. In many instances, evacuees have good and valid reasons for wishing to divert to other paths. They may be going to pick up children or relatives so that the family can evacuate as a unit. Traffic-control personnel should not unduly delay these people but rather should accommodate them within the overall traffic-control plan being implemented.
Escaping the Point
The Indian Point Energy Center (IPEC) is located on the eastern shore of the Hudson River in Westchester County, N.Y. The site is situated approximately 35 miles north of New York City. The emergency planning zone (EPZ) contains parts of four New York state counties: Orange, Putnam, Rockland and Westchester.
Under peak evacuation scenarios, a full evacuation of the entire EPZ would involve over 300,000 vehicles; 239,000 vehicles would evacuate from the area at risk, while an additional 63,000 vehicles would move on roads in the surrounding areas (i.e., the “shadow” evacuation region). These shadow vehicles are modeled because their presence on the road can potentially delay the departure of evacuees from the area at risk. Evacuation modeling results indicate that a full evacuation of the area would take from about 7 hours (in the evening) to about 12 hours (during the midweek period in snow). These times represent the estimated time for the last vehicle to leave the area. On average, an individual evacuation trip would last between one and two hours.
A traffic control point (TCP) is generally an intersection identified in the evacuation plan where special traffic-control treatment by traffic guides is needed. The selection of TCPs for the Indian Point EPZ was made by KLD using its expertise in traffic engineering, based on repeated surveys of the highway system and on the results of the simulation analysis of evacuation-traffic flow. The simulation model created an animation display that shows the changing levels of service on all links in the evacuation network.
After identifying the locations where the need for TCPs was indicated, KLD prepared detailed schematics of all intersections defining the traffic control that should be implemented to facilitate and safely control the flow of evacuating traffic. These schematics were sent to police departments for review. Police officers have an intimate knowledge of these intersections and how they operate. Thus, they offer added insight and experience into traffic control at these intersections. KLD and law enforcement personnel worked together to develop the final traffic-management plans. The resulting TCP schematics were enhanced by adding overhead photos of the TCP locations. These photos serve to provide landmarks to ensure that the traffic guide will travel to the correct location. Finally, booklets containing the final TCP schematics and overhead photos were delivered to each county.
An important concept discussed with police agencies is the location priority assigned to each TCP and presented on the schematic. The priority was established to help agencies decide where, and in what sequence, resources will be allocated in the event of an emergency at Indian Point. Priorities originally established by KLD were adjusted with county and law-enforcement input. The priorities range from 1 through 3, with 1 being the most important and 3 the least important. Those intersections with Priority 1 will be manned first in the event of an emergency. If additional manpower is available, then Priority 2 TCPs will be manned and so forth. Clearly, if resources permit, all TCPs will be manned.
The Indian Point case study is an example of the traffic-management procedures and tools available to planners and police personnel at fixed sites where the evacuation planning is in response to a known hazard and is performed in accordance with regulatory requirements. Pre-determined traffic-management plans are essential to effectively implement evacuation routing because required resources, direction of traffic flow and priority of staffing have been identified during an extensive planning process before the event.
In a real-time evacuation-planning system, traffic-control plans would need to be created and transmitted to field personnel with minimal user input. Due to the dynamic nature of events, real-time planning techniques and implementing procedures must be flexible enough to support changes in routing instructions.
Although real-time evacuation routing is software-driven, there still is a significant role for engineers to play in the emergency planning process. Routes and traffic-control procedures, created in real-time by software, must be vetted by traffic engineers with local expertise and by local law enforcement and emergency responders with intimate knowledge and experience.
Communication between planning software and software used in the field is important. Traffic-control plans need to be transferred to police computer systems for implementation in real-time.
Most of the tools required to create no-notice evacuation routing plans are available as commercially available, off-the-shelf software. Among the issues that need to be addressed are those of linking the pieces together to form a cohesive system and of providing the proper training and preparation for local emergency-response planners in the use of such a system. TME
RSS: Roads & Bridges Articles
