Managing Pedestrians During Evacuation of Metropolitan Areas > Table of Contents

3  Interviews with Practitioners

Interviews were conducted by telephone with practitioners to obtain information from the perspective of representatives of transportation, emergency response, and emergency planning agencies. Contacts included agency officials who had operational experience with an emergency evacuation in an urban area or had engaged in planning for urban evacuation where pedestrians were considered.

3.1    Selection of Evacuation Events

The most relevant information on pedestrian evacuation identified in the literature review came from the transportation case studies (see Section 2.4). From these studies, it is reasonable to expect that emergency evacuations involving large numbers of people on foot might be necessary in urban areas with a relatively high use of mass transit for commuting to work. The extensive blackout of the Northeast in 2003 and the terrorist events of September 11, 2001 stand out as notable events where appreciable numbers of people evacuated part of a large city without using a private vehicle.

Very little information was found using key word searches and online data bases for reports on evacuations. The searches covered newspaper articles, the National Transportation Safety Board accident investigations, and federal disaster declarations. The results were mainly examples of evacuations associated with transport-related hazardous material spills and toxic releases in rural or small community and industrial settings, and evacuations prompted by hurricane warnings and flooding disasters.

The search for practitioner thinking on pedestrian evacuation was broadened to include evacuation planning. In recent years, the emergency preparedness community has moved away from plans for specific events, to an “all-hazards” philosophy. This approach is grounded in the observation that a basic concept of operations serves to guide response to terrorism attacks as well as emergencies related to natural events and technological breakdowns, with a few adaptations made depending on the nature of the event. The 9/11 attacks provided many important lessons to transportation officials and emergency planners because of the self-evacuations that occurred. Images in the media showing outbound pedestrians crowding the bridges in New York City may have prompted emergency planners in other cities to consider pedestrians as part of the evacuation mix. This expanded search yielded a few examples of emergency evacuation planning.

Fifteen representatives of transportation, emergency response, and emergency planning agencies agreed to participate in an interview about a relevant emergency event or planning activity. These interviews were conducted over the phone by the principal investigator. The notes from the telephone interviews were reviewed by the respondents and corrected as necessary. The urban area and the related emergency or planning activities are listed below.

Table 1.  Evacuation Situations Selected for Interviews with Practitioners

Although New York City officials have the most relevant experience with pedestrian evacuation, most agency representatives that were contacted declined the invitation to participate because of the high demand for interviews that they have experienced since 9/11. Consequently, some information from other anecdotal and published sources has been included in order to present observations about the evacuations in New York City.

3.2    Findings from Interviews with Practitioners

3.2.1   Washington, D.C., Metropolitan Region

The greater metropolitan region of Washington, D.C., includes the District and parts of the adjacent states of Maryland and Virginia. Although the Pentagon suffered the direct impact of the 9/11 attacks, a much more general evacuation was experienced in the District. Some information is presented here related to the evacuation of the Pentagon, but the interviews with the practitioners focused more on what happened in the District that same day. No interviews were attempted with Pentagon officials or Arlington, VA emergency response agencies.

The crash of the hijacked commercial airliner into the Pentagon on 9/11 represents a “no-notice” event in the Washington, D.C., metropolitan region. By the time this event took place, the major commuter routes in the vicinity of the Pentagon were filled with traffic that was barely moving. This traffic was caused by people who began to leave the District upon hearing about the attacks on the World Trade Center. These traffic conditions did delay some emergency units that deployed to the Pentagon, but no mention has been found that pedestrian traffic interfered with emergency responders.

Many people commute by Metrorail and bus to the Pentagon, which is served by a station under the facility. Metrorail services were programmed to by-pass this station after the Pentagon was hit, but commuter rail and bus services were still available at the next station on the line, which can be reached by a pedestrian underpass. Pentagon evacuees without cars who took this route to find transportation would have avoided the congested roadways. This is an example of how urban planning and design for pedestrian safety and convenience under normal conditions can be of particular benefit to pedestrians in emergency evacuations.

Telephone interviews were conducted with transportation and transit agency representatives for the Washington, D.C., area, and information was obtained during the subsequent project-related workshop. The interviews provided observations related to the sort of spontaneous evacuation that began in the District as people began hearing the news of an airliner hitting a World Trade Center tower. When the second tower was hit in the same way, the evidence that this was an attack and not an accident quickly changed the interpretation of the situation for federal agencies and workers.

The typical high-volume traffic and transit period of the morning commute was nearly over when the traffic began to reverse direction as people in the District began to leave. It became even more pronounced once the Pentagon was hit. The sidewalks and congested streets in some areas filled with people leaving office buildings in large numbers. The assumption is that most people left for home. At around 10:30 a.m., the Office of Personnel Management notified federal government workers that they could leave work, which further increased the outbound traffic.

No report was found that any kind of a formal recommendation was made by authorities for people to leave the District, but many other pre-planned security arrangements were implemented. Evidence for this is the sudden appearance of physical security perimeters around high-probability target areas like the White House and Capitol area, which forced pedestrians and vehicle traffic to take detours from their usual routes. At least one report states that a recommendation was made to the transit authority that Metrorail services be suspended, but transit officials declined to do so based on their assessment of the nature of the threat and the importance of transit services for people leaving the District. Some route changes were made, including by-passing the stations at the Pentagon and Reagan National Airport. In-bound traffic was deflected in several places. Since no impact area was located in the District itself, there was no major convergence of emergency response personnel to any specific location in the District.

 

 

The summary observations given below provide a few different perspectives, but do not cover all the agencies that were directly involved.

Challenges Described from the Perspective of Traffic Management

• Shortly after the hijacked airliner hit the Pentagon, the downtown area of the District experienced large numbers of pedestrians crowding sidewalks and often spilling into the roadway, where vehicles were gridlocked and at a virtual standstill. It is likely that the pedestrians were people who had used mass transit to get to work and people who knew that the traffic congestion would greatly slow down any effort to leave the District by car. [Note: it is not clear if vehicle traffic could not move at times because pedestrians were in the street, or if pedestrians spilled over into the street because the traffic was at a standstill.]
• The huge number of people leaving the District at the same time that morning is believed to represent voluntary collective behavior mainly on the part of people who worked in the District. Some people started to leave, and soon others made the decision to do the same. At that point, it probably would have been difficult to instruct people to not leave and would have heightened confusion and stress already evident because of the attacks. [Note:  The various motivations for leaving cannot be known from the information available. The situation was ambiguous. The already completed attacks suggested that there might be further targets, and Washington, D.C., is considered a major potential target. One motivating factor cited in disaster research documents is that people tend to seek out family and friends in threatening conditions.]  
• At the time of the 9/11 attacks in 2001 there were no traffic cameras at intersections in the District and on-line sources of information were unavailable. The transportation official made a reconnaissance trip to the busiest part of the District to conduct a visual inspection. The first-hand observation of the rapidly developing outbound vehicle traffic gridlock provided information important to the District DOT (DDOT) on which to base decisions about the deployment of their various assets.
• The mass media had wrongly reported that Metrorail was not operating; only two stations were closed for security reasons. This misinformation is thought to have caused many people to ignore the option of Metrorail, which would have provided a faster means of egress from the District than attempting to walk out or to use surface bus transportation. Besides the inconvenience, these slower modes could be regarded as extending an evacuee’s exposure to potential risk in the District. Also, these options may have been less familiar to evacuees than their usual commuting mode.
• The surface transit systems of the Maryland and Virginia counties adjoining the District are designed to link with the District Metrorail and Metro bus routes. On 9/11, commuters suddenly started back for the suburbs just at the end of the morning rush hour commute. It was difficult for the surface transit agencies beyond the District to know what to expect in the way of passenger surge, because a situation at the Washington Metropolitan Area Transit Authority (WMATA) Operations Center hindered communications.
• The mass media had wrongly announced that the bridges were closed. The traffic manager reported the vehicle and pedestrian scene along 14^th Street as chaotic, with some people behaving aggressively as drivers and pedestrians became annoyed with each other.
• The traffic manager observed that the security perimeters established around such places as the White House and Capitol Hill, and perhaps other federal buildings, served as obstacles to familiar routes, and created longer and unfamiliar routes for pedestrians as well as for vehicles.
• One official stated that Washington, D.C., has to be considered a target for future terrorist acts, and consequently needs to engage in aggressive preparedness activities. It is important to have the capability to move workers and residents out of the District quickly, to reduce exposure to secondary events. The typical week-day commuter traffic congestion illustrates the difficulty of achieving this goal if large numbers of people attempt to use private vehicles.
• The emergence of a “walking event” as a major evacuation mode seemed inevitable, given the number of people who attempted to leave work at about the same time to go home and the gridlock that developed with the surge of vehicle traffic which was beyond usual rush hour levels.

Challenges Described from the Perspective of the Transit Police

• Many people suddenly wanting to use Metrorail for evacuation didn’t seem to have any kind of a personal plan; they seemed to be relying on the authorities to give them explicit guidance.
• Metrorail on a routine day runs near the saturation point. On the morning of 9/11, when people began to evacuate the District, the passenger demand was even higher than normal. This placed extra demands on the transit police to maintain crowd control and passenger safety. On the other hand, since the New York City attack occurred fairly early in the morning, some portion of Washington commuters may have never even started their ride into the District as they monitored the news about New York City.
• One of the two Metrorail stations that were closed due to security reasons also had the largest bus station in association with it. Since passengers had to go on to the next station before leaving the train, the location for the Metro buses that were to connect with the Metrorail had to quickly be re-established at the other station to keep the passengers moving.

Responses or Adjustments Implemented

• On 9/11. Emergency plans developed during preparations for Y2K for traffic management were implemented. The Mayor for the District requested the National Guard be deployed to provide traffic management at the intersections designated in the plan as most critical, so that the police would be free to carry out a broader range of safety and law enforcement activities.
• On 9/11.  Metrorail was able easily to extend service at its rush hour capacity (provide more trains with more cars) because the rush hour shift of operators was still in place when the surge of people wanting to leave the District by Metrorail began.
• On 9/11. The Maryland transit agency was able to quickly institute its evening rush hour procedures just at the end of the morning rush hour because buses and operators were still on duty. This capacity was important to have in place to meet the demand of large numbers of passengers that had originated in the District headed for the Maryland residential suburbs.
• Existing Routine Procedure:  The transit company can provide “bus bridges” to take care of expected increases in passenger load, and to serve as a substitute for Metrorail services if they are interrupted.
• Post-9/11: Cameras have been installed at many critical intersections in the District which permit the remote assessment of emerging problems with respect to not only vehicles but pedestrians.
• Post-9/11: The traffic management agency has participated with other agencies in efforts to provide information to the public about alternatives to becoming part of the traffic congestion. These include education activities about how to shelter in place, and suggestions like having a pair of walking shoes available if walking home becomes the preferred adjustment to the traffic congestion.
• Post-9/11: The Citizens Emergency Response Teams (CERT) have been engaged to provide public education about evacuation options and the use of the subway, and have been trained on how to assist people with the subway system during an evacuation.
• Post-9/11: The Transit Police work with local police jurisdictions, fire departments, and public works agencies so these personnel will be familiar with Metro stations and procedures in the event of needed emergency assistance to passengers and to the Transit police.
• Post-9/11: Traffic managers still use spotters to go out to see what is happening on evacuation routes, so the District’s Traffic Management Center will know about such things as pedestrians on the bridges along with the cars and can decide what to do.
• Post-9/11: A RICCS (Regional Incident Coordination and Communication System) has been developed for the entire region (District, Maryland, Virginia) to provide a central point to which various agencies, including transportation agencies, can direct information about some emerging situation. Posted information can be viewed by other agencies, or conference calls can be arranged among selected agencies. This information dissemination enables a wide array of agencies to have information that might be useful for decisions about adjusting their operations to meet changing conditions. For example, the information that large numbers of people on foot are likely to begin crossing over into an adjoining state could be posted on the RICCS, and then used by transit agencies in the destination jurisdictions to deploy special buses for these pedestrians.

Potential Adjustments Not Yet Implemented

• With the Federal government representing 65% of the work force in the district, it could be useful to get the agencies to work together to work out phased releases of workers to even the flow over a longer time, or, alternatively, to engage in preparations to shelter workers in place. As part of this the Office of Personnel Management (OPM) for federal agencies could develop plans in advance for the message they will give in a similar circumstance and ways to provide guidance for staggering worker release times in order to help distribute the flow of people to the subway more evenly. [Note: Some experts are skeptical that people will comply with a staggered release; others even hold that it is unethical to consider asking people to delay their departure in the face of immediate danger.]
• Consideration needs to be given to the trade-off between slowing the evacuation of tens of thousands of people and the increased security measures that have resulted in a growing number of obstacles (fences and bollards) to evacuation being created by the introduction of security perimeters around Federal buildings.
• Better coordination and quicker decisions about public information messages from across multiple Federal agencies is needed, in order to provide important guidance to pedestrians and other people evacuating.
• In response to a question, the traffic manager didn’t think consideration was being given to bike paths as options for pedestrian evacuation, even though he could see how they might be useful as alternative routes for pedestrians.
• The situation that emerged on 9/11 demonstrated the critical role the Metrorail can have in expediting a necessary evacuation of the District. However, the WMATA emergency planners do not assume that Metrorail will be available. Considerable attention is paid in their emergency planning and exercises to the scenario of Metrorail being the target of a terrorist attack. This scenario results in the thousands of riders on the system being evacuated from the trains and out onto the streets, in need of an alternate mode for continuing their evacuation. When some portion of Metrorail is disabled temporarily, buses are put into service to “bridge” the disabled gap in the overall system. However, this alternative would not have adequate capacity in an urgent and large scale evacuation, so WMATA assumes walking will be an important alternative. WMATA emergency preparedness brochures for the public urge transit users to think about walking routes as one of their alternatives in an emergency.
• Even though considerable effort has been made by many agencies to develop a Regional Emergency Coordination Plan, this Plan in its earlier stages did not acknowledge the likely scenario of an evacuation from the District that includes a major “walk out” by people, as an alternative to driving or using mass transit. Recently, the need to include measures for facilitating pedestrian evacuation has been receiving more attention in the District Department of Transportation (DDOT), and in the regional evacuation planning.
• One of the participants in the development of the Washington Metropolitan Regional Emergency Coordination Plan (RERC) believes plans should be made for keeping the walkers and emergency vehicles separated out onto different routes, to prevent the pedestrians from impeding the response. Modeling that has been done of the load likely to be put on the transportation system during a sudden evacuation of the District suggests that many people will resort to walking out so options for designating some roads for evacuation and others for counter-flow of emergency vehicles need to be planned.
• The group that is planning for regional emergency response is considering the inclusion in the plan of sheltering-in-place as an alternative to the evacuation of an area but are not certain what type of a shelter-in-place option people would be likely to comply with given the importance to most of getting to the same location as other family members.
• The Maryland transit official indicated that the transportation planners are considering plans for designating staging areas for transit along the state-District border, in order to have transit available for large numbers of people leaving the district, especially those on foot. 

3.2.2   Cleveland, Ohio

An internet search on the topic of pedestrian evacuation identified an evacuation plan for the City of Cleveland that addressed both pedestrian and vehicle evacuation. According to representatives of the Emergency Preparedness Office, this plan had been prepared following the events of September 11, 2001. On that day, as information about the hijackings emerged and it appeared that perhaps one of the planes was headed in the general direction of the city, the mayor ordered the evacuation of the few (six to nine) high-rise office buildings in the downtown area. This evacuation prompted a general, self-initiated, and essentially full-scale evacuation of downtown Cleveland, resulting in considerable traffic congestion. Although additional police officers were dispatched to manage the traffic, it took about ninety minutes for the traffic to clear out of the downtown. The Mayor requested the development of an evacuation plan that could be put into effect to ensure a much more rapid and efficient evacuation of the downtown area, should it become necessary in the future.

Cleveland also was affected by the prolonged Northeast Blackout in August 2003. A high level of traffic congestion ensued when the power outage prompted people to leave offices and stores for home. This was viewed and handled as a heavy outbound commute and not as an evacuation, since there was no urgency to clear the downtown. With the light rail out of service because of the loss of power, the transit authority implemented a contingency plan for replacing the light rail service with surface buses to accommodate those commuters. The interviews highlighted the difference in the concept of operation for an evacuation as detailed in the evacuation plan for the downtown compared to a commuter surge or sports event traffic.

Interviews were conducted with representatives of Cleveland’s Emergency Preparedness Office in the Division of Fire, the Division of Police, and the Regional Transit Authority.

 

 

Challenges Presented by the Events on 9/11

• When the mayor ordered the evacuation of the major high rise buildings, the emergency organization found they did not have pre-established contacts for advising building management of this action.
• Downtown Cleveland is mainly commercial, with a low residential population; only about 4% of persons who come into the city during the day do so on mass transit. The officials interviewed described it as a “car town.” Thus, as people made the decision to leave the down town, beginning with those evacuated from the high rise buildings as a precaution, most used a private vehicle. Traffic congestion was considerable and progress slow.
• Cleveland is bordered on the north by Lake Erie, which bars egress to the north, and to the west by the Cuyahoga River; west-bound traffic must be funneled to several bridges.
• During the 9/11 evacuation, an initial order to shut down the rail transit created some confusion among the rail commuters; it was soon determined to have been inappropriate, and rail transportation was restored about an hour later.
• Even though the city agencies are accustomed to the congestion related to events at the Cleveland Browns stadium on the north edge of the downtown, an evacuation of the downtown proper, spontaneous or directed by authorities, was unprecedented and was not directly addressed in the emergency preparedness activities. Thus, various agencies did what seemed right at that moment, but with less than adequate integration of information and coordination.

Responses and Adjustments Made

• On 9/11. Although the recommended evacuation of the high rise office buildings appeared to have been a major impetus for a much larger segment of the downtown worker population to also leave the downtown area, this had not been anticipated. At the same time, nothing threatening had actually happened in Cleveland, so the people leaving did not exhibit any great sense of fear or urgency. Drivers were calm and courteous.
• Post-9/11. A protocol for who decides and who implements any shut-down of any part of the mass transit system has been developed and is part of the emergency plan. The decision will be made at the city’s Emergency Operations Center, based on available information about the situation, and with a representative of the Regional Transit Authority present.
• Post-9/11. The emergency management organization has put in place a procedure for alerting the members of the Business Owners and Managers Association about necessary emergency actions so they can relay the information to the business community.
• Evacuation Planning. Public agencies and private organizations were pulled together to develop an evacuation plan for the city. They started by thinking in terms of the evacuation plan taking into consideration four categories: people in cars; people who use public transit; people who walk; and people who live downtown. The panel included the city’s public safety agencies, the Greater Cleveland Regional Transportation Authority, the Building Owners and Managers Association, and the Northeast Ohio Area-wide Coordinating Agency which is the metropolitan planning organization (MPO).
• Evacuation Planning. The planning discussions led to a general consensus that the public safety agencies need to be concerned with the big picture of most of the population and not the special needs of various individuals. They feel that the evidence from other situations, such as New York City on 9/11, indicates that individuals having problems will be helped by others as the evacuation takes place.
• Evacuation Planning. The emergency planners believe it is necessary to avoid over-planning, that is, avoid trying to think of every little thing and addressing it. The preferred approach is to lay out and share among the players the basic concept of the response, and then deal with the various unanticipated aspects within that framework as they come up.
• Evacuation Planning. The Cleveland planners did not include the option of sheltering in place as an alternative to evacuation in their plan because several hold the opinion that it is not realistic to think that some people will remain in place, even if recommended to do so, if others are evacuating.
• Evacuation Planning. The officials noted that working on the plan brought together different agencies and groups that had previously had not worked together for emergency planning. Working relationships emerged among agency representatives that are expected to facilitate major emergency response in the future. As one of them put it, if agency representatives already are used to working together, “the people involved are more likely to keep the broad objective in mind when they are working together to respond to an emergency.”
• Evacuation Plan Distribution. Citizens have been provided with brochures and web links describing the evacuation plan and providing a map. The downtown is divided into quadrants, each having routes pre-established for moving outward from each part of the business district. If an evacuation is ordered, incoming traffic and cross-town traffic will be prohibited by law enforcement personnel. The plan assumes an evacuation is most likely to be vehicular, although in certain circumstances it may be necessary to leave some areas on foot. For people who evacuate the downtown on foot, the evacuation routes take them to a designated transportation hub within a mile or less, where busses will take them to some further staging area that has necessary amenities. The central concept of the plan is to clear people out of the central area in 30 minutes or less if there is an actual threat, by having them move along a specified direct route from their location to a safe distance. At that point, each evacuee can then make the necessary alterations in route to their home or other preferred destination.

Challenges Presented by the Blackout in 2003

• For the power outage, the plan developed for the evacuation of the downtown area was not used, because the situation did not call for a rapid dispersal of the population for safety reasons. The emergency evacuation plan calls for moving people along specified direct routes to staging areas away from the threatened downtown area; the evacuees then must find their way home from there. The response for the blackout was to facilitate mobility so people could get from downtown to their home or other desired destination before nightfall. This included the use of the normal contingency plans used by the transit authority for any type of disruption in rail services, such as fires or derailment.
• Because traffic signals were out, law enforcement personnel were dispatched to major intersections to keep traffic moving and thereby minimize the traffic congestion in the city. However, commuters who made their way to one of the freeways then encountered major congestion and slow moving traffic.
• In the power outage, the light rail did not run. However, because the outage began in the late afternoon, the transit system staffing was at rush hour capacity and could be held over to operate the extra buses needed to accommodate the rail commuters.
• The transit authority tracks where every rail car is on the system at all times, and thus could move efficiently to send workers to each car to evacuate the passengers from the stalled trains.  They then could dispatch buses to these locations to pick up these passengers and take them to the park-and-ride locations along the rail line.
• As a result of problems encountered with the city’s improvised emergency operations center during the blackout, the city built a better equipped Emergency Operations Center.

3.2.3   The San Francisco Bay Area

Interviews were conducted with responder agency officials who had direct experience with two major disasters in the Bay Area—the collapse of a portion of a double-deck highway in Oakland during the Loma Prieta Earthquake in 1989, and the sudden evacuation prompted by the unpredictable behavior of the Oakland/Berkeley Hills Wildfire (also called the Tunnel Fire) in 1991. An interview also was conducted with a representative of the Metropolitan Transportation Council (MTC), the agency that serves as the Metropolitan Planning Organization (MPO) for the Bay Area. This organization has been active in efforts to develop an integrated response among the transportation agencies and in conducting drills and annual exercises to test these plans.

 

 

Challenges Posed by the Collapse of the Cypress
Elevated Highway in the Earthquake, 1989

• Uninjured drivers whose cars were trapped in the freeway collapse climbed down from the freeway. They most likely were not familiar with their location and needed to get to a place where they could contact relatives or friends for transportation. One respondent referred to the people who are separated from their vehicles while going from one area to another as “freeway refugees,” because they are suddenly in strange territory and need assistance to find certain facilities.
• There were too few neighborhood people or other onlookers at the response scene to have constituted an impediment to emergency vehicles and activities. However, the media quickly converged at the freeway collapse scene with large vehicles and other equipment, blocking the access of emergency responders. [Note: Because the event occurred at the beginning of a World Series game in San Francisco that evening, there was an unusually large contingent of national media in the area.]

Challenges Posed by the Earthquake Damage to theBay Bridge, 1989

• The earthquake caused a piece of the bridge at mid-span to drop below the road surface. Cars that were on the bridge but not yet past mid-span could not continue across the bridge. One respondent reported that some people abandoned their cars and walked back across the bridge the direction they had come. Another respondent said that eventually a procedure was established for having drivers turn their cars around and drive off the bridge, using the empty lane going the other direction. [Note: It is easy to imagine bridge damage, especially to the bridge approaches, that would necessitate all drivers having to walk off the bridge, leaving their cars stranded. Depending on the situation, emergency responders might need to arrange for transportation for such bridge evacuees.]
• Many workers in downtown San Francisco live in the East Bay area (e.g., Berkeley, Oakland). The type of earthquake damage that occurred to the Bay Bridge eliminated the use of that bridge as either a vehicle or pedestrian route. In this instance, the underground BART system between San Francisco and the East Bay was determined to be operable after initial assessment, so it could be used by commuters going between San Francisco and the East Bay communities. Other bridges that provide the shortest routes between San Francisco and other Bay Area communities were not damaged.

Challenges Posed by the Oakland/Berkeley Hills Fire, 1991

• The residential areas in the Oakland and Berkeley Hills had started out as areas of summer homes, many with small winding driveways to individual houses. This maze of driveways and roads was extended and paved as more and more houses were built in the area. In many instances, groups of houses had only one ingress/egress option. Little attention was given to planning for the evacuation of a large proportion of the residents in the event of a wildland fire in the area. Winding roads that already created challenges for fire fighting equipment with long wheelbases were inadequate for any large surge of two-way traffic. When the fast-moving and unpredictable firestorm enveloped large portions of the area in 1991, the residents had little preparation for evacuation.
• More than 3,300 homes were destroyed by the fire. It is estimated that approximately 10,000 people evacuated their homes during the Oakland Hills fire. Many of the 25 people who died were trying to evacuate. A responder who had been with the Oakland Fire Department at the time described how quickly the situation changed, causing even the firefighters to evacuate the position they had established for fighting the fire. As the fire began spreading in several directions, their role changed from that of firefighter to that of advising residents to evacuate immediately. Thus, a mix of fire department vehicles, residents in cars, and residents on foot all moved together to get away from the hillside neighborhood that was threatened.
• The police chief noted that he had discovered, through the personal experience of his teenage son, that the police had gone through the area in which his home was located to prompt the residents to evacuate because the fire had suddenly started spreading in that direction. The son, home alone and too young to drive, did not have any means other than on foot, to evacuate. The chief has since then taught police officers that when they are advising people to evacuate an area, they need to consider what might need to be done for people who do not have the means to do so. They cannot assume everyone has access to a vehicle.

Adjustments Implemented

• Cypress Freeway collapse:  The police called the Transit dispatch center to request buses to take the evacuees from the collapsed freeway to a nearby military facility where they could rest and make telephone calls. This is a standard practice implemented between the police and the transit company whenever buses are suddenly needed at a specific point to move groups of people. [Note: Given the widespread use of cell phones nowadays, people likely will use these to arrange private transportation. However, in order to keep traffic congestion away from the response activities, it may be advisable to establish a staging area where people who have no access to a vehicle can be picked up.]
• Crowd Management—General: The Police Chief noted that he believes people are fairly self-reliant in emergencies and can decide on a course of action without explicit instruction from authorities. Also, the police generally have other things to do besides focus on people on foot in the area. He believes that it is when authorities try to be over-controlling with crowds that it causes trouble. The emergency manager of another Bay Area community echoed these sentiments. In thinking about a scenario of an explosion or other situation requiring the evacuation of 70,000 from the local stadium, he expects that people who do not have vehicles and live elsewhere will likely begin moving toward the BART line; they will not need, nor would they pay much attention to, efforts of officials to specify a route for them to take.
• Mass Transit Characteristics:  One respondent noted that in many cities, mass transit is designed to get people from the suburbs to the city center and back. This limits its use for taking people shorter distances in between. Other means of transit will have to be used for that purpose. Planning and exercises for coordinating Bay Area multi-modal transportation providers in emergency and disaster conditions have been carried out for many years, in conjunction with the Department of Transportation and the emergency management officials. Emergency planning calls for the Metropolitan Transportation Council to compile and disseminate information on mobility options available throughout the Bay Area. However, the observation was made that transportation workers will be concerned with managing their assets, so it will most likely be law enforcement personnel who would manage a pedestrian evacuation, should that emerge, for example, if the subway had to be evacuated.
• Short-term Evacuee Needs: One respondent observed that, in his community, specific locations have been pre-designated as short-term shelters that can be used in the event that people have to be evacuated from an area because of a hazardous materials spill, toxic release, or fire danger. He thinks these shelters could also be opened when people are evacuating on foot away from the freeway because of freeway closure due to damage such as overpasses collapsing in an earthquake.

Adjustments Planned for Future Implementatio

• In the course of a hypothetical discussion of alternative modes for evacuation from San Francisco, the San Francisco respondents note that the several ferry routes between San Francisco and other points offer an important alternative mode for pedestrian evacuation, should one or more of the bridges be closed. The ferry system is characterized by a great deal of flexibility. There are docks in many places for loading and unloading, and most ferries also can be accessed from simple docks quickly set up for that purpose. The ferry system can be called into service for evacuation.
• Because of the similarity of the Berkeley Hills residential area to the area affected in the Oakland Hills Fire, the emergency manager of Berkeley wants to improve existing pathways between houses and link portions of circuitous roadways in order to provide alternative routes for pedestrian evacuation. These walkway rights-of-way are considered the property of the city. The narrow roadways in the residential area can become clogged with traffic or a roadway might suddenly become cut off by the fire, making the use of a vehicle for quick evacuation problematic. He has been able to obtain a Federal Emergency Management Agency/Department of Homeland Security (FEMA/DHS) Fire Prevention and Safety Grant to use for this purpose, illustrating the need for creative thinking about the dual use of various grant programs related to emergency response and response safety. The project activities funded by the grant will be carried out in conjunction with the activities of a volunteer group working on enhancing the walkways for recreational uses.

3.2.4   New York City Region

The events of September 11, 2001, when terrorist-controlled airliners were flown into the World Trade Center towers in a major business district of a large metropolitan area, present perhaps the most basic scenario for conditions likely to result in a large-scale pedestrian evacuation. Although city officials declined to provide interviews for the research, a general description can be pieced together with information from news accounts, news photos, and various research reports. This decidedly threatening event occurred with no warning. It also occurred in two stages—the impact of the planes on each of the towers, and then the collapse of each tower. People who evacuated from each tower after it was hit apparently arrived at the street in a fairly steady stream, most not even aware of exactly what had happened to the building. Some needed medical attention; others apparently dispersed in various directions on foot or initially lingered in the general vicinity as observers. Initially, no other buildings were evacuated.

On a typical work day a large proportion of the worker population present in lower Manhattan that morning will have arrived there by mass transit rather than by private vehicle. Soon after the initial attack, the subway was shut down and passengers were evacuated from it. All traffic inbound to Manhattan was prohibited, with the exception of emergency response vehicles. The compactness of the Borough of Manhattan, joined to the more residential boroughs by bridges, tunnels, and ferries, made feasible a decision to head out on foot for other parts of the city. Buses had continued to run and by afternoon subway operations were re-established in areas other than Lower Manhattan. Given that Manhattan is an island, all routes for leaving the island represented major choke points for both vehicle and pedestrian traffic.

Street access to the immediate area around the WTC had quickly been closed off by law enforcement for all but emergency response vehicles. One report stated that the evacuees from the towers and other observers who lingered around the scene early on stayed out of the way of the emergency responders. The initial exodus from the area of the two WTC towers was characterized by people emerging from the buildings in a continuous stream after an arduous building evacuation experience, walking away from the area, and then running for safety as the first tower unexpectedly collapsed. Within the cordoned-off area the pedestrians spread out across the sidewalks and street in an apparent effort to accommodate individual walking rates. The pressure wave from the collapse knocked pedestrians to the ground, and the dust cloud made it too dark to see for some period of time. Many crouched along buildings or cars, or were pulled inside buildings by other people. After the dust clouds cleared people continued their efforts to get home.

Once these pedestrians proceeded some distance from the impact site, they settled into streams of pedestrians heading toward their destinations of choice. For the most part, these were people who had left their place of work and were headed for home. This situation is essentially the reverse of what happens during hurricane evacuations, where entire families are leaving their homes, headed for some safer place to stay temporarily until the storm is over. Accounts indicate that some degree of official management may have been used to contain these streams so that usually at least one lane was available for emergency vehicles, both on the surface streets and the bridges. As pedestrians reached areas farther north of Canal Street, they encountered the normal mix of activities of people in areas that were not directly affected by the damage. Evacuees, themselves covered with dust from the collapse of the towers, remarked on how strange it was to find other parts of the city going about normal activities.

At about 11:00 a.m., after the collapse of both towers, the Mayor urged that everyone who was still south of Canal Street should evacuate that area. Evacuation in this instance was viewed as a general protective action, given the uncertainty about conditions associated with the destroyed towers. This secondary evacuation involved a heavy stream of private cars; the traffic flowed smoothly and as directed by law enforcement, who had received prior notice that it was to occur.

A no-notice emergency urban evacuation on the scale of that associated with the attack on the WTC towers was an unprecedented event in the United States. Interest in all aspects of the event was intense on the part of the media, criminal investigators, forensic engineers, building evacuation experts, behavioral researchers, and any number of other types of experts. In August 2003, the Northeast Blackout in August prompted another evacuation from the office buildings in Manhattan to other parts of the city by people on foot. Many also had to be evacuated from the stalled subway system. Increased bus and ferry service was fairly quickly implemented to accommodate the demand. Many people chose to walk, and bridge routes were provided for those headed to the east. Lessons from 9/11 led to the implementation of measures to separate pedestrians and vehicles by assigning them to different bridges and tunnels.

 

 

Challenge to the Transit System

One interview was conducted with a representative of a transit agency. This interview served to provide further evidence about pedestrian evacuation as a multi-modal alternative, unless the distance people need to move to be safe is short and the time they need to remain there is brief. Urban transit systems, especially rail systems, can provide alternative evacuation routes to avoid traffic gridlock and endangering pedestrians. However, urban transit systems must address certain challenges.

• The NYC subway system has a clear policy and procedures for evacuating individual trains or all trains in the system. Whenever a train is stopped for whatever reason, including such things as power failures, chemical releases, or fires, anywhere other than a station, the Fire Department or transit operational personnel assist passengers to walk out to the nearest station because of the danger represented by the “hot” third rail. The focus is on achieving a safe and orderly evacuation of the passengers and minimizing damage to the system.
• On 9/11. The transit authority initially focused on security because of the uncertainty about the scope of the attack. After the first tower was hit, the transit authority focused on evacuating the entire subway system in order to secure it from a potential attack. There was no specific plan in place at the time to provide information or alternate transit for the evacuated passengers or persons arriving at stations in hopes of taking the subway.
• Immediately following the impact of the airliners, buses continued to run except in the vicinity of the WTC.  However, there were not enough buses to handle the people evacuated from the subway. This added to the volume of people on the streets who were evacuating from the area of the WTC complex.

Responses and Adjustments Made

• On 9/11.  The Subway Control Center was able to use its communications system and procedures to evacuate the subway system. Once the extent of the attack became clear, the Control Center was able to communicate with transit personnel to put the subway system back into operation, bypassing the area of the WTC. Buses had continued to operate, with some alteration in routes around the WTC.
• During the 2003 Blackout, the backup power for the Subway Control Center and communication functions was operable, so trains could be located and the evacuation of the system carried out smoothly. Subway service could not be re-established until power was again available.
• One reason given for the good response of the transit system was the fact that the transit agency’s emergency procedures include a comprehensive and up-to-date list of contacts in various agencies. This helped establish direct access to people in a position to provide reliable information and to commit resources during an emergency.
• The transit agency official observed that a greater number of people walked sizeable distances to their preferred destination during the 2003 blackout than during the evacuation from the WTC attack on 9/11, presumably because fewer options for transportation were available during total blackout conditions.
• Post-9/11. The NYC Transit agency has adopted a policy of preparing all operations and administrative personnel to be involved in subway evacuations in order to offer better assistance to passengers. While the operations personnel work with the system, office personnel can be deployed to provide direction and assistance to the passengers emerging from the stations. This can be particularly useful for those passengers who find themselves in an unfamiliar part of the city due to the sudden evacuation of the system.

As was indicated in the earlier literature review, several observations about adjustments made to accommodate the unusually large numbers of pedestrians were gathered during the interviews for the FHWA-sponsored case study of the transportation challenges and responses in conjunction with the WTC attack on 9/11. A major concern of the transportation officials was that the congestion on the bridges associated with having large numbers of pedestrians on the bridges as well as cars might hinder incoming emergency vehicles. Recommendations were developed for ways to keep pedestrians and vehicles separated, such as assigning certain roads and bridges only to vehicles and others to pedestrians, or gathering pedestrians in staging areas where they would be provided bus transportation for crossing the bridges.

Anecdotal Information from First Hand Evacuation Observations

The evacuation situation on September 11, 2001 can be pieced together relatively well from news stories and photographs. Personal web logs on the internet provide another interesting, albeit anecdotal, source of information on the situation encountered by people walking out of the WTC area, their protective actions, and their strategies for getting home. Some insights into the options people had and used can be gleaned from these first hand observations.

• A person leaving the first tower that was struck was directed by law enforcement to ‘walk north to the train station.’
• When the tower collapsed, the force of the cloud of dust and debris pushed people off their feet and to the ground.
• The dust was so dark and thick, people thought they might have been blinded.
• When people who were walking away from the scene were caught by the dust-laden pressure wave from the building collapse, they took refuge in doorways and shops or were pulled into these places by others inside.
• A person who had reached the Wall Street area was directed by a police officer to walk toward South Street Seaport and go over the Brooklyn Bridge, at which point she would be transported across the river to Liberty State Park. When she arrived there, she was bused to Penn Station in Newark. Hazmat teams decontaminating people coming from New York confiscated her clothes, shoes, and purse, showered her, checked her vital signs and provided her with hospital scrubs to go home in. Relatives she had been keeping contact with by cell phone picked her up there.
• After the first plane hit, the situation seemed surreal to one observer, who was in the lobby of one of the towers at this time. This person describes himself as going into a sort of survival mode, and going out into the street where he just kept walking, telling himself that he had to get out of the city. He walked south from the WTC, then east to Broadway, and then north. At the Christopher Street Port Authority (PATH) station he got a train to Hoboken where his car was parked, and from there he drove home.

3.2.5   Seattle, Washington

In 2003, Seattle was one of the locations for the TOPOFF-2 exercise. The scenario for Seattle included a dirty bomb explosion in the downtown area. An interview was conducted with a representative of Seattle Emergency Services to determine what sort of protective action was included in the exercise scenario for an event that involved a radioactive plume associated with the destructive explosion of a “dirty bomb.”

Challenges Described

• The exercise scenario involved the explosion of a dirty bomb at the south end of downtown Seattle, less than a mile from Interstate-5. The emergency response entailed making an assessment of the situation before deciding on appropriate protective actions for the public. For the TOPOFF-2 exercise, Interstate-5 through downtown Seattle was closed, presumably as a precaution to avoid exposure of travelers. The downtown Washington State Ferry Terminal was not accessible because it was located in the projected plume path from where the bomb exploded.
• In response to a question about the experience of congestion around a response scene that hindered access by emergency responders, the respondent recalled an incident where a bus went off the edge of a bridge when a passenger shot the driver. The Fire Department and EMS responders were unable to get in good position to respond because of the police cars that had converged on the scene. He recalled that the same thing had occurred at the collapse of the Skybridge at a Hyatt Regency Hotel. The police had converged at the scene, and left their cars parked and locked in the circular drive in front of the hotel when they went into the hotel, thereby hindering access by other emergency responders.

Adjustments Assumed in the Simulation

• For the dirty bomb scenario, the emergency managers ordered (in simulation) the public in the metropolitan area to shelter in place, in appropriate areas of their offices and residences. The public was expected to remain sheltered in place unless the incident commander ordered people to be moved from an area where the radiation was expected to intensify as the plume from the explosion moved across the city, depending on the wind currents. The logistics for moving people, mainly by bus, would be implemented from the Emergency Operations Center. The public was ordered not to use vehicles to evacuate, but rather to wait to be evacuated by bus to an area not likely to be downwind from the bomb site. In such a situation, some might leave the area on foot instead. [The public was not actually participating.] The response objective was to minimize population exposure to the radioactive plume by moving only those persons who might otherwise be in its direct path. The hazard zone was being determined on an ongoing basis by plume modeling experts.
• The emergency services department holds that it is not appropriate to predetermine specific evacuation routes, because the location and nature of the event are seldom known in advance. Emergency management officials, however, will be aware of the most reasonable alternative routes and transportation modes as well as which agencies will help implement an evacuation using these routes. The Ferry system is one of the options to be considered if movement to the west is appropriate.

Adjustments Planned for Future Implementation

• A warning system that the City is putting in place along the Seattle waterfront was described during the interview. It will be used to warn people in the area if sensors have detected a chemical or biological hazard, or, for example, the threat of a tsunami. For many months of the year, this waterfront area is crowded with tourists and other pedestrians making use of the waterfront shops, restaurants, and tour boats. This warning system will have to be accompanied with signage and other public education approaches, so people will know what to do if the warning system is activated. [Note: The waterfront area lies between the water and a steep hill or high rise housing, which limits the options for easy evacuation.]

3.2.6   Observations from a Regional Planner

 

 

One respondent, affiliated with a regional planning organization, requested that remarks made during the interview be anonymous. The planner talked broadly about evacuation in general. He made the important point that plans for evacuations where evacuees will be crossing from one jurisdiction to another will have to address many important issues. These need to be worked out in advance. His observations were borne out in at least one incident from the Hurricane Katrina evacuations in 2005 when a group of evacuees were stopped from crossing the boundary into a particular jurisdiction by law enforcement authorities, apparently at the request of elected officials. The planner’s observations included the following points:

• Planning for an evacuation of persons that involves moving from one jurisdiction into another can be a sensitive issue to the public. Evacuation planning policy needs to be dealt with by statutorily responsible agencies and not planning bodies. One issue has to do with an evacuation that may result in one jurisdiction becoming a host for the evacuees of another jurisdiction in the event of a long term evacuation.
• Specifying the destination of an evacuation is not a traffic engineering issue; it is a political issue.
• “Evacuation” can be a confusing term because it is a layperson’s term. When considering planning assumptions in reference to evacuation, it is necessary to clarify whether the objective is traffic management in emergencies or protective action for the public.
• The planner suspects that what prompts the evacuation makes a difference, e.g., whether it is a response to terrorist or military action or to some more common event, like a hurricane or flood. The specific threat or hazard may determine the use of particular options.