The World Trade Center Evacuation: INTRODUCTION
The evacuation of the WTC complex represents one of the largest full-scale evacuations of people in modern times. As such, it is of fundamental importance to the understanding of the complex interaction between structure, procedures, environment, and human behavior, and how these factors interact to determine evacuation performance. The WTC evacuation provides an unrepeatable opportunity to probe into and understand the very nature of evacuation dynamics, and with this improved understanding, contribute to the design of safer, more evacuation-efficient, yet highly functional, high-rise buildings.

Following 9/11, the Fire Safety Engineering Group (FSEG) of the University of Greenwich embarked on a series of studies centered on the evacuation of the WTC. These include a study of printed accounts from survivors of the WTC evacuation;^1, 2, 3, 4 numerical simulation studies of the evacuation of WTC Tower 1; ⁵ project HEED, a study to collect and analyze data from face-to-face interviews with survivors of the WTC evacuation; ⁶ and a proposal to investigate the possible use of elevators and sky bridges for the evacuation of high-rise buildings. ⁷ This paper reports on a selection of the published findings of the first study.

Background to Project
The survivors of the WTC disaster hold a tremendous amount of information concerning their experiences of the conditions within the structures and the evolving evacuation scenario. Ideally, this information should be gathered from face-to-face interviews conducted as part of a scientific study. An alternative, and less desirable, approach relies on firsthand accounts that have appeared in the mass media. These are usually the result of press interviews conducted by journalists or personal accounts produced by survivors in Web sites or books. The difficulties with relying on the media is that specific groups are not targeted, interviewees self-select, journalists tend to only report the more sensational parts of people's stories, questions posed are inconsistent, questions posed by journalists are not necessarily known, and an inability to ask specific questions. Furthermore, individual survivor accounts of traumatic events may be influenced by the nature of the event, leading to an inaccurate perception of the facts by the survivor. One way of addressing this issue is through corroboration of facts from analysis of multiple survivor accounts. In effect, the accounts that appear in the mass media provide an uncontrolled snapshot view of the incident, and what is not known from these accounts is as important as what is known. Nevertheless, the data contained in such accounts can prove useful in providing insight into behavior during such incidents. Furthermore, these accounts were recorded very close to the event, some accounts being made a matter of days after the incident. Studies involving live interviews with survivors usually view the incident after the passage of a considerable amount of time (in the case of the WTC, years) and so may be tainted by information gleaned from other accounts that have appeared in the public domain, memory lapses, or selective amnesia. Therefore, the data collected from published accounts, while not ideal, potentially contain valuable information.

Following the WTC disaster, the Building Disaster Assessment Group (BDAG) of the UK Office of the Deputy Prime Minister funded FSEG to gather, collate, categorize, electronically store, and finally analyze data concerning human behavior during the WTC evacuation. Reports were gathered from the literature published in the public domain. Material sources ranged from survivor accounts printed in newspapers and newspaper Web sites, interviews in the electronic media, survivor Web sites, and books. Over 250 separate accounts were gathered that described occupant behavior. Information appearing in print newspapers represents 70 percent of the accounts, while information from Web sites (news and personal) represents 16 percent of the accounts. The remainder of the accounts have appeared in books, journals, and the electronic media. These accounts provided information concerning 120 people from WTC1 (north tower), 119 from WTC2 (south tower), and 21 of unknown origin.

Number of Occupants in the WTC Towers
There are various estimates for the number of people in the building and the number of fatalities. Denis Couchon of USA Today estimates that there were between 10,000 and 14,000 people in the buildings at the time of the impact,⁸ while NIST, in their interim study, estimates that there were 17,400 +/- 1,200 people in the buildings.⁹ Couchon estimates that 1,432 building occupants perished in WTC1 and 599 in WTC2,10 while NIST estimates that 1,560 and 599 building occupants in WTC1 and WTC2, respectively, perished⁹ .

The Database of Human Experience
The collected accounts were entered into a specially developed database. Each individual experience described within the account was stored and assigned specific behavioral references. This is similar to traditional qualitative analysis tools that allow users to categorize portions of textual accounts during the input process. The rationale for the database was that all information was centered on an experience. Each experience was assigned a main category and a subcategory that described the nature of the experience. A distinguishing feature of the database is that it is not only able to store experiences but also the location of the experience and a time reference for the experience.

The database contains reference to a total of 3,291 experiences from 260 people (1,869 accounts from WTC1, 1,411 from WTC2, and 11 from unknown locations). Gender information was available for 240 people, 164 of which were male and 76 female. The quality of these data varied enormously. While some accounts were several pages long, others were only a couple of paragraphs in length. Of more importance, some accounts provide important detailed information such as a detailed description of events, locations at which events took place, and reference to key time markers. The reports mainly came from occupants that began their evacuation in the upper floors of either tower. Within the database, 73 ( 61 %) and 91 ( 76 %) of the occupants from WTC1 and WTC2, respectively, were initially located on or above the 78th sky lobby. It is likely that this bias originates from the media's natural desire to focus on accounts that described the most extreme conditions during the disaster.