1. INTRODUCTION
Due to its specific spaces and complexes in usages and management, the underground MRT station does not accommodate to regular fire-safety regulations in Taiwan. TRTS adopts NFPA (National Fire Protection Association) 130 as the design criteria, evacuating all the passengers in a platform floor within 4mins and those in a station within 6mins, for evacuation safety. The Prescriptive Code, NFPA130, proposes fixed parameters and a simplified available safety egress time, which cannot reasonably describe the evacuation process and adjust the parameters for the real fire scenarios. Thus, to hold a full-scale maneuver of evacuation, due to a huge cost trial, does not match the principle of economic efficiency. Therefore, to adopt the computer simulation approach, an appropriate alternative, can raise the reasonability and feasibility in the evacuation research. Simultaneously, its results can be used to check the safety of specific architectural spaces and to persuade officers and clients for reaching the goals of saving cost and encouraging creative designs.
In order to construct a more complete evaluation structure for evaluating the egress safety of an underground station, this paper took HDSS as an example and utilized the buildingEXODUS software as an analytic tool for evacuation safety analysis. Those results can be the strategies proposed to TRTS in emergency response management.
2. VERIFICATION OF BUILDINGEXODUS COMPUTER SIMULATION 2.1 Verification and Application of EXODUS
buildingEXODUS is designed for applications in the built environment including supermarkets, hospitals, rail stations, airport terminals, high rise buildings, and so on.
It can be used to demonstrate compliance with building codes, to evaluate the evacuation capabilities of all types of structures and to investigate population movement efficiencies within structures. It also takes into consideration for people-people, people-fire and people-structure interactions. This model tracks the paths of individuals as they make their way out of the enclosure, or are overcome by fire hazards such as heat, smoke and toxic gases. Thus, the behavior and movement of each individual is determined by a set of heuristics or rules. For additional flexibility these rules have been categorized into five interacting sub-models, OCCUPANT, MOVEMENT, BEHAVIOR, TOXICITY and HAZARD [1]. Above all, the results can be the important references for simulating evacuation time when fire occurs. There are several verification and application as following:
I. In Paulsen’s research [2], the experimental evacuation time for the 1.5m exit is quoted as 30 seconds. This is apparently for a single trial. Five repeat runs of buildEXODUS produced evacuation time of 28.8, 29.1, 29.6, 31.1 and 31.4 seconds with a mean time of 30 second, resulting in a difference of 0%.
II. In Taiwan, the Graduate School of Fire Science, Central Police University applied buildingEXODUS to simulate the evacuation in underground MRT stations Ching-Mei [3] and Gung-Guan [4].
2.2 Functional Analysis of EXODUS
buildingEXODUS provides some measures to examine the outputs, which can be used to prove its reality. Two measures for evacuation efficiency are given as following [5]:
I.Optimal Performance Statistic (OPS)
OPS is measured from 0.0 to 1.0, a lower value indicating a more efficient evacuation. OPS is defined by the equation:
1
An OPS value of 0 signifies a well-balanced evacuation, i.e. all exits completing at the same time, while an OPS value of 1.0 indicates a poor evacuation with at least one exit not attracting any occupants. Though the OPS is a useful measure of optimal performance, it only provides performance information at a time point, i.e.
at the end of the evacuation. As such, it provides no insight into how the evacuation has progressed. To overcome this deficiency another measure has been developed to complement OPS, which is known as the Mean Non-flow Statistic or MNS.
II. Mean Non-flow Statistics (MNS)
The MNS provides a measure of the amount of time that each exit is not in use during an evacuation, i.e. no person negotiating the exit. The MNS can be determined from every exit and used as a measure of overall evacuation efficiency.
The formulation of the MNS for an exit is as follows:
= i
TNFi = total time (seconds) exit “i” is in non-flow conditions, i.e. no person negotiating the exit
TFTi = Total Flow Time of exit “i” (seconds) given by EETi – EFTi
EETi = time last person leaves exit “i”
EFTi = time first person leaves exit “i”
The formulation of MNS for a building is as follows:
1
MNS is the mean non-flow time for the building, and n is the number of exits during the evacuation. The MNS must be less than 100%. An evacuation in which MNS = 0 indicates that all the exits achieved a continuous evacuation flow without a breakdown. Generally, we consider MNS values less than or equal to 10.0% as being optimal. Above all, an ideal evacuation would have OPS = 0.0 and MNS = 0.0.
3. CONSTRUCTING THE PARAMETERS OF HDSS
Through buildingEXODUS, this paper simulated the evacuation in the platform floor of HDSS. The parameters are described as follows:
3.1 Architectural Characteristics of Platform Floor of HDSS
The length of platform floor is 141m and the width is 11.9m. In that area, two 1.8m width emergency exits at two ends of the floor connect ground floor and platform by stairwells. Three escalators, of which the width is 1.2m, one stair, of which the width is 1.8m, and one elevator (not for the purpose of evacuation), connect the lobby floor and the platform.
To draw the plans of HDSS on AUTOCAD DXF format is imported to the Geometry Mode of buildingEXODUS. Simultaneously, the Tools menu convert the DXF file to two-dimensional grid of nodes, such as Node, Arc, Seat Nodes, Internal Compartments, Obstacles and External Exits, and so on. Before constructing
staircases, the user has to key in the related values to define the size of the staircases.
In staircase dialogue box, the values must be adjusted appropriately, such as the Width, Drop, Lane Width (default value- 0.76m) and advances data – Hand Rails, Riser High and Riser Depth. In this project, the Land Width is adjusted to TRTS default value – 0.56m. The plan of Hsing-Dein Subway Station is illustrated on Fig. 1, in which Exit5
~ Exit8 mean one stair and three escalators.
Fig. 1. The Plan of Hsing-Dein Subway Station (HDDSS)
The number of the evacuation occupants was based on those predicted by Department of Transit System, Taipei City Government during the rush hours [6].
According to the scenario on the TRTS design and planning handbook 12th Edition [7], the total number of occupants – 2731 was estimated in this paper. The occupant number was obtained from a train with 1900 passengers and 1.5 times of the number of occupants entering the station. Through Population Panel Editor (PPE), 2731 occupants were randomly distributed on the platform floor. Their response times were among 0 – 30 seconds. The individual attributes of genders and ages were adopted from the questionnaires for TRTS passengers, research results of The Graduate School of Traffic Management, Central Police University [8].
3.3 Defining A Scenario
Because there are more than two stairs or escalators on the platform floor, Attractor and Discharge nodes are used to control the Exit Attractiveness for occupants. For the HDSS simulation, the Patience Attribute Value, 1~5 seconds, were adopted; the Unit Flow Rate default value – 1.33 occupant/m/sec was used. Exit
Status (default – Open) and Exit Activity (default – Yes)
were also adopted. When set to OPEN, any occupant queuing may pass through.
If the active attribute is YES, occupants will be attracted to the exit and queue, even if it is closed. In additional, all Exit Potential and Exit Attractiveness (default value – 100) were used to simulate the evacuation.