Evacuation Modeling: An Overview Presentation

1
Evacuation ModelingAn Overview Presentation

• Rani Muhdi

2
Outline

• Introduction
• The development of evacuation models
• Evacuation models
• Simulation examples and Life Safety Code
• Research opportunities

3
Introduction

• Crowd Dynamics
• Pedestrians
• Traffic control
• Cities evacuation
• Natural disasters
• Egress
• Fire evacuation
• Workplace violence

4
Introduction

• How important to us?

5
Introduction

• Buenos Aires Club
• Some 4,000 fans at a night concert fought to
  reach the exits as burning debris fell on them.
• The concert crowd was nearly three times the
  venue's capacity.
• Someone from the crowd tossed a flare and there
  were immediately flames.

6
Buenos Aires Club

• Once the fire erupted, everyone ran for the
  doors, but there was only one, very narrow one
  open at the exit closest to us. Another wider
  door next to it was locked.
• Emergency exits at a nightclub packed with
  teenagers were padlocked or wired shut when a
  flare ignited the foam ceiling, sparking a blaze
  that killed 175 people and injured more than 700,
  at least 102 were in critical condition.

7
Stampede at Hindu Procession

• About 200,000 Hindu worshippers gathered Tuesday
  afternoon along the narrow road leading up to the
  Mandhara Devi shrine.
• A fire caused by a short circuit in a makeshift
  shop near the temple created panic among the
  pilgrims.
• The stampede has killed at least 300 people in
  western India, according to officials. Most of
  those who died were women and small children.

8
Example I Herding
9
Example II Pedestrian Arch Formation
10
Example III Disruptive Interference
11
Example IV Wall seeking
12
The development of evacuation models

• Sources of data
• Videotaped observations
• How long it takes to evacuate
• Record movement
• Characteristics of individuals
• Lab experiments
• Restrictions and ethical issues
• NIOSH
• Reality!

13
The development of evacuation models

• Post-incident survey and interviews
• Survivors of actual fire
• Recollections and descriptions are subjective
• Elapsed times are not recorded objectively
• Details can be lost overtime
• Evacuation models
• Key tools for the evaluation of engineered
  designs.
• Features and capabilities
• Modeling methods and structures
• Limitations

14
Evacuation models

• Features and capabilities
• Purpose
• Availability for public use
• Occupant movement
• Occupant behavior
• Use of fire data
• Output
• Use of CAD drawings
• Special features

15
Evacuation models

• Modeling methods and structures
• Flow-based modeling
• Cellular automata
• Agent-based modeling

16
Flow-based modeling
17
Flow-based modeling
18
Cellular automata

• Discrete space into cells
• Individuals on a grid
• Series of time-steps
• Tracking the movement of individuals
• Lack of social behavior

19
Cellular automata
20
Agent-based modeling

• Attributes assigned to each individual
• Rational agent to assess optimal escape route
• Social force model
• Exodus-FSEG

21
The social force model

• Newtons Second Law
• Repulsive interaction force
• Interactions with the walls

22
Simulation Models

• No Panic
• Normal desired velocity vo 1 m/s.
• Regular and efficient outflow related to good
  coordination in leaving.
• 200 people leaving without panic
• Panic
• High desired velocity vo 5 m/s.
• Irregular and inefficient outflow due to arching
  and clogging at the bottleneck (door).
• 1000 people leaving with panic

23
Simulation Models

• Dwindling Available Space due to a Fire Front
• Pedestrians have a normal desired velocity v0 1
  m/s, but flee the fire which propagates linearly
  from left to right with velocity V (red).
• The fire front is modeled analogously to the
  psychological repulsion effect by walls, but its
  strength A and interaction range B are assumed 10
  times stronger.
• When the fire comes closer, people press stronger
  towards the exit.
• As a consequence, the outflow becomes
  disorientated and irregular.
• 200 people try to escape a fire

24
Simulation Models

• INDIVIDUALISM vs. HERDING
• People are searching for an exit they can't see
  because of smoke (gray), the desired velocity of
  individuals is v0 5 m/s.
• Pedestrians recognize the door and the walls from
  a distance of 2m, while the range of
  pedestrian-pedestrian interactions is assumed to
  be 5m.
• For pure individualism (unrealistic), people find
  an exit only by chance.
• For strong herding, people follow the mass which
  may move into the wrong direction.
• Most efficient is a mixture of individualism and
  herding, for which small groups are formed
  Successful strategies that individuals found by
  chance are imitated by a reasonable number of
  people.
• 1 Individualism (p0.01)
• No interaction except for collision avoidance.
• 2 Mixture of individualism and herding (p0.4)
• Small groups follow "leaders" who found the exit
  by chance.
• 3 Herding (p0.8)
• Everyone follows the mass, even if it makes fatal
  mistakes.

25
Life Safety Code
26
(No Transcript)
27
Life Safety Code
28
Life Safety Code
29
Research opportunities

• Movement capabilities of a wide cross-section of
  society
• Evacuation of disabled people
• Delay times before beginning evacuation
• Behaviors
• who decides to stay and who decides to go?
• what is the basis for exit choice?
• how can we predict stopping and turning back
  behaviors?
• who queues and who doesn't?
• Perception of risk
• what factors impact perception of risk?
• how does risk perception impact judgment?
• Interaction between people
• how do the presence of social groups impact
  evacuation delays and movement?

30
Research Opportunities

• Number and location of exits
• External disasters and facility layout design
• Layout design of machines, transportation and
  material handling, raw material, and work in
  process have direct effect on human behavior in
  evacuation
• Workplace violence and layout design
• Reliability of exit systems in a facility
• Average weight up 25 since 1960s
• Implementation of the Life Safety Code