A methodology for an aggregate analysis of evacuation of buildings, di Massimo Di Gangi, Corrado...

Massimo Di GangiMassimo Di GangiUniversità degli Studi di MessinaUniversità degli Studi di [email protected]@unime.it

A METHODOLOGY FOR AN A METHODOLOGY FOR AN AGGREGATE ANALYSIS OF AGGREGATE ANALYSIS OF EVACUATION OF EVACUATION OF BUILDINGSBUILDINGS

Corrado RindoneCorrado RindoneUniversità Mediterranea di Reggio Università Mediterranea di Reggio [email protected]@unirc.it

M. Di Gangi, C. Rindone - A Methodology For An Aggregate Analysis Of Evacuation Of BuildingsA Methodology For An Aggregate Analysis Of Evacuation Of Buildings

STRUCTURE OF THE PRESENTATIONSTRUCTURE OF THE PRESENTATION

I. INTRODUCTION

II. PROPOSED APPROACH

III. EXPERIMENTATION

IV. ANALYSIS OF RESULTSSimulation resultsComparison between simulations and on site experimentation dataComments and perspectives



I. INTRODUCTIONClassification of pedestrian evacuation tools

II. PROPOSED APPROACHES


IV. ANALYSIS OF RESULTSSimulation results

Comparison between simulations and on site experimentation data

Comments and perspectives


• Gwynne, S. Galea, E.R., Lawrence, P.J., Owen, M. & Filippidis, L. A review of the Methodologies used in the Computer Simulation of Evacuation from the Built Environment. Building and Environment, 34, 741-749, 1999.

• Fire Model Survey. International Survey of Computer Models for Fire and Smoke [on line]. Updated 2007 http://www.firemodelsurvey.com/EgressModels.html

• Kuligowski, E.D. & Peacock R.D. A Review of Building Evacuation Models. Technical note n. 1471. National Institute of Standards and Technology, Gaithesburgh, MD, 2005.

Main references

INTRODUCTION - INTRODUCTION - Pedestrian evacuation tools inventoryPedestrian evacuation tools inventory

http://www.firemodelsurvey.com/EgressModels.html

http://www.firemodelsurvey.com/EgressModels.html


perspective of model; perspective of users; modelling method structure of supply model; users’ behaviour; measurable outputs and visualisation

capabilities.

Classification of the models (Kuligowski and Peacock)

INTRODUCTION – INTRODUCTION – Classification of pedestrian evacuation toolsClassification of pedestrian evacuation tools


Perspective of modelPerspective of model

Explains how the model views the users

Individually: tracks the movement of individuals throughout the simulation and can give information about those individuals (e.g. their positions at points in time throughout the evacuation).

Globally: sees its occupants as a homogeneous group of people moving to the exits.



Perspective of usersPerspective of users

explains how the users (occupants) view the building

Individual: the user does not know the building’s exit paths and decides his/her route based on information from the floor, personal experience, and in some models, the information from the users around him/her.

Global: users are familiar with the building and automatically know their best exit path.



Modelling methodModelling method

three labels can be individuated:

Movement models: move users from one point in the building to another (usually the exit or a position of safety).

Behavioural models: incorporate users performing actions, in addition to movement toward a specified goal (exit).

Partial behaviour models: primarily calculate user movement, but begin to simulate behaviours.



Structure of supply modelStructure of supply model

how users move throughout the building

Coarse: divides the floor plan into rooms, corridors, stair sections, etc. and the users move from one room to another;

Fine: divides a floor plan into a number of small grid cells that the users move to and from;

Continuous: applies a 2D (continuous) space to the floor plans of the structure, allowing the users to walk from one point in space to another throughout the building.



Users’ behaviourUsers’ behaviour

None (N): only the movement aspect of the evacuation is simulated.

Implicit (I): attempt to model behaviour implicitly by assigning certain response delays or occupant characteristics that affect movement throughout the evacuation.

Conditional (rule-based) (C): assign individual actions to a person or group of users that are affected by structural or environmental conditions of the evacuation;

Artificial Intelligence (AI): attempt to simulate human intelligence throughout the evacuation;

Probabilistic (P): many of the rules or conditions are stochastic, allowing for the variations in outcome by repeating certain simulations;



TOOL COUNTRY PERSPECTIVE OF MODEL PERSPECTIVE OF USERS

ALLSAFE Norway Global Global

ASERI Germany Individual Individual

BuildingExodus United Kingdom Individual Individual

CRISP3 United Kingdom Individual Individual

EESCAPE Austria Global Global

EGRESS United Kingdom Individual Individual

EVACNET4 United States Global Global

EXIT89 United States Individual Individual

EXITT United States Individual Individual

FPETool United States Global Global

GridFlow United Kingdom Individual Individual

Legion United Kingdom Individual Individual

PathFinder United States Individual Global

PedGo Germany Individual Individual

PEDROUTE United Kingdom Global Global

SICURO Italy Individual Global

Simulex United Kingdom Individual Individual

STEPS United Kingdom Individual Individual

TIMTEX United States Global Individual

WAYOUT Australia Global Global



TOOL MODELLING METHOD STRUCTURE OF SUPPLY

USER BEHAVIOUR

ALLSAFE Partial Behaviour Coarse Implicit

ASERI Behaviour ContinuousRule-Based / Conditional

BuildingExodus Behaviour Fine

Rule-Based / Conditional

CRISP3 Behaviour FineRule-Based / Conditional

EESCAPE Movement Coarse None

EGRESS Behaviour Fine Conditional

EVACNET4 Movement Coarse None

EXIT89 Partial Behaviour Coarse Implicit

EXITT Behaviour CoarseRule-Based / Conditional

FPETool Movement Other None

GridFlow Partial Behaviour Continuous Implicit

Legion Behaviour Continuous Artificial Intelligence

PathFinder Movement Fine None

PedGo Movement / Partial Behaviour Fine Implicit

PEDROUTE Partial Behaviour Coarse Implicit

SICUROMovement / Partial

Behaviour Coarse None

Simulex Partial Behaviour Continuous Implicit

STEPS Movement / Partial Behaviour Fine None

TIMTEX Movement Coarse None

WAYOUT Movement Coarse None



TOOL VISU CAP. MAIN MEASURABLE OUTPUTS

ALLSAFE None Time to fire detection, to react and to interpret the situation Time for users to decide where to escape Time to evacuate a room or corridor and the building

ASERI 2D / 3D Evacuation time Detailed information on the structure and congestion situation that lead to delay Mean egress time, along wish their corresponding variances and confidence intervals

BuildingExodus 2D / 3D A data analysis tool (askEXODUS) allows to extract specific data from the output files

CRISP3 2D / 3D

Detailed information about each person at every time step Route information, fire conditions in certain locations Evacuation time Pictorial output

EESCAPE None Total evacuation time

EGRESS 2D Visualisation of congestion points Visualisation of bottlenecks Visualisation of merging flows

EVACNET4 None

Time to evacuate building, average time for evacuee to egress building, average number of evacuees per specified time period, number of successful evacuees

Number of evacuees that passed through a particular exit to safety List of arcs and number of people travelling through each arc Location of queues and time length of the queue Floor and node clearing time Building and destination evacuation profile Number of people not evacuated by a specified time

EXIT89 None

User movement table (track the time and corresponding node position of each user throughout the simulation)

Total evacuation time Number of occupants trapped Stair and floor-clearing times

EXITT None

Number of users out of the building Number of occupant trapped Total evacuation time Action of individual users at all time periods of the simulation



TOOL VISU CAP. MAIN MEASURABLE OUTPUTS

FPETool None Horizontal and stair travel time Time for all users to pass through exit doors

GridFlow 2D / 3D Outputs that can be imported into spreadsheet programs. Details about population in every space at every logging interval after each run. Detailed aspects of the buildings and users.

Legion 2D / 3D Usage maps (space, utilisation, density and speed, etc) Graphs on outflow characteristics Animations

PathFinder 2D

Number of people that have used an exit Statistics on times for people to exit from a given room Time for a stair and a floor to become empty Total evacuation time

PedGo 2D Text files that can be imported into spreadsheet programs (limited documentation on

this model)

PEDROUTE 2D / 3D Details of peak occupancy and average delay per passenger

SICURO 2D / 3D

Time to evacuate building, average time for evacuee to egress building, average number of evacuees per specified time period

Flow characteristics for each arc Location of queues and time period that arc had a queue Travel times for each path Destination evacuation profile Number of people evacuated by a specified time

Simulex 2D 2D visualisation of evacuation Overall evacuation time of all users reaching the exits Number of people passing through each exit

STEPS 2D / 3D Total evacuation rime Number of users in certain areas, planes and paths Number of people that have left the different fields versus time

TIMTEX None Total evacuation rime Individual floor clearing time

WAYOUT 2D Complete movement time Individual time when each compartment is evacuated




I. INTRODUCTION

II. PROPOSED APPROACHMacroscopic






MODELSMODELS

Macroscopic approach

DEMAND MODEL

described in terms of people occupying the building


SUPPLY MODEL

topological representation

– the network is represented using fundamentals of graph theory

representation of outflow conditions

– specific relationships

– functional dependence of speed, density and characteristic flow

MODELSMODELS



COMPUTATION OF EVACUATION TIME

Adaptation to the building of the IMO (International Maritime Organization) guidelines on evacuation analysis for passenger ships [IMO Ref. T4/4.01 MSC.1/Circ.1238 30 October 2007 - Annex 1]

MODELSMODELS




Schematization of escape routes as an hydraulic network where:

pipes → corridors and stairways,

valves → doors and restriction in general,

tank → public spaces.

MODELSMODELS



MODELSMODELS

Schematization of escape routes and data on occupants

Densitiy for each element

Geometrical characteristics

Occupants

Travel time for each escape

route

Occupants on each element

Specific flow

Speed evaluation

Travel time for each element

Escape routes

Evacuation time

Tevaq = max(T)




MODELSMODELS


WHY?

Simplified approach

Easily implementable on a worksheet

Uses easily available data

A simple tool can make DSS more attractive

Incentive in planning operations adopting more friendly DSS



I. INTRODUCTION


III. EXPERIMENTATIONOperations





Test siteTest site

EXPERIMENTATIONEXPERIMENTATION

Building type # Buildings Population

Residenziali 23 89

Public

School 1 155

Town hall 1 82

Court 1 7

Other 3 21

Mixed 28 262


Primary school

Evacuation area

Test site – location of the primary Test site – location of the primary schoolschool



1 1 1 1 1

1

1

1 1

1

1 1 2 3 4 5 6

7 8

9

10

11

27

28

29

30

31

32

33

35

34

36

38

37

39

41

40

42

44

43

47

45

46 50

48 49

53 51 52

54

55

56

57

58

59

60

61

62

63

64 65

112 113

114 115

116

117

Graph of the pedestrian network – first Graph of the pedestrian network – first floorfloor



1 1 1 1

1 1

1 1

1

1

1 1

1

12 13 14 16 17

18 19 20

21 22

23

24

66 67

68 69

70

71 72

75

76 77

78

79

80

81

82

83

86

85

84 87

88 89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107 108

1

109

111 110

26

1

15

74 73

118 119

120

CallFirst gathering place

Sec

on

d g

ath

eri

ng

pla

ce

121122

Graph of the pedestrian network – ground Graph of the pedestrian network – ground floorfloor



Demand - Primary schoolDemand - Primary school


Floor # centroid Functional type N.

ground

15 Secretary’s office 2

16 Secretary’s office 3

17 Headmistress office

1

first

1 Classroom 1a C 20

2 Classroom 1a A 20

3 Classroom 2a A 25

4 Classroom 2a B 24

6 Classroom 2a D 26

8 Classroom 2a C 21

People to be evacuated: 142

Location


1. Evacuation of the building

3. Transfer to the second gathering placeEvacuation phasesEvacuation phases

EXPERIMENTATION EXPERIMENTATION

2. Call of the pupils

Evacuation


Primary school

Evacuation area

SurveysSurveys

First gathering place

Second gathering placePositions of cameras



EVENT Clock time

Event notification 10:14:21

The Mayor reaches the operative center 10:23:48

The Mayor alerts the headmistress of the primary school 10:27:09

Alarm signal is activated at the town hall 10:37:31

Alarm signal is activated at primary school 10:39:26

All the pupils of primary the school reach the first assembly point 10:43:40

All the pupils of primary the school reach the second assembly point 10:48:45

Pupils begin to board 10:51:00

Bus starts from the assembly point 10:52:10

Recorded timesRecorded times




I. INTRODUCTION







Simulated phases:Simulated phases:

1. Evacuation of the building reaching first assembly point

2. Call of the pupils at first assembly point

3. Transfer to second assembly point




All occupants will begin evacuation at the same time and will not hinder each other;

Occupants will evacuate via the defined escape route;

Initial walking speed depends on the density of persons;

Evacuation is only in the direction of the escape route, and that there is no overtaking;

Full availability of escape arrangements is considered;

Hypotheses - Macroscopic approachHypotheses - Macroscopic approach


Phase DescriptionMeasur.

time Macro simul.

Sicuro (meso)

1 Evacuation of the building reaching first assembly point 4’14’’ 4’23” 5’47’’

2 Roll-call of pupils at first assembly point 3’05’’ 3’00” 2’58’’

3 Transfer to second assembly point 2’00’’ 4’00” 1’43’’

Total time 9’19’’ 11’23” 10’28”

0.00

2.00

4.00

6.00

8.00

10.00

12.00

min

utes

Measured time Macro simulation Meso simulation

Evacuation of the building reaching first assembly point Roll-call of pupils at f irst assembly point Transfer to second assembly point

ANALYSIS OF RESULTSANALYSIS OF RESULTS


Mesoscopic approach (Sicuro)

main advantage possibility to explicitly simulate queues and spill backs;

drawback necessity to use a specific software.


main advantage possibility to be easily implemented on a spreadsheet

drawback aggregate representation of flow conditions does not allow a detailed analysis

SOME CONSIDERATIONSSOME CONSIDERATIONS


PERSPECTIVESPERSPECTIVES

WEB WEB PortalPortal

ActionsActions Registration

Input data file on building

On line evacuation time computation

Return file with results

Potential usersPotential users Public administration

Schools

…

Massimo Di GangiMassimo Di GangiUniversità degli Studi di MessinaUniversità degli Studi di [email protected]@unime.it

A METHODOLOGY FOR AN A METHODOLOGY FOR AN AGGREGATE ANALYSIS OF AGGREGATE ANALYSIS OF EVACUATION OF EVACUATION OF BUILDINGSBUILDINGS

Corrado RindoneCorrado RindoneUniversità Mediterranea di Reggio Università Mediterranea di Reggio [email protected]@unirc.it

A methodology for an aggregate analysis of evacuation of buildings, di Massimo Di Gangi, Corrado...

Documents

Transcript of A methodology for an aggregate analysis of evacuation of buildings, di Massimo Di Gangi, Corrado...