EVACNET+: A computer program to determine optimal building evacuation plans

Fire Safety Journal, 9 ( 1 9 8 5 ) 211 - 220

211

EVACNET+: A Computer Program to Determine Optimal Building Evacuation Plans THOMAS M. KISKO and RICHARD L. FRANCIS

Department of Industrial and Systems Engineering, 303 Weil Hall, University of Florida, Gainesville, FL 32611 (U.S.A.)

SUMMARY

E V A C N E T + is a user-friendly interactive computer program that allows the modeling o f emergency building evacuations. A n E V A C N E T + model is a network consisting o f a set o f nodes connected by arcs. The nodes represent building components such as rooms, halls, landings, stairs and lobbies. The arcs represent the passageways between the building components. The program identifies optimal evacuation plans for user defined buildings.

INTRODUCTION

There are a number of reasons why the emergency evacuation of a building may be necessary. While the threat of smoke and/or fire is perhaps the most obvious reason, others may include the threat of an earthquake, a toxic or natural gas leak, a power blackout and/or elevator failure, a b o m b threat, and a civil defense emergency. Also, it is n o t unusual for large buildings to undergo regular practice evacuations, and it is natural to plan so that the practice will go well. In addition to their impact on building occupants, the foregoing reasons may well be of direct concern to building managers, building designers and architects, to public safety official s responsible for enforcing building safety codes, and to insurance companies. As a large building may contain thousands of people, it is certainly clear that accepting the responsibility for the successful emergency evacuation of a building is a substantial undertaking. One useful means of analyzing building evacuations is EVACNET+, a user-friendly interactive computer program that allows the modeling of emergency building evacuations. 0379-7112/85/$3.30

The program identifies optimal evacuation plans for user defined buildings.

QUESTION AND EVACNET+

ANSWER

SUMMARY

OF

The following questions and answers give an overview of EVACNET+. What is E V A C N E T + ? EVACNET+ is a user-friendly interactive computer program that models building evacuations. The program accepts a network description of a building and information on its initial contents at the beginning of the evacuation. From this information, EVACNET+ produces results that describe an optimal evacuation of the building. Each evacuation is optimal in the sense that it minimizes the time to evacuate the building. People are evacuated as quickly as possible. What is required as input to E V A C N E T + ? EVACNET+ requires a network description of a building and information about the initial placement of occupants at the beginning of the evacuation. The network is called an EVACNET+ network model. The network model consists of a set of nodes and arcs. The nodes of the network model represent building components such as rooms, halls, stairs, and lobbies. The initial contents (people) in each node must also be specified. The arcs represent the passageways between the building components. What does the n e t w o r k model look like? Figure 1 illustrates the graphic representation of an example network model. The rectangles represent the nodes, and the linking arrows represent the arcs. The numbers in the © Elsevier Sequoia/Printed in The Netherlands

212

6,1

LA13,25 3,4

q

WP2.3,92@1 ~ WP3.3,98 5,4

I 3,4

I twP1.2,211 [ @t I DSI.>1

l

.A2.1,,,711 i I14,4

~ 16,4

6,4

•o1,1,11oJ

HA3.1,191 3,1

6,4

HA1.1,66I I

W

3,3

;j DS2.1>

Fig. 1. E V A C N E T + representation of a three story building.

Figure represent the data required for the nodes and arcs.

What data needs to be supplied? F or each node, y o u will need to define a capacity. This is the upper limit on the n u m b e r o f people that can be contained in the building c o m p o n e n t the node represents. You can also (optionally) specify an initial contents o f a node. This is the n u m b e r of people in the ' n o d e ' at t he initiation of the evacuation. The initial c ont e nt s of a node will default to zero unless the user specifies otherwise.

For each arc, y o u will need to supply an arc traversal time and arc flow capacity. The traversal time is the num ber of time periods it takes to traverse the passageway the arc represents. The arc flow capacity is the upper limit on the n u m b e r of people that can traverse the passageway the arc represents during a time period. EVACNET+ breaks up time into time periods of fixed length. The length of each time period is user-definable. The default time period length is 5 seconds. Traversal times and flow capacities are based on this time period.

213 Where do I get this data? The EVACNET+ User's Guide provides information on supplying node and arc requirements to y o u r model. Figures, graphs, and tables in the guide provide a m e t h o d of estimating the parameters of y o u r model. You must determine the initial contents and capacities of each node. Therefore, y o u must estimate where the occupants will be at the beginning of the evacuation and collect information on the dimensions of the building. N o w that I have a model, how do I run EVACNET+? EVACNET+ is a user-friendly interactive computer package. The use of EVACNET+ requires no computer skills. EVACNET+ is EVACNET+ BUILDING EVACUATIONANALYSIS PROGRAM MASTER OPTION LIST CODE REQUESTEDACTION IL'IT---- ENTERNODE DEFINITIONS EA ENTER ARC DEFINITIONS LN LIST NODES LA LIST ARCS DN DELETE NODES DA DELETE ARCS SYS DEFINE OR REDEFINE SYSTEM ATTRIBUTES SAVE SAVECURRENTMODEL RM RETRIEVEDEFINED MODEL RUN RUN MODEL EXAM EXAMINERESULTS QUIT TERMINATEEXECUTION OF EVACNET HELP WHENEVERYOU HAVE QUESTIONS ENTER CODE OF REQUESTED ACTION

Fig. 2. EVACNET+ master o p t i o n list.

totally 'menu driven' in the sense that all that you have to do is select options and answer associated questions. When you run EVACNET+, you sit at a computer terminal and first enter the node and arc information representing the building of interest. After reviewing the input and making any necessary changes, the model can then be run interactively. Results can be displayed on a terminal or printed. If y o u wish, the network model m a y be modified and rerun. Figure 2 is the master menu that is presented to y o u when y o u run EVACNET+. What does E V A C N E T + do when it runs a model ? EVACNET+ takes the network model t h a t y o u provide and determines an optimal plan to evacuate the building in a " m i n i m u m " a m o u n t of time. This is done using an advanced capacitated network flow transshipment algorithm, a specialized algorithm used in solving linear programming problems with network structure. From y o u r point of view, all y o u do is supply the model, ask EVACNET+ to run it, and then examine the results. What kind o f results does E V A C N E T + provide ? The EVACNET+ results menu has 14 options for y o u to choose from in selecting specific results. Figure 3 is a display of this menu.

EXAM OPTION LIST FOR MODEL ID - "EXAMPLE THREE STORY BUILDING" I SUMMARYOF RESULTS: BASIC STATISTICS OF EVACUATION 2 DESTINATION ALLOCATION: NUMBER OF EVACUEESBY DESTINATION 3 TOTAL ARC MOVEMENT: TOTAL MOVEMENTTHROUGHAN ARC BY ARC 4 BOTTLENECKS: IDENTIFICATION OF BOTTLENECK ARCS 5 FLOOR CLEARING TIME: TIME TO CLEAR A FLOOR BY FLOOR NUMBER 6 UNCONGESTED TIMES: UNCONGESTED EVACUATION TIME BY NODE 7 NODE CLEARING TIME: TIME TO CLEAR A NODE BY NODE 8 BUILDING EVACUATION PROFILE: NUMBER OF EVACUEESBY TIME PERIOD 9 DESTINATION EVACUATION PROFILE: NUMBER OF EVACUEESBY TIME PERIOD FOR EACH DEST 10 NODECONTENTS PROFILE: PEOPLE WAITING AT END OF TIME PERIOD BY TIME PERIOD 11 NODECONTENTS SNAPSHOT:' PEOPLE WAITING AT END OF TIME PERIOD BY NODE 12 ARC MOVEMENTPROFILE: MOVEMENT THROUGHAN ARC BY TIME PERIOD 13 BOTTLENECKPROFILE: BOTTLENECK ARC INFORMATION BY TIME PERIOD 14 NON-EVACUEEALLOCATION: NUMBER OF NON-EVACUEESBY NODE ENTER OPTION NUMBER, OR "END" TO RETURN TO MAIN MENU

Fig. 3. Exam o p t i o n list.

What kind o f buildings can I model with EVACNET+ ? EVACNET+ has been designed to be flexible enough to model the evacuation of almost any structure representable as a network. This includes office buildings, hotels, skyscrapers, auditoriums, stadiums, retail establishments, restaurants, and schools. Entire structures or selected parts of a structure may be modeled. The cause of evacuation may be any reason requiring the quick removal of people from the building. How do I acquire EVACNET+? EVACNET+ is a public domain program that is written in ANSI 1966 standard FORTRAN. The program requires a minim u m of 100 000 bytes of memory. A version of EVACNET+ is also available for the IBM PC (or compatible microcomputer).

214 100'

ROOM 301

1

I I

I I

I

ROOM 302

I-*--15'--~1 '

:

.

ROOM 303

40' - - ! ~ 3 rd L E V E L

35'

.~!~- 10"-*-I O1~,

"~

144"'

ROOM201

I

I'

tl

ROOM I I

L=. '~ 5'

66" t

-

l

I

1--15'~-5'-1-5'-I,

T

2nd LEVEL

I

.

ROOM 101

6 o ' ~ GROUND LEVEL

0' I

5' 10' 15' 20' l I I / SCALE 1"=20'

Fig. 4. Layout of example building (10 ft. = 3.05 m).

Requests for EVACNET+ should be directed to the authors. H o w m u c h does it c o s t to run an E V A C N E T + model?

The cost of running an EVACNET+ model depends on three factors: (1) The charging algorithm and speed of your computer. (You may be lucky and have access to a 'free' computer.)

(2) The size of your model in terms of the number of nodes and arcs. (3) The number of time periods you allow EVACNET+ to simulate. There is no easy way of predicting exact costs on y o u r machine. However, experiments on representative computers have shown costs to range from pennies for simple example experiments to tens of dollars for experiments with an eleven story building.

215 I00'

i ,l

'I

W~ ROC /-/A

6

ROOM 302

IN

1

J"~---15'-"~ '~

40' - - 1 , 3rdLEVEL

35'

:!~ 10'-'~'-I

~. ~

~,,

P

#-,"P.,(.Z ROOM 201 (e~tJ

HAZ 2

~P2.2 ROOM 202

-'~ 66"

F 5'

~

zoz.z

2 nd LEVEL

I ~

1--16'~-6"-1-5'.1, 0' 5' 10' 15' 20' I

I SCALE

I

I

I

.

SO'GROUNDLEVEL

I

1"" = 2 0 "

Fig. 5. L a y o u t of example building with node outlines (10 ft. = 3.05 m).

AN EXAMPLE BUILDING The purpose of this section is to show an example of the application of EVACNET+. The building we will model is a hypothetical three story building. Figure 4 shows the layout of the building. The network model we saw in Fig. 1 is the corresponding EVACNET+ model. The example presented is the same one that is used in the EVACNET+ User's Guide. This presentation is meant to be a summary

of the EVACNET+ modeling process. Details are available in the User's Guide. The data for nodes and arcs that are presented here and in the User's Guide are founded on work done by Fruin, and by Pauls. Node definitions and data The first step in building an EVACNET+ model is to relate the physical building components to EVACNET+ node definitions. What we do is divide the building up into

216 TABLE 1 Node information for EVACNET+ example Node

Floor

Description

UA

LOS

APAD

NC

IC

HAl.1 HA2.1 HA3.1 LO1.1 WPI.1 HA1.2 LA1.2 LA2.2 SWl.2 SW2.2 WP1.2 WP2.2 HA1.3 LA1.3 LA2.3 SWl.3 SW2.3 WP1.3 WP2.3 WP3.3

1 1 1 1 1 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3

South hall North hall East hall Lobby Room 101 Hall West landing East landing West stairwell East stairwell Room 201 Room 202 Hall West landing East landing West stairwell East stairwell Room 301 Room 302 Room 303

330 585 95 620 2100 665 125 80 235 190 1690 1545 665 125 80 180 110 1690 740 785

D D D D C D D D D D C C D D D D D C C C

5 5 5 5 8 5 5 5 8 8 8 8 5 5 5 8 8 8 8 8

66 117 19 124 262 133 25 16 29 23 211 192 133 25 16 22 13 211 92 98

0 0 0 0 72 0 0 0 0 0 36 34 0 0 0 0 0 36 16 18

UA = usable area (sq ft) (x 9.3 X 10 -2 -+ m : ) LOS = level of service APAD = average pedestrian area occupancy (sq ft/person) (× 9.3 × 10 -2 -~ m 2/person) NC = node capacity (people) IC = initial contents (people) UA (sq ft) NC (people) = APAD (sq ft/person)

logical s e c t i o n s . T h i s d i v i s i o n is b e s t a c c o m plished by drawing the node boundaries right on the blueprint of the building. Refer t o Fig. 5 f o r a n e x a m p l e h o w w e d i d t h i s f o r our example building. I n Fig. 5 a n d c o r r e s p o n d i n g l y i n Fig. 1 we use a special n a m i n g c o n v e n t i o n to identify nodes. EVACNET+ node references consist of a two character n o d e t y p e , a

sequence number, and a floor number. For e x a m p l e ' W P 2 . 3 ' refers to Work Place n u m ber two on the third floor. T h e n e x t s t e p is t o d e t e r m i n e t h e n o d e data. T a b l e 1 is a list s u m m a r i z i n g t h i s p r o c e s s . As t h e T a b l e s h o w s , t h e n o d e c a p a c i t i e s are a f u n c t i o n o f t h e u s e a b l e a r e a associa t e d w i t h a n o d e a n d t h e level o f service e x p e c t e d f o r t h e n o d e . T h e level o f service is

EVACNET+ SUMMARYOF RESULTS FOR MODEL ID "USER GUIDE THREESTORY 34

TIME PERIODS TO EVACUATE BUILDING (

22

TIME PERIODS FOR UNCONGESTED BUILDING EVACUATION (

1.5 18.9

170 SECONDS)

CONGESTED FACTOR (RATIO OF BUILDING EVACUATION TIME TO UNCONGESTED BUILDING EVACUATION TIME) AVERAGE # OF PERIODS FOR AN EVACUEE TO EVACUATE (

6.2

AVERAGE NUMBEROF EVACUEES PER TIME PERIOD

212

NUMBER OF SUCCESSFUL EVACUEES

35 I

110 SECONDS)

MAXIMUM # OF TIME PERIODS ALLOWED FOR EVACUATION ( UNNECESSARY TIME PERIODS (

S SECONDS)

Fig. 6. Example output : summary of results.

95 SECONDS)

175 SECONDS)

217 BOTTLENECKS: IDENTICIATION OF BOTTLENECK ARCS FOR MODEL ID "USER GUIDE THREE STORY"

ARC SPECIFICATION

# OF TIME PERIODS ARC IS A BOTTLENECK

TOTAL BOTTLENECK MAGNITUDE

4 2 5 5 8 8 10 9 21 13 5

5 3 23 25 116 24 55 31 140 36 12

HAO3.0OI-DS02.001 LOOI.OOI-DSOI.O01 WPOI.OOI-HA03.001 WPOI.0OI-LO01. OOl HAOI.OO2-LA01.O02 HAOI.OO2-LA02.002 LAOI.002-SW01.O02 LAO2. OO2-SWO2.0O2 SWO2.OO2-HA03.OOl HAOI.OO3-LAOI.OO3 SWO2.OO3-LAO2.OO2

Fig. 7. Example output: bottlenecks.

measure of average area pedestrian occupancy. A level o f service 'D' corresponds to 3 to 7 square feet per person. (Source: Fruin's Pedestrian Planning Procedures Manual.) The node initial contents are based on expected room occupancies at the initiation of the evacuation. In this example the building has 212 occupants.

Arc definitions and data Arcs are relatively easy to define. They represent the possible flows of people from node to node. Table 2 lists the arcs as we defined them for this example. If the direction of flow is not known, a second oppoTOTAL ARC MOVEMENT: TOTAL MOVEMENTTHROUGHAN ARC BY ARC FOR MODEL ID "USER GUIDE THREE STORY"

ARC HAOI.OOI-DS02.001 HAO2.0OI-HA03.0Ol HAO2.0OI-LO01.OOl HAO3.001-DS02. OOl LOOI.0OI-DSOI. 0Ol LOOI.0OI-HA01.0Ol LOOI. OOI-HAO2.OOl WPOI.OO1-HAO2. OOl WPOI.OO1-HAO3.O01 WPOI.OOI-LO01.001 HAOI.OO2-LAOI.OO2 HAOI.OO2-LAO2.002 LAOI.002-SW01.OO2 LAO2. OO2-SWO2.OO2 SWOI.OO2-LOOI.OOl SWO2.OO2-HAO2.0Ol SWO2.OO2-HA03.OOl WPOI.002-HAOI.OO2 WPO2.002-HA01.002 HAOI.003-LAOI.OO3 HAOI.OO3-LAO2.0O3 LAOI.OO3-SWOI.OO3 LAO2.003=SWO2.003 SWOI.OO3-LA01.0O2 SWO2.003-LAO2. OO2 WPOI.003-HA01.OO3 WPO2.003-HA01.OO3 WPO3.003-HA01.0O3

# OF PEOPLE MOVING THROUGHARC

% OF NUMBER

0 3 27 64 148 0 0 17 19 36 46 24 85 55 85 13 42 36 34 39

0.00% 1.42% 12. 74% 30.19% 69.81% 0.00% 0.00% 8.02% 8.96% 16.98% 21.70% 11.32% 40.09% 25.94% 40.09% 6.13% 19.81% 16.98% 16.04% 18.40%

OF EVACUEES

31

14.62%

39 31 39

18.40% 14.62% 18.40%

31 36 16 18

14.62% 16.98% 7,55% 8.49%

Fig. 8. Example output: total a r c m o v e m e n t .

sitely directed arc can be added between the two nodes. This is the case in some of the halls o n the first floor of our example building. The first step in determining arc data is determining the Width Restriction (WR) associated with each arc. This is usually a doorway o f some sort between the nodes of the arc. For arcs linking stairwells to landings, or arcs linking hallways to hallways, we use the minimal width of the stairwell or hallway. BUILDING EVACUATION PROFILE NUMBER OF EVACUEES BY TIME PERIOD FOR MODEL ID "USER GUIDE THREE STORY" TIME PERIOD

# OF EVACUEES

I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

O O 0 9 9 9 9 9 11 11 5 0 0 5 5 5 5

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

5 5 5 5 8 8 8 8 8 8 8 8 8 8 8 6 6

:

EACH * :

REPRESENTS i PERSON : : : : :

********* ********* ********* ********* ********* *********** *********** ***** ***** ***** ***** ***** ***** ***** ***** ***** ******** ******** ******** ******** ******** ******** ******** ******** ******** ******** ******** ****** ******

i TIME PERIOD = 5 SECONDS

Fig. 9. Example output: building evacuation profile.

218 TABLE 2 Arc i n f o r m a t i o n for EVACNET+ example Arc

LOS

WR

AFV

DC*

DIST

AS

TT~"

HAI.I-DS2.1 HA2.1-HA3.1 HA2.1-LOI.1 HA3.1-DS2.1 LOI.I-DSI.1 LOI.I-HAI.1 LOI.I-HA2.1 WPI.I-HA2.1 WPI.I-HA3.1 WPI.I-LOI.1 HA1.2-LA1.2 HA1.2-LA2.2 LA1.2-SWl. 2 LA2.2-SW2.2 SWl.2-LOI.1 SW2.2-HA2.1 SW2.2-HA3.1 WP1.2-HA1.2 WP2.2-HA1.2 HA1.3-LA1.3 HA1.3-LA2.3 LA1.3-SWl.3 LA2.3-SW2.3 SWl.3-LA130 SW2.3-LA2.2 WP1.3-HA1.2 WP2.3-HA1.3 WP3.3-HA1.3

D D D D D D D D D C D D D D D D D C C D D D D D D C C C

24 48 108 24 60 48 108 48 24 60 24 24 54 32 54 24 24 72 72 24 24 54 32 54 32 72 48 48

18 18 18 18 18 18 18 18 18 13 18 18 12 12 12 12 12 13 13 18 18 12 12 18 12 12 13 13

3 6 14 3 8 6 14 6 3 6 3 3 5 3 5 2 2 7 7 3 3 5 3 5 3 6 5 5

45 60 60 10 15 65 60 55 45 45 45 45 30 30 40 75 45 20 20 45 45 30 30 30 30 20 20 20

215 215 215 215 215 215 215 215 215 240 215 215 110 110 110 110 110 240 240 215 215 110 110 110 110 240 240 240

3 4 4 1 1 4 4 4 3 3 3 3 4 4 5 9 5 1 1 3 3 4 4 4 4 1 1 1

SPTP = seconds per time period (assumed 5 s/time period) DIST = distance (ft) (x 0.305 ~ m) LOS = level of service AS = average speed ( f t / m i n ) (x 5.1 x 10 -3 -~ m / s ) TT = traversal time (time periods) WR = width restriction -- minimal width (in) (x 2.5 x 10 -2 ~ m) AFV = average flow volume (people/ft-min) (x 5.5 x 10 -2 -+ people/m-s) DC = dynamic capacity (people/time period) 1 ft 1 min *DC (people/time periods) = WR X x AFV X - X SPTP 12 in 60 s DIST 60 s 1 t T T (time periods) = x -x - AS min SPTP DIST 1 ( * t U s i n g SI units as indicated DC = WR x AFV x SPTP: TT = as- x s--F~

As w e d i d w i t h n o d e s , w e t h e n s e l e c t a l e v e l o f s e r v i c e . T h i s t h e n r e l a t e s t o a n average flow volume and an average speed. As T a b l e 2 s h o w s , t h e d y n a m i c c a p a c i t y is now a function of the width restriction and the average flow volume. The arc trav e r s a l t i m e is a f u n c t i o n o f t h e a v e r a g e s p e e d and the average distance traveled. Distances are estimated to be the median distance that people would be required to travel

)

in passing from one node of an arc to the other.

Construction of graphic model F i g u r e 1 is t h e g r a p h i c m o d e l o f t h e b u i l d ing. T h e m o d e l , o f c o u r s e , has t h e s a m e nodes and arcs that we just defined. Special symbols are used for each node type to make it e a s i e r t o v i s u a l i z e t h e b u i l d i n g . W e h a v e transferred the appropriate node and arc

219 data to the graphic model. This Figure is the only d o c u m e n t now needed to input our model into EVACNET+. It is also a very useful d o c u m e n t when results are analyzed. R u n n i n g the model

After initiating the execution of EVACNET+, the nodes and arcs are entered into the computer with the 'EN' and 'EA' options of Fig. 2. Then the model could be run by typing ' R U N ' and the results examined by typing 'EXAM'. Figures 6 through 9 are actual o u t p u t produced by the EXAM option. After looking at the results of an EVACNET+ experiment, changes may be made and the experiment may be rerun. ASSUMPTIONS OF EVACNET+ EVACNET+ takes the network model that y o u provide and determines an optimal plan to evacuate the building in a 'minimum' amount of time. This is done using an advanced capacitated network flow transshipment algorithm, a specialized algorithm used in solving linear programming problems with network structure. The formulation of an EVACNET+ model requires certain assumptions to be made. These assumptions can lessen the realism of the model. Understanding these assumptions is necessary in order to produce valid results. The principle assumptions of EVACNET+ include: (1) EVACNET+ is a linear modeling system. Arc capacities and arc traversal times do not change over time, and DO NOT DEPEND ON THE ARC FLOWS. Smoke and fire spread cannot currently be modeled over time. (2) EVACNET+ does not model behavioral aspects. The only actions that are modeled are those that lead to achieving the minimum evacuation time. Sometimes y o u can represent behavioral aspects by your choice of data, e.g., specifying a lower bound on the number of people to go to a destination. (3) EVACNET+ is based on a global viewpoint; not an individual viewpoint. This means that in achieving an optimal evacuation plan, EVACNET+ has the capability to 'see' everything over all time periods.

In an actual evacuation, it is more likely that evacuees will view an evacuation from their individual perspectives. One major use of EVACNET+ can be to inform potential evacuees and/or the floor wardens of globally optimal building evacuation plans. (4) The evacuation plan produced by EVACNET+ is only one of potentially many possible optimal solutions. You can be assured that there is no other plan with a smaller building evacuation time. However, the details of the routing of evacuees can vary significantly between t w o optimal solutions. (5) EVACNET+ counts people. It does not trace individual people. (6) Ioitial locations of people must be known. (7) The building modeled by EVACNET+ is what the user defines it to be. CONCLUSIONS EVACNET+ is a powerfultool that will allow fire safety engineers to make objective decisions about the evacuability of buildings. It is designed to find an optimal evacuation plan that can be used to make relative comparisons with other alternatives, or to give evacuation planners more insight in developing actual evacuation procedures. We believe, if you are aware of the assumptions EVACNET+ is based on and do not rely upon results blindly, that you will find EVACNET+ a very useful quantitative tool for analyzing emergency evacuation problems.

ACKNOWLEDGEMENTS The development of EVACNET+ was sponsored in part by the Center for Fire Research of the National Bureau of Standards under Grant number NB81NADA2057. BIBLIOGRAPHY 1 R. L. Francis and L. G. Chalmet, Network Models

for Building Evacuation: A Prototype Primer, NBS-GCR-81-316, National Bureau of Standards, Center for Fire Research, Washington, DC 20234, March 1981.

220 2 L. G. Chalmet, R. L. Francis and P. B. Saunders, Network models for building evacuation, Management Science, 28 (1) (January) (1982) 86 105. Also reprinted in Fire Technol., 18 (1) (February) (1982) 90 - 113. 3 Basic Building Code, Building Officials and Code Administrators International, Inc., Homewood, IL, 1980. 4 John J. Fruin, Designing for Pedestrians-- a Level-of-Service Concept, Ph.D. Dissertation, The Polytechnic Institute of Brooklyn, June 1970. 5 John J. Fruin, Pedestrian Planning and Design, Metropolitan Association of Urban Designers and Environmental Planners, New York, 1971 (out of print). 6 Life Safety Code, NFPA 101, National Fire Protection Association, Quincy, MA, 1981. 7 J. Pauls, Evacuation of high rise office buildings, Buildings, 72 (5) (1978) 84 - 88. 8 J. Pauls and B. Jones, Building evacuation: research methods and case studies, in J. Canter (ed.), Fires and Human Behavior, J. Wiley, New York, 1980, pp. 227 - 249. 9 J. Pauls, Building evacuation: research findings and recommendations, in J. Canter (ed.), Fires

10 11

12

13

14

and Human Behavior, J. Wiley, New York, 1980, pp. 251 - 275. J. Pauls, Building design for egress, J. Architectural Education, (Summer) (1980) 38 - 42. J. Pauls, Effective-width model for crowd evacuation flow on stairs, Sixth International Fire Protection Seminar, Karlsruhe, West Germany, September21 - 24, 1982, Vol. 1, pp. 295 - 306. A Pedestrian Planning Procedures Manual, Report No. FHWA-RD-79-46, (NTIS: Springfield, VA 22161), Charles F. Scheffey; Dir., Office of Research. Prepared for Federal Highway Administration, Offices of Research and Development, Environmental Design Control Division, Washington, DC 20590. M. Y. Roytman, Principles o f Fire Safely Standards for Building Construction, published for the National Bureau of Standards, Washington, DC, by Amerind Publishing Co., Pvt. Ltd, New Delhi, India, 1975. J. Templer, Stair Shape and Human Movement, Ph.D. Dissertation, Columbia University, New York, 1974. (A Text, Designing for Pedestrians, to be published by the Whitney Library of Design, is in preparation.)